Y. Tamarin-Protective Coatings for Turbine Blades (2002).pdf

© 2002 ASM International. All Rights Reserved. Protective Coatings for Turbine Blades (#06738G)

Protective Coatings for Turbine Blades

Y. Tamarin

Materials Park, Ohio 44073–0002 www.asminternational.org

www.asminternational.org


Copyright 䉷 2002 by ASM International威 All rights reserved No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the copyright owner. First printing, September 2002

Great care is taken in the compilation and production of this book, but it should be made clear that NO WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE GIVEN IN CONNECTION WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot guarantee that favorable results will be obtained from the use of this publication alone. This publication is intended for use by persons having technical skill, at their sole discretion and risk. Since the conditions of product or material use are outside of ASM’s control, ASM assumes no liability or obligation in connection with any use of this information. No claim of any kind, whether as to products or information in this publication, and whether or not based on negligence, shall be greater in amount than the purchase price of this product or publication in respect of which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE THE EXCLUSIVE AND SOLE REMEDY OF BUYER, AND IN NO EVENT SHALL EITHER PARTY BE LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES WHETHER OR NOT CAUSED BY OR RESULTING FROM THE NEGLIGENCE OF SUCH PARTY. As with any material, evaluation of the material under end-use conditions prior to specification is essential. Therefore, specific testing under actual conditions is recommended. Nothing contained in this book shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters patent, copyright, or trademark, and nothing contained in this book shall be construed as a defense against any alleged infringement of letters patent, copyright, or trademark, or as a defense against liability for such infringement. Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International. Prepared under the direction of the ASM International Technical Book Committee (2001–2002), Charles A. Parker, Chair ASM International staff who worked on this project included Steve Lampman, Manager of Book Acquisitons; Bonnie Sanders, Manager of Production; Carol Terman, Production Project Manager; and Scott Henry, Assistant Director of Reference Publications. Library of Congress Cataloging-in-Publication Data Tamarin, Y. Protective coatings for turbine blades / Y. Tamarin. p. cm. Includes bibliographical references and index. 1. Aircraft gas-turbines—Blades. 2. Protective coatings. 3. Gas-turbines—Materials. I. ASM International. II. Title. TL709.5.B6 T36 2002 629.134⬘353—dc21 2002027690 ISBN: 0–87170–759–4 SAN: 204–7586 ASM International威 Materials Park, OH 44073–0002 www.asminternational.org Printed in the United States of America Cover: Test blades (shiny) in a low-pressure turbine. Source: Advances in Turbine Materials, Design, and Manufacturing, Proceedings of the Fourth International Charles Parsons Turbine Conference, The institute of Materials, 1997




Contents Foreword....................................................................................................................... v Preface......................................................................................................................... vi Chapter 1: Introduction.................................................................................................... 1 Chapter 2: Choosing Optimum Coatings for Modern Aircraft Engine Turbine Blades................. 5 Conditions of Turbine Blade Operation ............................................................... 5 Requirements Imposed on Turbine Blade Coatings................................................ 8 Principles of Choosing Coatings for Aircraft Engine Turbine Blades ........................ 8 Causes of Coating Failures on Aircraft-Engine Turbine Blades.............................. 10 Chapter 3: Technological Processes for Deposition of Protective Coatings to Turbine Blades .... 25 Diffusion Coatings ........................................................................................ 25 Overlay Coatings .......................................................................................... 38 Chapter 4: Phase Composition and Structure of Coatings on Superalloys ............................... 55 Phase Composition and Structure of Diffusion Coatings....................................... 55 Phase Composition and Structure of Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al Alloys and Overlay Coatings Made of Them...................................................... 69 Phase Composition and Structure of Overlay Coatings......................................... 71 Chapter 5: Phase and Structural Changes in Coatings during High-Temperature Tests .............. 79 Changes of Phase Composition and Structure in Diffusion Coatings....................... 80 Changes of Phase Composition and Structure in Overlay Coatings at HighTemperature Tests ......................................................................................... 87 Chapter 6: Turbine Blade Coating Protective Properties ...................................................... 97 Protective-Properties Evaluation Methods .......................................................... 97 Heat Resistance of Aluminides and Alloys for Overlay Coatings ..........................100 Heat Resistance of Coated Superalloys.............................................................106 Resistance of Overlay-Coated Alloys to Hot Corrosion .......................................109 Resistance of Coated Superalloys to Hot Corrosion ............................................113 Chapter 7: The Effect of Protective Coatings on the Mechanical Properties of Superalloys .......119 Thermal Expansion Coefficients and Elasticity Modulus of Coating Alloys ............120 Mechanical Properties of Coating Alloys ..........................................................124 Thermal Stresses in Superalloy Coatings ..........................................................128 Effect of Coatings on High-Temperature Strength of Superalloys..........................134 iii


iv / Protective Coatings for Turbine Blades

Effect of Coatings on Fatigue Strength of Superalloys ........................................138 Effect of Coatings on Thermal Fatigue of Superalloys ........................................142 Coating Effect on Thermomechanical Fatigue of Superalloys ...............................150 Chapter 8: Electron Beam Thermal Barrier Coatings..........................................................161 Ceramics for EB Evaporation .........................................................................162 Main Features of TBC Deposition Technique ....................................................165 Thermophysical Properties of Condensed Ceramics ............................................175 TBC Ceramic Layer Durability .......................................................................180 Chapter 9: Some Principles of Strength Designing for Turbine-Blade Protective Coatings ........195 Calculation of Stresses and Strains in Coatings..................................................195 Thermal Barrier Coatings ..............................................................................203 Appendix....................................................................................................................211 Index .........................................................................................................................217

iv


ASM International is the society for materials

engineers and scientists, a worldwide network dedicated to advancing industry, technology, and applications of metals and materials. ASM International, Materials Park, Ohio, USA www.asminternational.org This publication is copyright © ASM International®. All rights reserved. Publication title

Product code


06738G

To order products from ASM International: Online Visit www.asminternational.org/bookstore Telephone 1-800-336-5152 (US) or 1-440-338-5151 (Outside US) Fax 1-440-338-4634 Mail

Customer Service, ASM International 9639 Kinsman Rd, Materials Park, Ohio 44073-0002, USA

Email [email protected]

American Technical Publishers Ltd. 27-29 Knowl Piece, Wilbury Way, Hitchin Hertfordshire SG4 0SX, In Europe United Kingdom Telephone: 01462 437933 (account holders), 01462 431525 (credit card)

www.ameritech.co.uk Neutrino Inc. In Japan Takahashi Bldg., 44-3 Fuda 1-chome, Chofu-Shi, Tokyo 182 Japan Telephone: 81 (0) 424 84 5550 Terms of Use. This publication is being made available in PDF format as a benefit to members and customers of ASM International. You may download and print a copy of this publication for your personal use only. Other use and distribution is prohibited without the express written permission of ASM International. No warranties, express or implied, including, without limitation, warranties of merchantability or fitness for a particular purpose, are given in connection with this publication. Although this information is believed to be accurate by ASM, ASM cannot guarantee that favorable results will be obtained from the use of this publication alone. This publication is intended for use by persons having technical skill, at their sole discretion and risk. Since the conditions of product or material use are outside of ASM's control, ASM assumes no liability or obligation in connection with any use of this information. As with any material, evaluation of the material under end-use conditions prior to specification is essential. Therefore, specific testing under actual conditions is recommended. Nothing contained in this publication shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters patent, copyright, or trademark, and nothing contained in this publication shall be construed as a defense against any alleged infringement of letters patent, copyright, or trademark, or as a defense against liability for such infringement.

Protective Coatings for Turbine Blades Y. Tamarin, p1-3 DOI: 10.1361/pctb2002p001

Copyright © 2002 ASM International® All rights reserved. www.asminternational.org

CHAPTER 1

Introduction THE DAMAGE TO blade surfaces caused by oxidation and hot corrosion results in considerable deterioration of the mechanical properties of blades and shortens their service lives. That is why protection of turbine blade surfaces against damage has been a common problem. Solving this problem is critical for designing cost-effective and reliable aircraft engines and stationary gas turbine units. The idea to apply a layer with protective properties to the surface of a nickel superalloy was first practiced in the 1960s and found a wide application area. Since then, aircraft engine turbine blades with protective coatings have been used. In the 1960s, extensive research into the properties of various coatings demonstrated that diffusion aluminide coatings had the best protective properties on turbine blades made of nickel superalloys. Owing to their properties, such as high oxidation resistance, high stability when deposited on nickel superalloys, and a favorable combination of physical and mechanical properties, the diffusion aluminide coatings have been the predominant type of aircraft engine turbine blade coatings used for several decades. The simplicity of the technique used in forming diffusion coatings on the blade surfaces in the powder mixtures containing free or bonded aluminum and aluminum halides contributed to their most common use. Translation of diffusion aluminide coatings in a commercial practice and the work on improvement of their protective properties and further development of their deposition techniques started concurrently. This further development was the result of the need for improved properties of protective coatings in the media containing sulfur compounds. Diffusion aluminide coatings with the increased chromium content in their outer layer (chromizing-aluminizing method) had found application in the protection of aircraft engine blades

suffering from hot corrosion. Alloying aluminide coatings with silicon (silicification in combination with aluminizing) turned out to be one more technique that allowed the improvement of coating resistance to hot corrosion and oxidation. Since the 1970s, both modifications of the coatings have been used for protection of aircraft engine turbine blades. The studies aimed at modifying diffusion aluminide coatings are ongoing. Some of the coatings designed (e.g., Pt-Al) are used widely in aircraft engines and stationary gas turbine units. Designing new-generation engines with increased inlet gas temperatures resulted in the shortening of the surface lives of the blades protected with diffusion coatings. Low ductility of diffusion coatings and high stresses generated in cooled blades cause the formation of numerous thermal fatigue cracks in the outer layer of diffusion coatings on the blade surfaces. In updating diffusion coatings, designers seek to improve primarily the resistance of the coatings to thermal fatigue. The problem of improving the coating property in question arises from the limited technological feasibility of changing its composition, structure, and consequently, its physical and mechanical properties. The studies aimed at increasing ductility of diffusion aluminide coatings resulted in the gradual decrease of the aluminum contents of the coatings from 34 to 36%, typical of the coatings used in the 1960s, down to 20 to 24%, typical of the coatings used since the 1970s. This trend, however, resulted in a decrease in oxidation resistance. High labor consumption and expenditure of energy on the formation of diffusion aluminide coatings on turbine blades in powder mixtures and the need for the use of halides in the technological process stimulate the seeking of alternative techniques for forming diffusion coatings

2 / Protective Coatings for Turbine Blades

on turbine blades. Coating deposition from slurry allowed considerable improvement of the process conditions under which the coatings were formed on the turbine blades. The method has experienced a wide application area in aircraft engine manufacturing. Further technique improvements resulted in the development of various gas aluminizing methods. These methods allowed less employment of powder mixtures and higher efficiency of gas transfer of aluminum to blade surfaces. By now, the techniques of blade aluminizing that eliminate completely the need for covering the blades with powder mixtures have been developed and are enjoying application. In this case, aluminum transfer to the blade surface occurs by circulation of gaseous aluminum halides formed in a special reaction zone of the unit. A further advancement in the practice of diffusion coating deposition was the use of electricarc evaporation of aluminum alloys and their transfer to the blade surface in the form of plasma. In this case, the need for the use of halides and powder components is eliminated completely, and it opens up the possibility for the accurate control of diffusion coating thickness and its location on the blade surface. This technique is used for forming the diffusion coatings modified with silicon, yttrium, and other elements on turbine blades. At the same time, it is noteworthy that all technological processes of diffusion coating deposition used in turbine blades production form the coatings with similar protective properties. As a rule, the choice of a certain technique for coating application depends on requirements for its quality, technological equipment available, and personnel skills. Although diffusion coatings have been enjoying wide application, they do have a limitation: They do not allow deliberate changes of their properties to meet the requirements imposed by specific blade service conditions. This restriction exists because all diffusion coatings always are based on Ni-Al and Ni3Al aluminides with typical combinations of physical-chemical and mechanical properties. This drawback follows from the principle of diffusion coating formation due to interaction of aluminum transferred to the surface from an aluminizing medium and the components of a nickel superalloy. It cannot be overcome by modernization of coating compositions and techniques of their deposition. That is why, as a following step in developing protective coatings for turbine blades, it was quite natural

to apply some new principles of their formation, such as deposition of the alloys, which compositions met complex requirements to protective coatings for turbine blades most adequately. Design of electron-beam, electric-arc, plasma methods, and equipment for their realization triggered the change to the new principles of development and application of protective coatings to turbine blades. Since the 1980s, some variants of overlay coatings have found application for the protection of turbine blades. Unlike diffusion coatings, overlay coatings admit the control of their composition, thickness, and combination of physical-chemical and mechanical properties within wide ranges. It allows the choice of overlay coatings, which have the properties optimal for certain engine types and their service conditions. New principles of coating deposition opened up new possibilities for purposive improvement of coating compositions and variation of their properties. However, a wide range of possible variants of coatings compositions and techniques for their deposition presents the problem of making an adequate choice in coatings for certain engine types. For this reason, development of calculation methods for determining stresses and strains in the coatings on turbine blades and thus their service lives under conditions of high thermal stresses further advanced the work on overlay coatings. To calculate the values in question, it is necessary to have certain information on chemical and mechanical properties of the coatings within a wide range of probable compositions. Basically, a new stage of progress in the area of turbine blades protection has been achieved with designing thermal barrier coatings. The outer layer is formed of stabilized zirconium oxide-base ceramics. This coating type allows not only protection of a coating surface against corrosion damage as all previously designed coatings did, but it was also the first to allow protection of the structure of a superalloy in cooled blades against changes caused by exposure to high temperatures and, hence, against its softening. An apparently simple idea of designing a ceramic layer on the blade surface called for extensive research and development for its realization that involved designing techniques for production of special ceramics, their evaporation with electron beam, special test methods, and calculations of service lives for thermal barrier coatings on turbine blades.

Introduction / 3

As is well known, metal structural materials properties depend on their chemical compositions and structures. Control of metal materials compositions and structures makes it possible to produce a great variety of structural alloys with required properties. Similarly, protective coatings properties depend on their chemical compositions and structures. And yet, the approaches worked out for structural materials hardly remain valid for coatings because the relationships connecting chemical compositions, structures, and properties in thin coating layers do not reveal themselves clearly. Besides that, a “coating-superalloy” system is a dynamic one. The chemical and phase compositions of a coating, as well as its structure, vary continuously under blade service conditions at elevated temperatures. The initial coating properties on the blade and those it has at any moment of its operation in the turbine may be quite different. Since the first use of protective coatings on the turbine blades of different designations, hundreds of papers that deal with their properties have been published. However, to create a database covering the papers published and storing the information required for the proper choice of coatings and for calculations of the service lives of coated blades, one should overcome the problems of coatings identification and lack of data on chemical and phase compositions and structures of the coatings studied. No unified standard methods for research into coatings properties are available. In most cases it makes impossible comparison and classification of the test results obtained. Special databases should be created to solve the problems of description and systematization of the steadily increasing number of protective coatings for turbines. The databases should provide systematized, accumulated information on the coatings properties studied and on the experience of their use in practice. Systematization of the results presented by laboratories and firms, which deal with the problems of turbine blades protection in the form of databases, may

allow further simulation of changes in coatings properties, as well as formalization of “composition-structure-property” relationships. Besides that, the characteristics required for coated blade service life calculations can be derived. This book presents knowledge accumulated by the author over many years of his work on the research and development of diffusion and overlay protective coatings for aircraft engine turbine blades. The book includes detailed descriptions of some technological processes for diffusion and overlay coatings deposition, together with compositions of the coatings used for turbine blades protection, coatings structures after their deposition on the blade surfaces, and their changes during high-temperature testing. The results of a number of tests for coatings and samples of the alloys used for coating deposition are discussed also. Some chapters deal with thermal barrier coatings, technological features of their deposition, and test results, as well as with the principles of calculations of stresses and strains in the coatings on blades. The “composition-property” regression equations derived on the basis of research into aluminide alloys of Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al systems used for deposition of overlay coatings are given in the Appendix. They allow simulation of aluminide alloys properties. These simulations cover the compositions of both as-deposited protective coatings and those after their long-term operation on turbine blades. In some cases the simulations obtained may be used for assessment and prediction of coatings properties. Many coatings dealt with in this book are used currently for turbine blades protection. The information on technological processes and coatings properties presented in the book may be useful to the specialists who work in the field of turbine blades protection, to the scientists interested in research into materials behavior at hot corrosion, and to the students whose specialty is protection against hot corrosion.



CHAPTER 2

Choosing Optimum Coatings for Modern Aircraft Engine Turbine Blades COATED TURBINE BLADES boast service lives of 2 to 5 times those of their uncoated counterparts. However, coating protection effects are manifested only if the coating has been chosen correctly, both in terms of blade design and engine running conditions.

Conditions of Turbine Blade Operation The main parameters determining the operating conditions of turbine blade surfaces of aircraft gas-turbine engines are the gas temperature at the turbine inlet; the pressure, speed, and composition of the gas flow; the stresses due to the effect of the centrifugal and gas dynamic forces; and the irregularity of the temperature field. The gas temperature at a turbine entry is the most important parameter determining the specific power, specific weight, and efficiency of an aircraft engine. During the last 20 years, the mean mass temperature of the gas at a turbine entry increased from 1200 to 1300 K to 1700 to 1800 K. Together with the turbine temperature growth, the pressure ratio of the compressor has grown significantly as well; it has reached 20 to 30. The combined growth of the temperature and the pressure ratio results in significantly more intensive heat flows in turbines; they make 1 to 2 MW/m2 for modern engines. One of the main properties of aircraft gas-turbine engines is a significant number of variable conditions when the level of temperatures and stresses differs significantly from those measured in stationary conditions. Statistical analysis shows that for aircraft engines, about 20% of

the flaws developed at operation are flaws resulting from the cyclic temperature variations (Ref 1). Figure 2.1 shows the operating diagram of an airliner for a typical medium-range flight. Every 1.5 h flight has four corresponding cycles of condition variation; that is, 10,000 h of engine life correspond to 27,000 cycles of variation in flight conditions. Every time an engine reaches takeoff power, the nonuniformity of the temperature field in the turbine blades increases significantly. Running the engine under cruising conditions does not eliminate the effect of nonuniform temperature and stress fields on the blades. The number of varied powers for warplane engines, especially fighters, is significantly greater than that for passenger aircraft engines. The surface layers of aircraft-engine turbine blades during operation contact the combustion products of the aviation fuel. This is hydrocarbon-based (98 to 99%), with the remaining 1 to 2% including sulfur, nitrogen, oxygen compounds, and trace metals as well as their compounds. A widespread impurity in fuels is vanadium oxide (V2O5), whose content depends on the deposit of the source oil. The fuel sulfur has the form of its various compounds, mercaptanes, sulfides, disulfides, and thiophenes. The total sulfur content of an aviation fuel is in the range of 0.05 to 0.25%. Aviation fuels usually get impurities from the environment and from the production process, resulting from corrosion and rubber decomposition and so forth. Typical impurities of jet fuels are as follows: C, H, S, N, Fe, Si, Sn, Ca, Mg, Pb, Na, Ba, Ti, Ni, Al, Cu, Zn, Mn, and O.


The quantity and the composition of the impurities depend on local operation conditions, on the fulfillment of requirements to fuel controls and purification, and on aircraft fueling conditions (Ref 2 to 4). In a combustion chamber, the process of fuel combustion is realized at the temperature of 2273 to 2473 K. Hydrocarbon combustion is a process of their oxidation by the oxygen in air, resulting in carbon dioxide and water, which may contain some carbon monoxide (CO), H2, methane (CH4), and hard carbon particles. The complete combustion of 1 kg of fuel requires 15 kg of air. The total factor of air excess in combustion chambers compared to the amount necessary for complete fuel combustion is 3.5 to 4.5; it may change in dependence of the flight conditions. The specific fuel consumption for modern engines provides 0.75 to 0.90 kg/kgf of thrust. The high-temperature gas flow at the turbine inlet contains mainly oxygen, carbon dioxide, water vapors, and some amounts of CO, H2, CH4, hard carbon particles, compounds of sulfur, and metals present in fuels. Various chemical compounds may come into the gas flow from the environment. These are sea salts from flight over sea regions and various compounds that are exhausts of industrial enterprise. As a rule, they contain sulfur, salts of alkaline metals,

Fig. 2.1

Operating diagram of an airliner engine. Source: Ref 1

and other compounds, depending on the industry situated in the flight area. The temperature pattern of the gas flow at the turbine inlet is highly inhomogeneous; the inhomogeneity is usually caused by the design of a combustion chamber, by the design and layout of burners, and by other factors determining the flow-gas dynamics and the perfection of the fuel combustion process. The measurements of the gas temperature before the nozzle guide vanes show that a difference between the extreme gas temperatures at a nozzle block inlet may be 400 to 500 C. The nozzle vanes of a high-pressure (HP) turbine situated just behind the combustion chamber are the most heated elements of an engine. The temperature of the cooled nozzle vanes is 200 to 300 C lower than that of the gas flow, but the thermal inhomogeneity of the gas at a turbine inlet results in some vanes working at temperatures differing significantly from the average ones. The temperature measurement of vanes has shown, that the thermal pattern of the vanes leading edges in the middle section differs by 200 to 250 C. The thermal inhomogeneity of the nozzle vanes results in significantly differing rates of flaw development in vanes situated in different parts of an engine relative to the combustion chamber baffles and burners. The gas temperature pattern inhomogeneity is observed not only by the hot-duct circumfer-

Choosing Optimum Coatings for Modern Aircraft Engine Turbine Blades / 7

ence, but by height, too. The characteristic temperature curve by the vane height shows an area of maximal temperature in the vane mid-part and lower temperatures at end areas. Nevertheless, other temperature patterns may also be observed. The greatest difference between the middle and peripheral areas of the vanes is 100 to 150 C. The temperature gradient value grows at the transition to the gas high-temperature conditions. The vane and blade thermal state depends on the engine power; it changes from 450 to 500 C to 1000 to 1100 C within a few seconds at the engine transition from idle power to take-off conditions. The temperature inhomogeneity of the vane and blade surfaces increases significantly under the transient conditions of startup, acceleration, and run. The maximal thermal stresses of the vanes, especially at their surfaces, within the coating areas are observed at these very conditions. The computed data that characterize variations in cooled-vane temperatures at the engine startup show that the difference in the temperatures between the vane back and trailing edges may reach 500 to 600 C. The computation results on the thermal stresses corresponding to the above temperature patterns show that their values attain 400 to 500 MPa (Ref 5). The analysis of the strains in a vane caused by such thermal stresses reveals that some areas of a vane are in a state of plastic flow. The level of thermal stresses and strains in a vane surface layer may exceed significantly the mean level of the vane stresses due to the presence of the temperature gradients along the cooled-vane wall section and due to different physical and mechanical properties of the protective coating and superalloy. The accumulation of thermal variations at operation results in an accumulation of residual strains, which, in their turn, result in thermal fatigue cracks on the vane surfaces, first on the leading and trailing edges. These cracks are the most widespread flaws of vanes. Analysis of the vane operating conditions shows that the combined effect of high temperatures, gas dynamic forces, and thermal stresses in vanes causes cracks due to thermal fatigue, burnouts, and warping. The higher temperature of the engine gas resulted in higher temperature gradients in the vane surface layers, and the use of cast nozzle blocks caused higher stiffness of restraint of individual vanes. As a result, the thermal fatigue cracks have become the most widespread flaw of the vanes during tests and

operation, despite the use of efficient film methods for cooling. One of the critical elements of an aircraft gasturbine engine is its turbine blade. It is its thermal state that determines the maximal admissible gas temperature in the turbine and the reliability and life of the turbine and the engine in whole. The blades are affected by centrifugal and gas dynamic forces causing tension, bend, and torsion of the blades. The high temperatures and stresses, the unstable conditions of heating and loading, and the possibility of resonance vibrations make a blade one of the most complicated elements of an engine. In contrast to the nozzle vanes, the gas temperature field before the turbine blades is more uniform. The high rotation speed of a blade results in a more homogeneous thermal state of individual blades, if compared with the vanes. The temperature gradient by the blade surface depends on the gas-temperature curve by height and by the specific operation of the blade cooling system. The highest temperature area is, as a rule, in the upper third of a blade airfoil portion; this is explained by the need to place it within the area with the lowest centrifugal loads. The temperature distribution by the turbine second stage blades is even more homogeneous, though the non-cooled, second-stage blades of some engines have higher average temperatures than the ones of the first stage. The limitations due to high-temperature strength (heat resistance) of the nickel-based superalloys and the necessity to improve the parameters of modern aircraft engines pushed the design of sophisticated and efficient cooling systems; they permitted lower average temperatures of a blade wall by 300 to 400 C, compared to the gas temperature. The higher temperature and compression rates in modern engines result in greater temperature irregularities in the blades. When the gas temperature increases by 100 C, the temperature gradient on the surface of a cooled blade increases by 40 to 50 C; the compression rate effect is similar. The temperature pattern of the modern cooled blades is complex enough: the areas of high temperature, 1000 to 1100 C, whose size is often 10 to 30 mm2, are alternated by the average (900 to 950 C) and low (750 to 800 C) temperature areas. The temperature gradient by the surface and wall section of the blades causes high thermal stresses, reaching maximal magnitudes at engine operation under non-cruise conditions (Ref 6 to 8). A blade is affected by tensile stresses caused by centrifugal forces. Their levels are not usu-


ally higher than 200 to 250 MPa. A blade is also affected by the bending stresses caused by gas dynamic forces. These stresses are determined by the difference in pressures at the concave (pressure) and convex (back) surfaces of a blade. The combined static stresses caused by the external forces affecting turbine blades are distributed irregularly along a blade. The highest tensions are observed near the blade root, the lowest ones exist at the end of a blade airfoil portion. The section with the least static strength security margin or with the maximal exposure to flaws due to the long-term static load is usually between the maximal stress and maximal temperature sections (Ref 9 to 11). The vibration loads are also critical for the blades; these loads result from the inhomogeneity of the gas flow hitting the turbine wheel rim. The low-frequency vibrations are caused by the flow thermal inhomogeneity due to the presence of burners and pillars; the high-frequency ones result from the finite number of nozzle vanes. The resonance vibrations are the most widespread cause of the surface fatigue cracks and rupture of the turbine blades. The level of variable stresses in the blades reaches 80 to 100 MPa (Ref 12).

Requirements Imposed on Turbine Blade Coatings Any coating deposited on turbine blades (or vanes) must offer protection within a specified period of service life against destructive attacks of high-temperature corrosion and erosion, when the said components are exposed to a flow of fuel combustion products containing aggressive ingredients and solid particulate matter. With this in mind, the coating must meet the following requirements: ●

It must withstand hot corrosion, oxidation, and erosion when placed into a flow of gas whose parameters are similar to those of turbine gases. ● It must safely withstand the static and alternate stresses applied to the blade surface; to this end the coating must have the requisite combination of strength and ductility. ● It must show good stability and not be destroyed by interaction with the substrate. ● It must not degrade the blade material mechanical properties.

Even after a long service life, coated blades and vanes must have better mechanical properties than their uncoated counterparts due to protecting the blade surfaces from damage and softening. In addition to the requirements to metal coatings listed previously, the ceramic coatings (thermal barrier coatings, or TBCs) must ensure lower average wall temperatures in cooled blades and protect superalloys against softening. TBCs must level off the temperature over the blade surface and reduce thermal stresses during engine transient running. The techniques involved in depositing protective coatings must guarantee that the coatings have the required composition, thickness, and structure, and, as such, the requisite set of physical-chemical and physical-mechanical characteristics. Of utmost importance is the need to maintain these characteristics depositing protective coatings on blades under the commercial conditions. Three technological principles are applied to form coatings on the surfaces of aircraft engine turbine blades: ●

Chemical-thermal treatment (aluminizing) of blade surfaces in the media, containing aluminum or aluminum with additives of other elements (such as chromium, silicon, yttrium, etc.). A diffusion aluminide coating is built up as aluminum or its compounds interact with the superalloy surfaces. At present, diffusion aluminide coatings are used to protect 80 to 90% of all aircraft turbine blades. ● Depositing the overlay coatings by evaporating special alloys under vacuum and condensing vapor or plasma with coating components on blade surfaces ● Forming coatings from powders by arc plasma spraying

Principles of Choosing Coatings for Aircraft Engine Turbine Blades The requirement to attain highly protective coatings (i.e., adequate resistance to hot corrosion and oxidation) conflicts with the demand to preserve high mechanical properties in the “coating-superalloy” system. Until recently, engine tests have remained the main method of examining the properties of and choosing optimized coatings for blades. How-


ever, the high cost of systematic studies of coating protectiveness using this approach has fueled the desire to seek shorter equivalent trials. Under these conditions, laboratory research into physical-chemical and physical-mechanical coating properties and their relation to blade service life facilitates significantly the selection of optimal chemical compositions of the coatings, reduces the number of their field tests, and promises high cost efficiency of the study. Shown in Fig. 2.2 are the key parameters to consider in selection of coatings for turbine blades. The starting points to be taken into account when choosing a coating of optimum life expectancy include: ●

Experience in applying protective coatings on blades running under similar conditions ● Analysis of causes of coating damage on blades after long-time tests or service operation ● Presumable alterations in running conditions of newly designed blades (changed parameters of gas flow, values of thermal stresses, running conditions, etc.) ● Results of research into physical-mechanical properties of coatings and their effects on the main properties of the superalloy from which the blades are made (resistance to thermal fatigue, endurance, high-temperature strength) Of prime significance for the protectiveness of a coating is its resistance to hot corrosion and oxidation in the temperature range of 600 to

Fig. 2.2

Scheme of the choice of a coating for aircraft engine turbine blades

1200 C, that is typical of the running conditions of modern-engine turbine blades. The mechanism of coating damage in this range can vary depending on the temperature and composition of gas in turbine, blade surface temperature, and design. Coatings are chosen from the laboratory results of coating tests for heat resistance and hot corrosion. Actual engine running conditions can result in blade surfaces being exposed to temperatures either below or higher than 900 C. Moreover, the demands on coating life expectancy at temperatures above 900 C may conflict with life expectancy requirements imposed on blades working in corrosive media at temperatures below 900 C. When choosing coatings, due attention should be paid to possible application of engines in coastal regions or in areas with air heavily laden with industrial impurities. Should the blade damage be mostly caused by hot corrosion, then coatings with high resistance to this type of corrosion should be preferred. Alternatively, multilayer coatings can be used, given that individual layers of such coatings offer high resistance to oxidation and hot corrosion. Coating thickness is undoubtedly a very important characteristic of its protectiveness; coating resistance to hot corrosion and oxidation is directly proportional to the thickness of coating used. It should be borne in mind, however, that the thickness of coatings applied to turbine blades are restricted, by both the probable decrease in mechanical properties of the blades with thick coatings and by technological problems involved in depositing thick coatings. Design problems must also be taken into account (e.g., the reduction of the open flow area in the turbine). As overlay coatings add to the mass of blades they are applied to, hence, the centrifugal loads on the blades and turbine disk grow. This must be counted when thick coatings are used. Variable thickness coatings are most advantageous; a thicker layer is deposited on the area that is most likely to be attacked by corrosion. Other areas are thinner-coated. No correlation between coating thickness and thermal fatigue resistance has been noted in overlay coatings up to a thickness of 120 lm. A positive correlation seems more likely in thicker coatings prone to formation of various defects, which, in turn, are conducive to the emergence of thermal fatigue cracks. When using coatings thicker than 120 lm, the blade fatigue limit must be taken into consideration.


Coating composition and thickness can be determined once the required service life has been specified. Once the coating composition has been determined, one must then consider how long the coating of given composition will stay steadily on the blade surface after long exposure. Coating protectiveness is limited by diffusion into the alloy of those elements that control resistance to oxidation and hot corrosion. The diffusion progress can cause the decrease in aluminum, chromium, and cobalt contents of the coating. It can call for the use of special methods for retarding diffusion processes by introducing barrier layers based on phases that are stable with respect to both the coating and the alloy. On the other hand, as coating elements diffuse into the alloy, a zone of poorer heat resistance is formed under the coating. It is a factor that, in thin-walled blades, leads to reduction of the margin of safety. No coating lacking the requisite stability can be used for turbine-blade protection. The life of a coated turbine blade depends on its mechanical characteristics and the stresses on the surface. Thermal fatigue cracks have become the commonest cause of service life reduction and discarding of blades. Calculation of the stresses and strains in turbine blade construction is mandatory in coating selection for modern turbine engines. Engine testing is the final step in coating selection and estimating useful life on turbine blades. Defects in a coating on blade surfaces may be caused by hot corrosion and oxidation, low resistance of blade construction to thermal

Table 2.1 Alloy composition for overlay coatings deposited on turbine blades Chemical composition, wt% Alloy

Ni

Co

Fe

Cr

Al

Y

Co20Cr12AlY … Base … 18–22 11–13 0.2–0.6 Co23Cr12AlY … Base … 22–24 11–13 0.2–0.6 Co26Cr9AlY … Base … 25–28 8–10 0.3–0.6 Co25Cr4AlY Base 24–26 3.5–5 0.3–0.6 Co30Cr6AlY 0–2 Base 28–32 5–7 0.3–0.6 Co28Cr10FeY … Base 8–12 26–30 … 0.1 Co22Ni23Cr12AlY 18–25 Base … 22–24 11–13 0.3–0.5 Co8Ni23Cr12AlY 6–10 Base … 22–24 11–13 0.2–0.6 Ni20Co20Cr12AlY(a) Base 18–22 … 18–22 11–13 0.2–0.6 Ni8Co20Cr12AlY(a) Base 6–10 … 18–22 11–13 0.2–0.6 Ni20Cr12AlY(a) Base … … 18–22 11–13.5 0.2–0.6 Ni20Cr5AlY Base … … 18–22 4–6 0.2–0.6 Ni36Cr5AlY … Base … 35–38 4–6 0.2–0.5 Fe25Cr5AlY … … Base 23–26 4–6 0.2–0.5 Fe22Ni24Cr5AlY 20–25 … Base 23–25 4–6 0.2–0.5 (a) Composition used in aviation. Source: Ref 13, 14

fatigue, poor high-temperature strength and endurance of the substrate, or by a combination of these factors. In order to assess life expectancy of coated blades in long-term tests, their temperature schedule must be as close to the actual engine running conditions as possible. The matter is that in the course of engine-equivalent quick tests, where the turbine blades are running at higher than actual service temperatures, the processes occurring on the blade surfaces, as well as between the coating and superalloy, may significantly differ from those occurring at lower temperatures. Elevation of temperatures during equivalent quick tests may be misleading due to reduced condensation of aggressive components from the gas flow on the blade surfaces and suggest longer-than-actual service lives. Alternatively, such tests may degrade coating resistance owing to sharp acceleration of diffusion processes and rapid destruction of protective oxides. Analyzing the results of engine tests pinpoints the coating-failure mechanism and enables, if necessary, appropriate modifications to the composition, thickness, and/or design of multilayer coatings. Chemical compositions of some alloys used for coating deposition on turbine blades are presented in Table 2.1 (Ref 13, 14). It is noteworthy that the coating composition depends on the currently used technological process of its deposition. It can differ from that of the initial alloy. In this book, the coatings are marked with the grades of the alloys used for their deposition. Otherwise, when the coatings are marked according to their chemical compositions, the respective notes are supplied. Chemical compositions of the materials used for applying diffusion coating to aircraft turbine blades are presented in Table 2.2 (Ref 15 to 21). Chemical compositions of the superalloys used for research into coating properties are presented in Table 2.3 (Ref 22, 23).

Causes of Coating Failures on AircraftEngine Turbine Blades The investigation of coating damage, detected on the blades of engines after a long time in service, is an important stage in the choice of an optimum coating. Careful investigation of the nature of damage and the sources of origin will enable the designer to choose a coating that has


Table 2.2

Chemical compositions of the materials used for forming diffusion coatings

Method of coating

Grade

Chemical composition of alloy (mixture), % mass

Pack calorizing(a) Pack calorizing(a) Pack calorizing(a) Pack calorizing under vacuum(b) Slurry aluminizing(c)

Al Al Al AlCr Al AlSi Al AlSiY

Granules of alloy of Fe–(35–50%)Al; 2% NH4Cl Granules of alloy of Fe–(70–75%)Al; 2% NH4Cl Powder 98%Al, 2% NH4Cl Powder (7–8%Al; 35–40%Cr; Al2O3); 0.3–0.4% NH4Cl-balance Powder of (100% Al) in colloxiline solution of amilacetate Powder of (Al–5–50%Si) in colloxiline solution of amilacetate Alloy: 100% Al Alloy: Al base; 4.0–15.0% Si; 0.6–2.0% Y

Electric arc vacuum deposition(d)

(a) Ref 15, 16. (b) Ref 17, 18. (c) Ref 16, 19. (d) Ref 20, 21

Table 2.3

Superalloys used for aircraft turbine blades Chemical composition(b), wt%

Alloy (a)

Cr

Ti

Mo

W

Re

Ta

Al

Co

Hf

Nb

B

Zr

C

Ni10CrWMoCo JS6K JS6U VJL12U(c) JS6F(c) JS30 JS26(c) JS32 MAR-M-200 MAR-M-002 CMSX-4

10.0 11.0 8.5 9.5 5.5 7.0 5.0 5.0 9.0 9.0 6.5

… 3.0 2.5 4.5 1.0 1.9 1.0 … 1.9 1.7 1.0

10.0 4.0 1.8 3.0 0.9 0.7 1.1 1.0 … … 0.6

5.0 5.0 10.0 1.4 12.0 11.8 11.7 8.3 12.5 11.0 6.0

… … … … … … … 4.0 … … 3.0

… … … … … … … 4.0 … 2.5 6.5

4.5 5.5 5.5 5.5 5.3 5.2 5.8 6.0 4.7 5.5 5.6

5.0 4.3 9.5 9.5 9.4 8.5 9.0 9.0 10.0 10.0 9.0

… … … … 0.9 0.8 … … … 1.5 0.1

… … 1.0 0.8 1.6 0.9 1.6 1.5 1.8 … …

… … … … … 0.015 0.015 0.015 0.015 0.015 …

… … … … … 0.05 0.05 0.05 0.05 0.05 …

0.10 0.15 0.15 0.15 0.11 0.15 0.15 0.15 0.14 0.15 …

(a) JS alloys are designed in All-Russian Institute of Aviation Materials (VIAM). (b) Average values. (c) These alloys contain 1.0% V. Source: Ref 22, 23

Depth of high-temperature corrosion in Ni20Cr12AlY overlay coating vs. test temperature at testing for heat resistance (●); 200 h hot corrosion (o), and engine test, 100 h (D)

Fig. 2.3

the highest life under certain running conditions and point out ways to improve existent coatings and develop alternative systems.

Two temperature regimes, one at 600 to 850 C and one above 1050 C (Fig. 2.3), have been singled out as the ones where coatings are subjected to intensive damage. This result was achieved while analyzing aluminide-based coatings for their protectiveness. At 600 to 850 C, hot corrosion was noted to develop in coatings owing to the presence of aggressive compounds constituted mostly of sulfur and vanadium. The temperature boundaries and intensity of corrosion will depend on a number of factors: composition of protective coating, quantity of aggressive compounds on the surface, and duration of tests. During engine tests, the intensity of hot corrosion will be affected (besides temperature) by the gas pressure in the turbine and by design features of the blades and vanes which can facilitate or impede condensation of aggressive compounds on different areas of the surface. In the interval from 900 to 1000 C, only modest oxidation has been noted due to the high protective properties of oxides Al2O3 • Cr2O3 forming on the surface of this type of coating. In laboratory heat resistance tests in the temperature range of 900 to 950 C, the diffusion aluminide-based coatings, 50 lm thick, with


aluminum content of 34 to 36%, have manifested their ability to protect superalloys from oxidation during 10,000 h. At temperatures above 1000 C, the rate of coating oxidation rises. At the same time, the diffusion processes between the coating and alloy speed up, resulting in the aluminum and chromium content of the coating dropping appreciably and the rate of oxidation growing. A diffusion coating 50 lm thick (34 to 36% Al) has exhausted its protective ability in 300 to 500 h when tested in the laboratory at 1100 C for heat resistance. The trends that have been noted during laboratory tests are also observed on engine tests. Hot corrosion develops when engines are run continuously. The rate of coating damage is generally lower when testing coated blades in the range of 700 to 850 C (in comparison with laboratory tests). This phenomenon is attributed to lower amounts of aggressive compounds from the gas flow in contact with the blade surfaces. Oxidation processes at 900 to 950 C go on turbine blades more intensively than during heat resistance tests. Generally, coatings are oxidized during high-temperature cycles of engine running. At such temperature, the life expectancy of a diffusion coating 50 lm thick (34 to 36% Al) spans between 6000 and 8000 h when the engine is run at maximum load. However, life expectancy dramatically drops when the temperature on blade surface is elevated up to 1050 to 1100 C. When the blades were rig-tested at 1050 C, their diffusion-type protective coating failed in 80 to 100 h during engine running at high temperatures. Hot Corrosion of Turbine Blades. Since aviation fuels contain insignificant quantities of sulfur and vanadium, they do not destroy coatings through hot corrosion below service lives of 5,000 to 10,000 h. The environment is the main source of aggressive components entering the hot channel of aircraft engines. This applies, first of all, to sea regions, where a diverse variety of compounds contained in the seawater get into the engine hot channel. Another source of aggressive components is the waste gas of industrial plants. The location of test rigs or the region of engine operation may be the cause of this mechanism of coating failure. Investigations carried out have revealed two modes of hot corrosion on turbine blades of aircraft engines. Those are (a) pit and (b) uniform corrosion modes (Ref 16). Pit corrosion emerges

and progresses first of all on those areas of the blade surface that work continuously at 620 to 760 C (Fig. 2.4). Such surfaces are located on the blade airfoil near the fir-tree root and shroud platform. As early as in the first dozen hours of testing or operation, numerous pits may appear on the said surfaces. These pits, filled with corrosion products and looking like small nodules, gradually, as time goes on, spread towards the blade central region. Pit corrosion proceeds on most-lengthy working cycles of engine operation. As the analysis of turbine work parameters has shown, the higher the gas flow velocity, the lesser the corrosion damage. Variation of pressure and air surplus coefficient produces no significant effect on blade corrosion. Of much greater significance are the blade design features and the local aerodynamics. For the same temperatures on the suction and pressure surfaces, the pressure surface will suffer more from pit corrosion than the suction surface where corrosion attack is much less severe. The poorer the fuel quality, the more severe is the corrosion. Research has been conducted into the microstructure of diffusion coatings in corrosion-inflicted areas on the blades removed from different engines, either rig-tested or service-run, in diverse conditions. The research has proved that the microstructure was the same in all cases studied. The oxides penetrating into coating outer zones are encircled with light edging, consisting of Ni3Al compound (Fig. 2.4b). The longer the time of testing or running, the greater the depth of penetration of oxides into the coating. This, however, continues only until inner zones with high contents of chromium-base phases are reached. At this point, the oxide penetration is lowered and the inner zone of the coating retards corrosion penetration into basic metal. The analysis of the chemical composition of pit corrosion products has shown that, apart from the coating elements, they contained from 5 to 10% V and from 0.5 to 1.0% S. The higher the blade working temperatures, the less the pit corrosion. Thus, no pit corrosion was observed on turbine blades from engines that had been running at 900 C for as long as 8,000 to 10,000 h. Elevation of temperatures, higher compression rates, and greater gas flow velocities in engine turbines have altered the outer appearance of corrosion-stricken blades (Fig. 2.5). This should be attributed, first of all, to the elevation of gas temperatures up to 1330 to 1400 K and


to the resulting changes in the composition, fluidity, and thermodynamic properties of aggressive compounds. The condensation of aggressive compounds is helped by a high temperature gradient between gas flow and surface of cooled blades. Higher gas flow pressure and velocity enhance erosion effects upon the corrosion products by briskly sweeping the latter off the blade surfaces.

Fig. 2.4

Instead of formation of individual corrosion pits, uniform corrosion has been noted. The latter affects large areas of the blade surface, particularly the pressure surface and the leading edge exposed to temperatures of 780 to 850 C. Consisting of aluminum, chromium, and nickel oxides, the corrosion products penetrate deeply into the coating and the superalloy (Fig. 2.6). A unique feature of hot corrosion is that a thin

(a) Pit corrosion location vs. blade surface temperature (service time around 4000 h); (b) Blade surface in the zone of pit corrosion. 20⳯; (c) Microstructure of diffusion coating in pit corrosion zone. 300⳯. (Light phase around oxides is Ni3Al.)


Fig. 2.5

Fig. 2.6

Uniform corrosion zone location vs. blade surface temperature (test time around 3000 h).

Microstructure of JS6U-alloy blade surface under layer of oxides. 5000⳯

layer of metal with sulfur precipitates emerges underneath the corrosion products in the hightemperature zone. Such microstructure was ob-

served in all blades stricken with hot corrosion. As for vanadium, its content in corrosion products is either insignificant or nil. The uniform corrosion is typical of continuous engine running. Uniform hot corrosion is noted on engine blades running at high temperatures (gas temperature from 1650 to 1700 K). This should be attributed to effective cooling systems providing, on certain surfaces (e.g., in the lower zone of the blade airfoil), favorable temperature conditions for aggressive compounds to condense. The choice of a coating to protect engine blades from hot corrosion within the required life period is generally based on the results of laboratory tests conducted in line with unified methods. “Coating service life” is understood as time to coating failure caused by hot corrosion. The correctness of coating choice and prediction of its service lifespan is totally dependent on how sound the selected laboratory methods are. Accelerated tests may be fraught with such serious errors that the results of the tests cannot be used to predict lives. Moreover, one cannot rule out errors involved in comparative analysis of the properties of different coatings. Life expectancies of coatings can be predicted from the results of tests described in “Phase and Structural Changes in Coatings during HighTemperature Tests,” Chapter 5. The approximate life expectancy of a new coating in terms of its corrosion-resistance can be inferred from


the rate failure of an existing engine-run coating and the results of comparative hot-corrosion laboratory tests of both the existing and new coatings. The suggested approach to inferring blade-life expectancy is simplified and substantiated by test experience in using the laboratory methods and then applying the data generated in selecting the most appropriate coatings for aircraft engines. The best way of improvement here (i.e., in estimating the coating life expectancies) is not to sophisticate further the relevant calculations but, rather, to select the right test methods and to accumulate data on the properties of different coatings. The problem of accumulation of data on hot corrosion can be solved, at reasonable cost, by using the alloys, imitating possible compositions of coatings, and applying statistical methods of processing the test results. Figure 2.7 illustrates the results of life-expectancy calculations for 100 lm thick coatings tested at 850 C under laboratory conditions for the Ni-Co-Cr-Al-Y alloys. The same data can easily be obtained for other systems too. Additional investigations will be required to infer from these data life expectancies for coated blades. Ceramic coatings provide new opportunities to lengthen blade-life expectancy in a hot-corrosion environment. The effects of such coatings can be dual: ● Adequate protection of metal coating from aggressive components with proportional lengthening of life expectancy of a double-

Service life of Ni-Co-Cr-Al-Y coatings 100 lm thick calculated from data on laboratory-tested Ni-Co-CrAl-Y alloys in GZT ash, 850 C, 200 h

Fig. 2.7

layer coating (ceramic-metal coating). In practice, the pertinent solution may be to increase coating thickness. ● The ceramic coating alters the thermal state of blade surface conditions (elevates the surface temperature). This reduces the amount of aggressive compounds condensing on the surface. It also enhances the fluidity of such aggressive compounds, thus facilitating their removal by gas flow. This factor has been confirmed by tests. When tested in the laboratory, the ceramic coatings had no significant effect, though tests on power units showed the increase in life expectancy of blades. Further investigations and development of special ceramic compositions will be necessary to bring ceramic coatings on a par with the requirements imposed on coatings for hot-corrosion protection. Oxidation of Turbine Blade Operating at Temperatures above 1000 C. As the temperature of a blade surface increases from 750 to 850 C to 900 to 950 C, the life expectancy of coatings rises. In this temperature interval, diffusion coatings guarantee adequate protection of aircraft-engine turbine blades within 6000 to 8000 h. Increasing blade surface temperature up to 1050 to 1100 C results in appreciable decreases in diffusion coating life expectancy. At such temperatures the coatings fail mostly due to consumption of aluminum for the formation of the protective oxide. Besides, the aluminum content was reduced owing to interdiffusion of aluminum and nickel between the coating and superalloy. As Fig. 2.8 shows, a turbine blade, after being tested for 100 h, displayed full oxidation of the coating (blade wall thereafter) in 30 to 40 h when exposed to a temperature of 1100 C. Figure 2.9 illustrates oxidation rates in a diffusion coating deposited on the blade from superalloy JS6U. The coating has been built up by slurry aluminizing and it had an “aluminum reserve” of 40 g/m2. After being tested for 100 h, the coating outer zone was found to be fully destroyed at a narrow strip on the leading edge 2 to 4 mm broad and 10 to 20 mm high when exposed to the maximum temperature of 1050 C. The consumption rate of aluminum for protective oxide formation in the leading edge region was 0.5 g/m2 • h. With the temperature at the blade inlet edges elevated to 1100 C, the diffusion coating with “aluminum reserve” of 60


Fig. 2.8

External appearance of a blade with diffusion coating after 100 h of testing

(1) Oxidation depth of diffusion and (2) Ni20Cr12AlY overlay coatings on the leading edge of a blade during testing

Fig. 2.9

to 70 g/m2 was found to be fully oxidized in 30 to 40 h of tests. Figure 2.9 illustrates the depths to which oxidation has penetrated in the overlay coating Ni20Cr12AlY, 120 lm thick, with aluminum reserve 200 g/m2. Measurements were taken on the leading edges of turbine blades. When running under maximum loads, the temperature on the leading edge was 1050 C. Test duration is indicated as a summed value for all running. In comparison with diffusion coatings, the overlay coatings had greater aluminum reserves, which made the life expectancy 4 times as long as that of diffusion coatings. The designed average consumption of aluminum and chromium for protective oxide formation in Ni20Cr12AlY coating varied within 0.5 to 0.6 g/m2 • h. As for the phase transformations and structural changes in the coating Ni20Cr12AlY occurring on the leading edge of the blade, they are similar to those typical of laboratory tests for heat resistance. After 140 h, a thin layer of csolid solution is formed on the coating surface, controlling the kinetics of oxidation processes. As test time gradually increases to 300 h, the csolid solution covers a greater zone and its boundary spreads further into the coating. The diffusion of aluminum, and to a much lesser extent chromium, from the coating central zones toward the surface makes the coating composition change in the direction NiAl Ⳮ c→Ni3Al Ⳮ c→c. In this case 10 to 15% of aluminum and chromium diffuse under the coating into the zone of superalloy interaction. As the Ni20Cr12AlY coating is becoming oxidized, the surface of a blade is deformed in a “rippled” or “folded” fashion (Fig. 2.10). This phenomenon can be attributed to gas flow pressure on the coating that is soft at these temperatures. Its deformation is affected by compression stresses occurring in the coating at temperatures above 1000 C. Should the working temperatures on turbine blades with the overlay coating Ni20Cr12AlY be raised to 1100 to 1150 C, the coating protection capability reduces dramatically, provided that the described mechanism of coating damage prevails. Figure 2.11 shows typical structures of Ni20Cr12AlY coating on turbine blades due to overheating from 1000 to 1150 C in certain zones. The oxidation mechanism remains the same on all surface areas. Its nature can be explained by consumption of certain elements (aluminum standing the first) for oxide formation. The rate


of aluminum and chromium consumption varies with the temperature level. While in zones at 1000 C, the coating has oxidized to an insignificant depth (Fig. 2.11a); in the zone at 1150 C, phase transformations associated with csolid solution formation have been completed. Aluminum content there has been reduced to 3%, and the processes of internal oxidation have started (Fig. 2.11d). Aluminum and chromium levels consumed to form the interaction zone do not exceed 10 to 15%, even at maximum temperatures. The heat resistance of diffusion and overlay coatings is discussed against the background of laboratory tests, enabling protective properties of different coatings to be compared, given that correct methods and proper criteria have been chosen. However, the test data, pertinent to specific changes of mass (g/m2) cannot be used in calculations of coating life expectancies on blades. For the same mechanism of oxidation, the extent of blade damage from oxidation will be many times that of laboratory samples.

Fig. 2.10

In laboratory tests and in tests on the blades, the protective function is performed not by the coating, but rather by the oxides that are formed on its surface. In fact, the coating acts as a reserve that supplies aluminum and chromium for the formation of oxides if they are damaged. In this case it will be the “aluminum reserve” (mAl) that emerges as a criterion for coating life expectancy (Ref 16). This has been substantiated by data, which showed that all aluminum contained in a diffusion coating on superalloys is eventually consumed for oxide formation. The “aluminum reserve” criterion correlates well with the results of both laboratory and field tests. “Aluminum reserve,” mAl in a diffusion coating is found from the following function: n

mAl ⳱ K •

兺 Xi • qi • hi i⳱1

in g/m2 where Xi is aluminum mass fraction in a definite zone of coating; qi is the density of

External appearance of Ni20Cr12AlY coating on a turbine blade after 500 h of testing. (a) Leading edge. (b) Pressure surface. 10⳯


that zone of coating; hi is the thickness of that zone of coating; and K is the alloying factor. “Aluminum reserve” in the coating equals the increase of specific mass of samples at alitizing,

and it can be found out experimentally. The effect of alloying the diffusion coatings can be counted by introducing the coefficient “K ”, which for aluminized coatings equals unity (K

Microstructure of Ni20Cr12AlY coating in the blade zones with gradual increasing of working temperatures from (a) 1000 C to (d) 1150 C. 500⳯. In (a), (b), and (c), a dark phase in the coating is c-solid solution; a light one is Ni3Al. In (d), a dark phase in the coating is oxides; a light one is c-solid solution.

Fig. 2.11


⳱ 1), while for the modified coatings alloyed with chromium, silicon, and yttrium, it takes the value KAl ⳱1.1ⳮ1.6. The “aluminum reserve” criterion can also be used in comparing life expectancies of overlay coatings, having the same or nearly the same system of alloying. When determining aluminum reserve in Ni-(Co)-Cr-Al system, Cr concentration must be taken into account too. Other elements must be counted by introducing suitable alloying factor K: mAl ⳱ (XAl Ⳮ 0.5 • XCr) • q • h • K 2

in g/m . The coefficient 0.5 for chromium mass fraction XCr in this case was taken from the regression equation, interlinking the effect of these elements (Al and Cr) on the oxidation kinetics in Ni-Co-Cr-Al system alloys. The tests were conducted at 1200 C. When aluminum consumption for oxide formation on turbine blades is known, then the coating life expectancy will be directly proportional to the aluminum reserve. The coating life parameter at temperatures above 1000 C will depend on the effects of interaction between the coating and superalloy. In both laboratory and engine tests, a part of the aluminum reserve is lost in diffusion of aluminum and chromium from the coating into the superalloy. In laboratory tests the losses maybe considerable compared with aluminum consumption for oxide formation, but in testing on turbine blades the oxide consumption will always exceed interaction losses. Since interaction zone dimensions (between diffusion/overlay coatings on the one hand and nickel-base superalloys on the other) do not exceed 30 to 40 lm when tested/run for up to 1000 h at temperatures below 1000 C, these may be disregarded. Thermal Fatigue of Turbine Blades. As a result of higher gas temperatures in aircraft engines and the development of turbine blades with highly efficient cooling, the thermal stresses in blades have risen. Thermal fatigue cracks in blades have become the commonest defect. It has been suggested that all thermal fatigue cracks occurring in turbine blades of different aircraft engines can be conventionally classified into two types, proceeding from either crack origins within the substrate or from thermal stress cracking in coatings (Ref 24): ● Thermal fatigue cracks of the first type ● Thermal fatigue cracks of the second type (fragmentation)

Differences in the appearance and microstructure between the two types are as follows: ●

The first type of thermal fatigue cracks are generally few. In the process of operation, they penetrate deeply into the blade wall, reducing the thermal stresses arising between certain zones of the surface and eventually leading to blade failure. ● The second type of thermal fatigue cracks (fragmentation) can appear on blade surfaces during the first hours of blade service life. Usually these are numerous fine cracks, occasionally covering the full blade surface. The depth of their penetration is usually limited to within the coating thickness; thereafter, the cracks either discontinue their growth or continue at a very slow rate. The unique feature of such cracks is that their distribution density is very high; distances between adjacent cracks are comparable with the coating thickness. The research into the fragmentation phenomenon shows that its occurrence is due to the following: ●

● ●

● ●

High gradients of temperatures across the blade wall section, the blades being provided with highly efficient cooling systems Increased transient running on the engine Sharp changes of heating on transient running (thermal shocks), during which the rate of temperature rise and drop is measured in hundreds of degrees per second Difference between the thermal expansion coefficients (TECs) of coating and superalloy Insufficient ductility of coatings in the area where maximum strains occur

Fragmentation emerges as a result of the coating-superalloy system exposure to high thermal stresses and strains. Illustrated in Fig. 2.12 is a turbine blade made from the superalloy JS6F, with a diffusion coating and after testing for 100 h. The entire pressure surface of the blade is covered with a net of the second type of thermal fatigue cracks, penetrating not beyond the coating thickness. The optimum way of removing fragmentation is to reduce thermal stresses in the blade coating to as low a level as possible. Bringing closer the coefficients of thermal expansion of the coating and superalloy can do this. While calculating the cyclic service life of a coated blade, the designer can predict the emergence of fragmentation


cracks and thereby choose coatings that offer the best life expectancy. The principles of a cyclic service life calculation for a coating deposited on a blade surface are dealt with in Chapter 8, “Strength Designing of Turbine Blade Protective Coatings.” The calculation is based on stresses and strains arising in the coating due to centrifugal forces, bending moments, nonuniformity of blade thermal conditions, and thermal stresses caused by different TECs of the coating and the alloy protected. In performing calculations, it must be borne in mind that the composition of the coating and the whole complex of its physical and mechanical characteristics are subject to continuous changes with time. The rate of such changes is higher when the temperature of the blade surface rises. The coating cyclic life expectancy must be calculated for the original and final coating compositions. Should the coating properties undergo essential changes with time, interim calculations may be necessary. While protecting the blade from oxidation, the coating simultaneously produces extra stresses and strains on the blade surface. These stresses and strains damage the coating and reduce its cyclic service life. This reasoning holds only with respect to an ideal blade, because in actual conditions, when uncoated blades are exposed

Fig. 2.12

to vigorous oxidation, the resultant loss of blade service life far exceeds the losses caused by the coatings. With high strains in blade coating (Deic 0.5%), the rate at which thermal fatigue cracks emerge in the coating (fragmentation) may exceed the rate of blade surface damage from oxidation. In this case, the deposition of coatings would bring about a shortened cyclic life of the blade. Figure 2.13(a) and (b) illustrate the results of cyclic tests conducted on engines whose turbine blades are made from superalloys JS6F and JS6U with diffusion and overlay coatings Ni20Cr12AlY, Ni20Co20Cr12AlY, and Co25Ni20Cr12AlY. The number of blades with fragmentation grows with test time. After 10 to 50 h, all blades are covered with a network of cracks. No cracks have been observed on the blades with Ni20Cr12AlY coating. Spalling of Ceramics on Turbine Blades with Thermal Barrier Coatings. The development of thermal barrier coatings (TBC) offers new horizons in further lengthening of turbine blade life expectancy. Not only does the TBC protect the blade surface from hot corrosion and oxidation, it also offers thermal protection and thus shields the superalloy from softening during high-temperature exposure. The TBC is basically a structure consisting of a bond coat and

(a) External appearance of a blade with fragmentation exposed by capillary flaw detection after 500 h of testing. (b) Microstructure of a coating in a fragmentation zone. 500⳯


an outer ceramic layer. The latter is based on zirconium oxide and serves as the TBC main element. The properties of the TBC ceramic layer are entirely dependent on the mode of its formation. If the electron beam (EB) technique is applied and if all process problems have been satisfactorily settled, the ceramic layer will have a specific columnar structure. The typical features of such a structure are low tensile strength and low elastic modulus in the direction normal to the crystallite-growth axis. As thermal stresses and strains arise, the ceramic layer freely breaks into fragments. The individual fragments are sized by such factors as stress level and the ceramic layer thickness. Any strains in the blade cause fragmentation. This property protects the ceramics from high stresses and spalling. Life expectancy of a TBC working at temperatures above 1000 C is totally dependent on the adhesion of the ceramic layer. The EB technique enables ceramic layers with original adhesion strength higher than 70 MPa to be obtained.

Number of (a) JS6F and (b) JS6U blades with fragmentation vs. time of testing. 1, Ni20Cr12AlY coating; 2, aluminized coating; 3, Ni20Co20Cr12AlY coating; 4, Co23Ni20Cr12AlY coating

Fig. 2.13

Depositing a TBC on the blade alters its characteristics. This must be noted when designing blades with TBC. The thermal barrier effect DT is controlled by the ceramic layer thickness, d; its thermal conductivity, k; and the thermal flux through the blade wall, Q such that D • T ⳱ Q • d/ k. There are certain restrictions on the extent to which ceramic thickness can be increased. In view of its low tensile strength, the ceramic layer is unable to carry mechanical loads produced by centrifugal forces. Hence, extra loading is applied to the turbine blades (and disk). This dictates that it is better to make the ceramic layer of variable thickness throughout blade surface. Maximum thickness is needed in the areas where most intensive thermal fluxes will be experienced or where the thermal field is nonuniform. In other zones, the ceramic layer may have minimum thickness or be absent. The optimum thickness of a ceramic layer on the blade varies within 120 to 250 lm. At lower thickness, there may be rapid oxidation of the ceramic-bond coat boundary and poor adhesion strength. At greater thickness, considerable stresses resulting in spallation are likely to occur. The composition of a bond coat and its thickness are chosen as when using metal heat resistant coatings. Since TBCs are employed on blades with high thermal stresses, preference should be given to bond coats offering higher resistance to thermal fatigue cracking. Once the ceramic layer has spalled, the bond coat has to protect the blade from oxidation. The main cause of TBC failure is spallation of the ceramic layer from the blade while being exposed to higher temperatures for a long time (Fig. 2.14). The spallation is assisted by tear-off stresses occurring on the convex blade surfaces when thermal stresses occur inside the ceramic layer. These internal compression stresses reach their maximum as the blade is cooled down to room temperature. The onset of TBC failure is betrayed by a network of microcracking (fragmentation) appearing on the surface of the blade. Fragmentation of a ceramic layer does not indicate its failure. The blades in such a state may be exploited for quite a long time. Fragmentation appears on some surface areas where tensile stresses can be generated in the ceramic layer under the certain engine operation conditions. TBC life expectancy can be estimated by adhesion strength that has to be maintained throughout the service life of the coating. According to this criterion, the adhesion strength


must exceed the shearing stress needed to separate the ceramic coating from the surface. Failure of the ceramic layer in the TBC is evidence of a reduced adhesive strength on the ceramic-

Fig. 2.14

Blades with a spalled ceramic layer after 800 cycles of testing at 200 ↔ 1100 C

bond coat boundary. Spallation of ceramics after high-temperature tests takes place even in the absence of stresses in the sample. The effect is associated with the formation of various Al2O3type oxides on the ceramics-metal interface. Such oxides reduce the adhesive strength and alter the stress condition on the boundary because their physical and mechanical properties differ from those of ceramic and metal layers in TBCs. All the above processes, leading to ceramic layer spallation, can be allowed for by the integral characteristic of adhesion strength, ra, found experimentally in relation to the temperature and duration of tests. The criteria applicable to the determination of adhesive strength for a ceramic layer on blades with TBC are discussed in Chapter 8, “Strength Designing of Turbine Blade Protective Coatings.” To retain adhesive strength with time is the most formidable challenge to further lengthening of TBC life expectancy. Two prerequisites, however, must be fulfilled for this: to reduce oxygen diffusion mobility in the lattice of zirconium oxide and to prevent fragmentation of ceramics in high-temperature applications. Benefits that TBCs provide include the opportunity to elevate the gas temperature at the turbine inlet without changing the temperature of the turbine blades. Most often, however, TBCs are used to extend blade service life. Estimation of the TBC benefit is a difficult task. The issue is that while comparing the blades with and without TBC, some side effects occur that affect the blade temperature (e.g., reduction of perforated hole diameter). Analysis of plastic strains in identical blades with and without TBCs, tested in the engine under similar conditions revealed lesser elongation in blades with TBC in comparison to their counterparts without TBC (Fig. 2.15); that is, they demonstrate higher high-temperature strength. Use of TBCs allows higher temperatures in engine turbines to be employed, though ZrO2 • Y2O3 oxide imposes significant restrictions on its use at the temperatures above 1150 C. These restrictions are due to high oxygendiffusion mobility in zirconium oxide. The development of new ceramics for high-temperature application is the most promising route to further TBC improvement. REFERENCES

Fig. 2.15

Elongation of blades with (1) Ni20Cr12AlY coating and with (2) TBC vs. service time

1. N.D. Kuznetsov, Problems of Thermocyclic Strength of Gas Turbine Engines Elements, Probl. Prochn., Vol 6, 1978, p 3–8


2. Y.B. Chertkov, K.F. Rybakov, and V.H. Zrelov, Impurities and Purification Methods of Oil-based Fuels, Moscow, Khimiya, 1970 3. Ya.B. Chertkov and V.G. Spirkin, Use of Jet Fuels in Aviation, Moscow, Transport, 1974 4. I.H. Shishkov and V.B. Belov, Aviation Lubricants, Fuels, and Special Fluids, Moscow, Transport, 1979 5. I.A. Birger, B.F. Schorr, and I.V. Demianushenko, Machine Elements, Moscow, Mashinostroyeniye, 1975, 455 p 6. N.D. Kuznetsov, Designing the Strength of a Long Life Gas Turbine Engine, Probl. Prochn., 5, 1976, p 3–9 7. N.D. Kuznetsov, Strength of Gas Turbine Engine Turbine Elements under Complex Loads and Related Problems, Probl. Prochn., 3, 1982, p 10–14 8. M.Ia. Ivanov and V.P. Pochuev, Problems of Designing High-Temperature Turbines of Modern Aircraft Engines, Conversion in Machine Building of Russia, 5, 2000, p 34– 46 9. V.H. Abiants, Jet Engines, Moscow, Mashinostroyeniye, 1985 10. Ye.N. Bogomolov, Service Processes in Cooled Turbines of Gas Turbine Engines with Perforated Blades, Moscow, Mashinostroyeniye, 1987 11. S.Z. Kopelev, M.N. Galkin, A.A. Harin, and I.V. Chevtchenko, Thermal and Hydraulic Characteristics of Cooled Gas Turbine Blades, Moscow, Mashinostroyeniye, 1993, 176 p 12. G.P. Dolgopolenko, M.D. Romanov, and V.V. Gatin, Plane and Helicopter Gas Turbine Engines, Moscow, Mashinostroyeniye, 1983 13. D.H. Boon, Overlay Coatings for Improved Oxidation/Corrosion Protection and Ductility for High-Temperature Applications, Airco Temescal, May, 1977, 12 p

14. B.A. Movhan and I.S. Malachenko, Heat Resistant Coatings Deposited in Vacuum, Kiev, Naukova Dumka, 1983, 232 p 15. A.G. Andreeva, V.V. Terekova, and G.D. Fomenko, Heat-Resistance Coatings of Nickel-Base Alloys, High-Temperature Coatings, Leningrad, Nauka, 1967, p 96– 110 16. Y.A. Tamarin, Heat Resistant Diffusion Coatings for Turbine Blades, Moscow, Machinostroenie, 1978, 133 p 17. P. Galmiche, Applications en Construction Airospatiale et Retomlees de Techniques Thermochimiques ONERA (The Use of Chemical Heat Treatment of ONERA in Aircraft and Ground Constructions), Aircraft and Space, 1973, Vol 41, p 33–42 18. P.T. Kolomytzev, Heat Resistant Diffusion Coatings, Moscow, Metallurgiya, 1979, 272 p 19. T.V. Levchenko, V.I. Moroz, and L.P. Buiyanova, Protective Coatings on Metals, Kiev, Naukova Dumka, 4, 1971, p 158–164 20. S.A. Muboyadjyan, I.A. Pomelov, Y.A. Tamarin, N.V. Zabrodina, and R.I. Beliakova, “Aluminum-Base Alloy for Coatings,” Author’s Certificate USSR N1067847, 1983 21. E.B. Kathanov and Y.A. Tamarin, Protective Coatings: Effective Path of Increase of Reliability of the Blades, Aviation Materials on the Eve of 21st Century, Moscow,VIAM, 1994, p 296–304 22. R.E. Halin, E.B. Kachanov, I.L. Svetlov, and V.N. Toloriya, Single-Crystals of Nickel Superalloys, Moscow, Mashinostroenie, 1997, 333 p 23. L.B. Getzov, Gas Turbine Components: Materials and Strength, Moscow, Nedra, 1996, 590 p 24. Y.A. Tamarin, V.G. Sundyrin, and E.B. Kachanov, Gas Corrosion and Thermal Fatigue of Protective Coatings for Turbine Blades, High Temperature Corrosion and Protection, China, Liaoning Science and Technology Publishing House, 1991, p 161–166



CHAPTER 3

Technological Processes for Deposition of Protective Coatings to Turbine Blades A GREAT VARIETY of techniques for deposition of protective coatings to aircraft turbine blades have been designed. The use of each technological process is warranted by a number of reasons including requirements to the properties of protective coatings, production equipment available, personnel experience, and common practice. The preference of new techniques to conventional ones is justified only in cases where fundamental changes in the properties of protective coatings are required. For example, new techniques are necessary in the case of designing a new turbine. As a rule, superseding in the field of technology involves considerable financial expenditure connected with the purchase of equipment and materials and training of personnel.

given in percents by weight). Coatings alloyed with chromium, silicon, and other elements can be formed. The method of aluminizing in the powder mixtures with activators offers such advantages as simple production process and equipment, a wide temperature range for aluminizing (from 500 to 1200 C), highly even thickness and homogeneity of the produced coatings, and low production cost. Powder mixtures for aluminide coating formation are composed of aluminum or its alloys with iron, chromium, and other elements (as an active component), aluminum oxide (as an inert filler), and a halide activator (usually NH4Cl). When the container with the saturating mixture and aluminized parts is heated, the following reactions take place: NH4Cl ↔ NH3 Ⳮ HCl

Diffusion Coatings Application of Diffusion Coatings in Powder Mixtures with Halide Activators The diffusion saturation of superalloys with aluminum and with aluminum combined with other elements (Al-Cr, Al-Si) from the powder mixtures with halide activators is a widely used method of aluminide coating formation on turbine blades of aircraft gas turbine engines. The wide application of this method is justified by the fact that different mixtures used for aluminizing ensure a wide range of aluminide coatings with aluminum contents varying from 18 to 69% Al (here and subsequently all compositions are

2NH3 ↔ 2N Ⳮ 3H2

The activator decomposition products displace the air from the container and reduce partly the oxides on the metal powder surface and on the aluminized alloy surface. Hydrogen chloride interacting with aluminum forms aluminum chlorides: 6HCl Ⳮ 2Al ↔ 2AlCl3 Ⳮ 3H2 2AlCl3 Ⳮ Al ↔ 3AlCl2 AlCl3 Ⳮ 2Al ↔ 3AlCl

These chlorides carry aluminum to the saturated surface on which the disproportionation reac-


tions occur, resulting in the formation of the corresponding aluminides and the highest valence chlorides: 3AlCl2 Ⳮ n Ni ↔ 2AlCl3 Ⳮ NinAl 3AlCl Ⳮ 2n Ni ↔AlCl3 Ⳮ 2NinAl

AlCl3 resulting from the reaction goes to the powder mixture where it interacts with aluminum again. This interaction feature is that aluminizing does not involve the consumption of chlorides. The equilibrium mixture of aluminum chlorides at 1200 to 1700 K has the prevailing AlCl2 content, the maximal amount of which is formed at approximately 1400 K. The aluminizing process occurs because of the thermodynamic equilibrium shift in the gaseous mixture resulting from absorption of aluminum and its diffusion into the saturated metal. The outcomes of aluminizing, however, such as dimensions and compositions of the coatings formed, depend not only on the amount of aluminum dichloride in the mixture but also on the rate of its supply to the saturated surface, the rate of the surface reaction, the rate of the reactionproduct discharge to the mixture, and the rate of the diffusion processes in the coating being formed. The quantitative analysis of all these processes is extremely complicated; hence, for practical purposes, the saturating mixtures are usually characterized by the experimental relationships between the mixture composition and the properties of the coating, such as its composition, thickness, and hardness.

Influence of NH4Cl activator content on thickness of coatings forming on nickel-base superalloys. Aluminizing in mixture: Al-Fe alloy (50% Al)-NH4Cl at 950 C for 4 h. Source: Ref 1

Fig. 3.1

The amount of a halide activator added has a strong effect on the outcome of aluminizing, such as coating thickness and surface quality characteristics (Fig. 3.1). When the amount of NH4Cl increases from 0.1 to 2%, a gradual thickening of the coatings formed on nickel-base superalloys occurs at their aluminizing from the powder mixture that contains an Al-Fe alloy with 50% of Al as an active component. The further increase in NH4Cl content up to 10% results in thinner coating due to a greater amount of chlorides displaced from the container and, respectively, to lower aluminum content of the Al-Fe alloy. This effect is particularly well seen at a repeated use of the powder mixtures with more than 4% NH4Cl activator. When a mixture with 10% of NH4Cl is used ten times (the activator has been introduced before every next process), the thickness of the formed coatings is four times lesser, while for 2% NH4Cl mixture, the coating thickness is reduced by a factor of 1.8 only. Hence, the optimal activator content of the powder mixture ensuring the thickest coatings at an insignificant depletion of the mixture is in the range of 1 to 3%. Other halide-containing compounds, such as NH4F, NaF, KF, (NH4)HF2, NH4I, NH4Br, N2H2-2HCl can also be used as activators at aluminizing of superalloys. The mechanisms of their activating effects are similar. The main parameter of the powder mixtures used for aluminide coating manufacturing is their activity determined by the amount of aluminum that the mixture feeds to the saturated surface. This parameter determines the concentration of aluminum and other elements in the coating, its thickness, and, hence, its protective and other properties. The saturating mixtures may be conventionally classified into high, average, and low activity ones on the basis of the aluminide structure and composition in the coating outer layer. A high-activity mixture is usually composed of aluminum powder and inert filler. As the content of free aluminum increases, the activity of such a mixture also increases, resulting in thickening the coatings formed on the surface of the aluminized alloy (Fig. 3.2). The aluminum content of the coating outer layer exceeds 40%. The mixtures with aluminum contents of 10 to 15% have found use in the application of aluminide coatings to turbine blades and other elements of aircraft gas turbine engines. The high activity of the powder mixtures containing free aluminum allows their use for alu-

Technological Processes for Deposition of Protective Coatings to Turbine Blades / 27

minizing at the temperatures below 900 C. If the aluminum content of the mixture increases up to 98%, the aluminizing temperature may be lowered down to 500 to 600 C. The mixture containing 98% aluminum powder and 2% NH4Cl forms thick enough coatings with aluminum content of 66 to 69% on nickel-base superalloy surfaces within the temperature range of 500 to 600 C (Fig. 3.3). The powder mixtures are widely used to obtain protective aluminide coatings on superalloy surfaces. They contain bound aluminum as a component of aluminum-iron or aluminumchromium alloys. These alloys may be obtained by presmelting followed by reducing to powder with the required particle size or by sintering the powders of their raw elements. In some cases

Influence of aluminum content of mixture of AlAl2O3-5% NH4Cl on the thickness of coatings forming on nickel-base superalloys at 900 oC

Fig. 3.2

Fig. 3.3

Kinetics of coating growth on nickel-base superalloys. Aluminizing in mixture: 98% Al and 2% NH4Cl

the saturating mixtures are composed of powders of individual elements and sintered at aluminizing. The halide formation and interaction with the saturated surface obey the previously mentioned laws for powder mixtures containing bound aluminum. In the gaseous phase forming at aluminizing, however, the corresponding halides of the elements added together with aluminum can be found. The amount of halides at aluminizing depends on the thermodynamic parameters of their formation and on the amount of the element fed. It is noteworthy that the thermodynamic potential varies significantly less when iron, chromium, and silicon halides form in the mixture than when aluminum halides form; hence, their amount in the gaseous phase of the mixtures with more than 30% Al is not large. The interaction between the saturated surface and halides of iron, chromium, silicon, and other elements added to the mixture results in a formation of the aluminide coatings alloyed with these elements. If iron, chromium, nickel, cobalt, and some other elements are introduced into the saturating mixtures together with aluminum, the mixture activity, the aluminum content of the resulting coating, and the coating thickness decrease. For example, an increase in the iron content of the saturating Al-Fe alloy from 30 to 50% reduces the thickness of the coatings formed on the JS6K alloy at 950 C for 4 h from approximately 1000 to approximately 60 lm. In the coating, the reduction of the aluminum content is usually accompanied by the increase in the second-component content of the saturating alloy. As the mixtures with iron, chromium, or nickel contents of more than 50% feature low activity, aluminizing is usually carried out at temperatures above 1000 C to obtain the required thickness of the coating. For aluminizing at temperatures above 850 C, Fe-Al alloy-based mixtures with aluminum content of 50% are convenient (Ref 1). The outer layer of the coating formed when these mixtures are used to saturate nickel-base alloys has aluminum content of 32 to 36%. The low rate of the coating thickening in these mixtures as compared with that observed in the high-activity ones allows accurate enough control of the thickness of the coatings formed within the temperature range of 850 to 1100 C currently used for superalloy aluminizing. The kinetics of diffusion coating formation on the nickel-base alloys is shown in Fig. 3.4. The mixtures com-


posed of Al-Fe alloy with aluminum content of 50% feature high stability that is particularly important for their commercial application. The Al-Fe alloy-based mixtures with initial aluminum content of ⱕ40% feature low activity. The outer layers of the coatings obtained on nickel-base superalloys at aluminizing from such mixtures at 950 to 1200 C have an aluminum content of 30% and less. If the aluminum content of Al-Fe alloy is more than 70%, such mixtures feature high activity. They may be used

to saturate nickel-base superalloys at 600 C. In this case, a 30 to 50 lm thick aluminide coating with aluminum content of up to 60% is formed within the period of 4 to 5 h. The mixtures composed of aluminum and silicon powders analyzed elsewhere (Ref 2) have somehow different properties. The interrelation of the aluminum and silicon effects on the properties of the coatings has been studied using the mixtures containing 50% (Al Ⳮ Si), 45% Al2O3, and 5% NH4Cl. The addition of silicon to the saturating mixture reduces the activity of the mixture and the thickness of the coating formed (Fig. 3.5), but this drop of activity is significantly less than that for the previously mentioned Al-Fe alloys. This is due to the absence of chemical interaction between aluminum and silicon and of intermetallide compound formation. At 1000 C, the increase in the silicon content above 50% results in coating thickening due to the intensification of the siliconizing process. The aluminide coatings alloyed with silicon may be obtained at the aluminum/silicon (Al/Si) ratio of the saturating mixture within the range of 1 to 9 to 9 to 1. The kinetics of coating formation on the JS6K alloy at its aluminizing from the mixture of powders containing 10% Al Ⳮ 50% Si, 38% Al2O3, and NH4Cl is shown in Fig. 3.6. The method of aluminizing in powder mixtures that contain halide activators allows the application of protective aluminide coatings of various compositions to turbine blades and other

Kinetics of coating growth on nickel-base superalloys. Aluminizing in mixture: 98% Al-Fe (50% Al) alloy and 2% NH4Cl. Source: Ref 1

Fig. 3.4

Influence of aluminum and silicon contents of mixture on thickness of the coatings forming on nickelbase superalloys. Aluminizing in mixture: 50% (Al-Si), 45% Al2O3, and 5% NH4Cl. Source: Ref 2

Fig. 3.5

Kinetics of coating growth on nickel-base superalloys. Aluminizing in mixture: (10% Al – 50% Si), 38% Al2O3 and 2% NH4Cl

Fig. 3.6


parts within a wide temperature range of 500 to 1200 C. The initial considerations, which allow the selection of the saturating mixture and conditions of aluminizing and diffusion annealing, are the requirements to the protective coating composition and thickness. The key problem is the selection of the proper thermal conditions for the coating formation. If the heating of the parts in the container with the powder mixture and their cooling after aluminizing are slow, the mechanical properties of the superalloy may worsen significantly. Hence, if it is possible, aluminizing should be combined with heat treatment operations that do not require accelerated cooling, mainly, aging. When the high-activity mixtures, which allow aluminizing in the temperature range of 500 to 800 C are used, diffusion annealing after aluminizing may be combined with any operation of the alloy heat treatment. Sometimes at aluminizing in low activity mixtures, when the process temperature is above 1000 C, special heat treatment is required to compensate for the reduction of alloy heat resistance after aluminizing. Many different mixtures are known to be used to obtain heat resistant coatings on superalloy elements. The mixtures saturating such critical elements as aircraft engine turbine blades should meet some special requirements; their properties should be considered from the viewpoint of commercial application where aluminizing is carried out in large containers (having room for 20 to 100 blades). The mixtures for turbineblade aluminizing should guarantee reproducibility of the coatings with the required composition and thickness, allow for the possibility of their repeated use (provided that the thickness and composition of the coatings in question do not change), and guarantee the same initial class of surface roughness of the finished blades. The first requirement is usually met by selecting a suitable mixture for the aluminizing temperature range in use and by using a highly homogeneous mixture. Those are the features of the mixtures composed of cast Al-Fe or Al-Cr alloys. The ingots are crushed into 2 to 5 mm granules. When the mixtures, which contain aluminum powder and oxides of aluminum, chromium, and other elements, are used, the initial powders should be fine (less than 60 lm) and thoroughly mixed. To increase stability of the powders, they are annealed under H2 atmosphere. The proper use of the activator is also very important in obtaining coatings of the preset

composition, thickness, and high surface quality. The ammonium chloride (NH4Cl) should be predried at 130 to 150 C, and then the crushed activator should be distributed uniformly in the mixture. To produce high quality coatings, thoroughly preparing the blade surfaces for aluminizing is of particular importance. The surface of the elements should be cleaned from oxides and impurities. To improve the quality of coatings, the blade surfaces may be treated with glass balls. Before the blades are arranged in the containers, they should be degreased. The homogeneity of the coating composition and the evenness of its thickness both depend on the proper and regular arrangement of the blades in the container for aluminizing. They should be positioned in such a manner that the thickness of the surrounding layer of the saturating mixture is not less than 25 to 30 mm. Reference specimens are placed in the middle and near the walls of the container. The excess of gases resulting from the activator decomposition should be pumped out of the furnace space. As a consequence of large-sized industrial containers and low heat conduction of the saturating mixtures, the greatest share of the process time (70 to 90%) is spent on heating the container to the required temperature and on cooling. The coatings form during both heating and cooling; this should be accounted for when computing the aluminizing time. Hence, if large containers are used, it is more convenient to use the low and average activity mixtures in order to get the coatings of required parameters on the blades situated in the center and near the walls of the container with the least possible tolerances. The amount of aluminum consumed when forming the surface coatings is small enough compared to its content of a mixture, but, when a powder mixture is used more than once, the reduction of its activity, the aluminum content of the coatings, and the coating thickness is observed. There are two reasons for this. The first is that the mixture is depleted of aluminum. It is spent on the formation of gaseous aluminum chlorides partly replaced from the container and on the formation of the coating. The second reason is that mixture oxidation with the air coming into the container while cooling oxidizes the mixture. The oxidation is extremely intensive in the mixtures composed of fine powders of aluminum, chromium, and other elements. In this case, a protective atmosphere (Ar, H2) is created in the container at its heating and cooling to make possible a repeated use of the mixture.


The multiple use of the mixtures consisting of 98% Al-Fe alloy (50% Al) and 2% NH4Cl results in the aluminum content of the mixture of up to 39% after the twentieth process. In this case the aluminum content of the coating formed drops from 36 to 24%, and the coatings themselves are about six times thinner than those obtained in the first process (Fig. 3.7). To use a saturating mixture repeatedly and to obtain the constant thickness and composition of the coatings, a certain mixture beneficiation process should be commercially used; that is, a certain amount of the mixture of its initial composition should be added to the reused one. When Al-Fe alloy-based mixtures are used, 5 to 25% of the initial alloy is added after every process. The particles equal to 1 mm or smaller are screened. The control of the added alloy parameters, that is, the mixture activity, allows the control of the aluminum content of the coating within the range of 24 to 36%. The powder method of aluminide coating application to turbine blades has some serious drawbacks. Preparation of powder mixtures, arrangement of the components in the container, and their removal from the container after aluminizing are labor consuming. It is difficult to protect some surfaces (e.g., blade roots) from aluminizing. There is some risk that the blade cooling channels will be fouled with the powder mixture. Blade cooling after aluminizing is slow enough. Considerable emission of halides to the atmosphere and a large volume of powder wastes should be subjected to special utilization. Updating the powder method may eliminate some of the aforementioned drawbacks. The powder mixture may be placed into special cavities in the container. The parts are arranged in

Influence of repeated use of saturating mixture on aluminum concentration in (1) Al-Fe alloy, (2) on aluminum content of the outer layer of coatings, and (3) on coating thickness on nickel-base superalloys. Aluminizing in mixture: 98% Al-Fe alloy and 2% NH4Cl at 950 C for 4 h. Source: Ref 1

Fig. 3.7

the container so that a free circulation of the halides released from the mixture at heating is ensured. The airtight container is blown by argonhydrogen or evacuated to the vacuum of 0.65 Pa before aluminizing to prevent oxidation of the parts and the mixture at heating. The transfer of aluminum halides becomes more efficient due to their forced circulation in the container. Aluminizing powder mixtures can be used for forming diffusion coatings on the inner cavity surfaces of the cooled blades. To form the protective coating on the intricate inner surface, it is necessary to use the process that guarantees the transfer of aluminum to this surface. It occurs when the inner cavity of the blade is filled with the powder mixture. The powder mixtures used for this purpose should be free-flowing bulk materials both initially and after their use for aluminizing of the inner cavity surfaces. Besides that, they should not cake and stick fast to the blade surface. The technological process for forming a diffusion coating on the inner cavity surface combines the elements of aluminizing under vacuum and aluminizing with the use of halides. It includes the following main stages: preparing a powder mixture; filling the blade inner cavity with the mixture; aluminizing; cleaning the blade inner cavity surface from the powder mixture; and inspecting the coating obtained. The chromium-aluminizing mixture is prepared of aluminum, chromium, and aluminum oxide powders (Ref 3, 4). Powders of individual components are mixed in the following proportions: aluminum, 7 to 8%, chromium, 35 to 40%; and aluminum oxide, balance. The mixture prepared is thoroughly intermixed and annealed in vacuum at 1000 C for 3 to 4 h. During that period of time, chromium and aluminum interact and form a solid solution of aluminum in chromium and lower chromium aluminides. If sintered, the mixture is crushed and sifted through a sieve. To activate the mixture, it should be annealed again after adding 0.3 to 0.4% aluminum chloride (AlCl3). After being well intermixed with aluminum chloride, the mixture is placed into a leakproof container from which air has been expelled by argon blowing-through. Mixture annealing is conducted for 2 to 3 h. During annealing, the formation of aluminum chlorides occurs in the mixture. At cooling, they condense on powder particle surfaces. The chloride compounds activate aluminum and chromium transfer from powder particles to the surface saturated.


Adequate preparation of the inner cavity of the blade for coating is a stage of great importance. It should guarantee required evenness and thickness of the coating on the inner blade surface. The matter is that these surfaces have imbedded oxides, ceramic inclusions (from cores), various impurities, casting porosity, and so forth. During preparation of the inner cavity for aluminizing, all impurities should be removed by abrasive blasting. Blasting in two directions (from the blade root and from its trailing edge) is considered to be good practice, thereby eliminating “shaded” zones where processing otherwise would have been of poor quality. The operation of filling saturation mixture should assure complete filling up of blade inner spaces: no coating is formed in the spaces unfilled with mixture. A relatively high fineness of the powder-mixture components (100 lm) and its good flowability enable most intricately shaped interiors of the blade to be fully filled. Since the mixture retains its free-flowing properties unaffected after the saturation procedure is finished, it is easy to remove the mixture from the blade inner spaces completely. To prevent mixture outpour from the cooling channels while filling, the channel outlets, where they open onto the surface, are sealed with a thick organic-based suspension. The process of aluminizing is carried out under (1 to 3) ⳯10ⳮ1 Pa vacuum at 1000 to 1030 C for 5 to 6 h. “Aluminum-to-surface” transfer occurs first through a gas phase and then through a vapor phase. The thickness of the coating formed will depend on mixture activity, process temperature, and time, as well as on the ratio of a certain area of the blade inner cavity surface being aluminized and the amount of mixture on this very area. The process of diffusion saturation involves the mixture depletion of aluminum, the reduction of its activity, and the drop of coating formation rate (Ref 3). In 8 h, when held at 1050 C, the aluminum content of the mixture near a saturated surface will drop from 7 to 2%, and the zone of changed composition will penetrate into the mixture depth by 2 to 3 mm. As a result, in some positions of the inner cavity where blade spacing is 1 to 2 mm, the saturating mixture rapidly becomes depleted of aluminum, and its activity lowers. The process of saturation ceases. This is why the blade inner surfaces that have been aluminized will inevitably show great variation in coating thickness. Saturating for 5 to 6 h at 1000 to 1030 C will form a coating 10 to

30 lm thick, containing 18 to 22% Al. The fact that coating thickness depends not only on time and temperature, but also on ratio between “the surface area saturated and the volume of the mixture on the surface” (this ratio being characteristic of the blade design type) dictates individual selection of saturation conditions for a certain blade design. To improve the process of aluminizing in powders containing halides, new techniques for diffusion coating formation have been developed.

Application of Diffusion Coatings in Circulating Gaseous Phase The circulation technique uses chemical processes similar to those that occur during aluminizing from powder mixtures containing halides. The difference is in eliminating the preparation and use of powder mixture. The technique is based on forced circulation of the gaseous medium that contains aluminum halides (AlCl3, AlCl2) in the space between the source of aluminum (i.e., liquid aluminum) and the blades being aluminized. The required gaseous medium is created in a special reaction zone by passing gaseous halide AlCl3 over melted aluminum surface having a temperature approximately 900 to 1000 C (Ref 5, 6). A fan directs the gaseous compound (AlCl2) formed in the reaction zone toward the aluminizing zone where the blades heated up to 900 to 1000 C are stationed. On their surfaces, reaction of disproportioning takes place. As aluminum reacts with the superalloy, a coating is formed whose aluminum content and structure are close to those obtained by aluminizing from powder mixtures with halide activators. The AlCl3 formed in the aluminizing zone is headed for the reaction zone again. Thus, the aluminizing process proceeds with halides continuously circulating over melted aluminum and the blades being aluminized. Temperatures in the zones of the unit are chosen such that optimum conditions are produced for the reactions in the course of which AlCl2 is first created and then decomposed on the blade surfaces. Besides temperature, the velocity and pattern of motion of gaseous halides markedly affect the kinetics of reactions. In order to obtain a uniform flow of saturation components and to ensure steady reaction, the velocity of halide flow must correspond to a minimum enough to impart turbulence to the gaseous mixture.


The design of the unit for the circulation technique of aluminizing is shown in Fig. 3.8 (Ref 7). The blades are placed on the shelves in the gas-tight muffle. A system ensures halide circulation in the muffle. It consists of a fan and a few special baffles. Aluminum used in the circulation technique of aluminizing is in a special bath in the lower part of the muffle. The halides that are necessary for the process form in the reactor and then head for the muffle. The halides that leave the muffle condense in the special system for exhaust gas purification. Diffusion coating growth kinetics in the circulation technique of aluminizing depends on the process temperature and time. The relationship between the coating thickness and the alloy composition is observed. Besides these wellknown factors, the coating thickness depends on gas flow velocity. The thickest coating forms when gas flow changes from laminar to turbulent. In the outer zone, aluminum content is 22 to 26%. The main advantage of the circulation technique is the possibility of deposition of diffusion coatings on inner space surfaces of cooled blades. In comparison with the blade outer surfaces, the inner surfaces work at lower (ⱕ900 o C) temperatures. Besides, severe erosion and

corrosion do not affect them. Instead, however, they require coatings with rather high ductility because the intricate profile of their inner surfaces features numerous stress concentrators. The main difficulty here is to ensure adequate thickness and composition of coating throughout full inner space. This task can be accomplished by blowing through the inner space a saturating gaseous medium, with the gas well agitated during the procedure. In this way a uniform mass transfer of aluminum from the carrier gas to the surface can be ensured. Uniformity of mass transfer in the blade inner space has been examined, revealing the dependence of this characteristic on gas flow velocity. The most uniform mass transfer is observed when blowing through the inner space from both ends, that is, from its root and trailing edge sides. The use of the gaseous medium containing aluminum iodides (AlI, AlI3) helps provide optimum conditions for aluminum mass transfer (Ref 6). Experimental studies of the processes involved in saturating the insides of blades with aluminum by employing the circulation technique have shown the following tendencies. The process temperature and fan rotation speeds (i.e., gas-carrier flow velocities) appear to have affected most essentially the coating thickness. Coating evenness was satisfactory in all the experiments as measured at four opposite points of blade interior. When applying the circulation technique for aluminizing, the aluminum content of coating varies within 16 to 25%, which usually meets the requirements imposed on the protective properties of the blade inside. The circulation technique can be used for depositing chromium-base diffusion coatings as well as diffusion coatings alloyed with chromium and silicon. The advantages of the circulation technique are ecological friendliness (as no powder mixture is required) and even coating thickness on both inner and outer blade surfaces. The disadvantage of this technique is corrosion damage of equipment parts contacting with halides.

Application of Diffusion Coatings Using Slurry The design of the unit for applying the diffusion coatings by the circulating technology of aluminizing. 1, furnace; 2, lower cover; 3, upper cover; 4, gas-tight muffle; 5, director baffle; 6, fan; 7, saturating metal (liquid aluminum in the circulating method of aluminizing); 8, reactor for halide formation; 9, turbine blades; 10, halide condensation system; 11, vacuum pump; 12, loading mechanism. Source: Ref 7

Fig. 3.8

The application of aluminizing slurries by painting followed by diffusion annealing is widely used for obtaining aluminide protective coatings on turbine blades and vanes (Ref 8, 9). The use of slurries has a series of advantages over aluminizing from powder mixtures. They


are as follows: a shorter thermal cycle of coating application due to quick heating and cooling of the treated part; a possibility of local aluminizing, for example, saturation of a blade airfoil portion when its root is finished; a possibility of applying coatings to large-size parts, for example, combustion chambers. The use of slurries increases significantly the output of aluminizing and allows process automation. The slurry is made of the saturating element powders, such as Al, Al-Si, Al-Si-(Y, Ce, B), and so forth, with the particle size of less than 40 lm (as its active component) and of an organic binder. The binder solutions are selected so they can ensure a good suspension of the active component, a quick drying of the slurry on the saturated surface, and an even layer with the required strength to withstand next operations. A readyto-use slurry should be good for a long-term storage without deterioration of its properties and have a viscosity facilitating its application by various methods, such as dipping, brushing, and spraying. Diffusion annealing should ensure a complete enough burnout or volatilization of the organic component without deteriorating the alloy and the resulting diffusion coatings. Slurries manufactured on the base of aluminum powder (as an active component) and of the solution of colloxiline in the mixture of amylacetate and dimetyloxolate (as an organic binder) are widely used. Colloxiline with its low flashpoint of 180 to 200 C has a high decomposition rate at diffusion annealing of the treated parts. This permits spending less of the slurry and reducing the nitrogen and carbon contents of it. The process features an intensive release of mainly carbon-containing compounds at the flashpoint. As the treated specimens are further heated, all the remaining organic fractions are removed when the temperature of 650 C is reached. At the same time a part of the powder is oxidized by the active component, resulting in a respective gradual increase in the slurry weight. As colloxiline burns out, the active component is concentrated. The surface-tension forces make the powder particles converge and partly take out the volume of the burned organic components. It allows metallic contacts between the particles of the saturating elements and the saturated surface. This condition is necessary for a diffusion coating formation. The presence of oxides on both aluminum particles and the saturated surface is an obstacle for the diffusion processes. Yet, as the elements are heated to temperatures close to the aluminum

melting point, the diffusion mobility of the atoms increases and surpasses these barriers. Vacuum is the most favorable condition to guarantee the contact formation. It facilitates the evaporation of the impurities and the removal of the adsorbed gases. The diffusion layers obtained under vacuum are of even depth and high surface-finish quality. At heating, the saturating elements enter into reactions with oxide films on the nickel-base alloy surface. These reactions favor the formation of even, high-quality coatings; that is why the aluminizing process under an atmosphere reducing the oxides is undesirable. The liquid phase is most favorable for reactions. It guarantees high rates of diffusion processes. That is why the best quality layers are obtained at saturating with several elements (Al-Si, Al-Si-Cr) when the slurry has a significant aluminum content. At saturation with aluminum, the liquid phase consists of Al-NiAl3 eutectics that spread over the surface, favoring the dissolution of the nearest solid particles and the absorption of the saturating elements on the whole surface of the alloy. The dissolution of the solid phase in the liquid one goes on to the point when the element contents in the melt reach the level when solid refractory compounds may form. Their formation begins at the metal-melt interface, resulting in a gradual solidification of the whole volume of the liquid phase. After solidification, a specific dendrite pattern appears on the surface. At further holding, the diffusion processes continue in the solid phase, but their rate is much lower. The diffusion coating forms actually during 10 to 15 min of the diffusion annealing; further holding thickens the coating. The interaction between the active component of the slurry and the saturated surface improves with the increase in the temperature of diffusion annealing; that is why higher temperatures result in the formation of the coatings with more even thickness and high quality finish. Thus the process of saturation with aluminum, or aluminum together with silicon, chromium, and other elements, using slurries is a variant of the liquid-phase aluminizing method, where the applied layer of the slurry limits the liquid-phase volume. The results of saturation, that is, the aluminum content of the coating formed, its alloying element contents, and, hence, hardness, ductility, and other parameters, depend on the thickness of the applied slurry layer, the ratio of aluminum and other metal-constituent contents in its active component, diffusion annealing


temperature, and time. The liquid phase on the saturated surface ensures favorable conditions for adsorption and diffusion processes, but it is not obligatory for diffusion coating formation. A diffusion layer may form at saturation from the solid phase as well. In this case the process is more favorable for the formation of intermetallide phases with high heat of formation. The slurries used for turbine blade aluminizing are prepared of a solution of colloxiline in the mixture of amylacetate and dimetyloxolate (as an organic binder) and of aluminum powder (as an active component). The aluminum powder and colloxiline solution are mixed in the ratio of 600 to 700 g of the powder per 1 L of the organic binder. Then it is filtered through a sieve with a mesh size of up to 40 lm. The slurry is thoroughly mixed before application. Its viscosity should ensure the possibility of its spraying. At the same time, its running off the saturated surface should be prevented. At aluminizing from slurry, particular attention should be paid to the preparation of the part surface. It should be cleared from dirt, grease, and oxide particles. If the surface is of a low class finish, if it has pits, freckles, and other flaws where the slurry cannot penetrate, or, on the contrary, where it will be thicker than required, all this can result in either uncovered areas or areas with too thick a coating after annealing. Hence, the saturated surface preparation determines to a great extent aluminizing parameters and final coating quality.

The evenness of the coating thickness after annealing depends completely on the evenness of the applied slurry layer. Taking into account the strict requirements to the evenness of the blade coating thickness, a special attention should be paid to the technique of slurry application; it should be automated if possible. After diffusion annealing, a thick layer of the slurry (80 lm) and its sags may cause a thick brittle coating on these areas, prone to cracking at cyclic loads. When the slurry with a complex active component is used, special attention should be paid to uniform distribution of the active component elements. When the saturating elements have noticeably different specific weight values (Al-Ta, Al-Nb), special mixers should be used. They should prevent heavy fractions from precipitation while the slurry is being applied to the alloy. The slurry is applied to the surfaces of the blades arranged in special facilities preventing its application to the blade roots or other elements not to be saturated. The slurry layer on the blade surface is then air dried and visually inspected. Then the parts are placed in special fixtures in the container; slurry damage should be avoided. The parts are annealed either in vacuum or under argon or air atmosphere (if mechanical finishing follows). The annealing conditions are selected in accordance with the heat treatment conditions of the saturated alloy. Annealing at 950 C for 4 to 6 h ensures good results. The thickness of the resulting aluminide coating, h, is a linear function of the applied slurry thickness, H; it may be approximately calculated from the following relationships: h ⳱ 0.7 • H for annealing at 950 C for 4 to 6 h h ⳱ H for annealing at 1200 C for 1 to 2 h

Influence of diffusion annealing temperature on hardness (lines 1 and 2), and aluminum content of the coating outer layer (line 3) on nickel-base superalloys. Aluminizing from slurry: 100% Al as an active component. Annealing time: (1) 1 h; (2) 6 h; (3) 4–6 h. Source: Ref 9

Fig. 3.9

A thicker layer of slurry (100 lm) may spall at diffusion annealing. The geometric dimensions of the parts change while saturated. The increase factor equal to the quotient of the specimen linear dimensions by the coating thickness varies in the range of 0.9 (at annealing at 650 C) to 0.3 (at annealing at 1000 C). The annealing temperature effects the surface finish. The best results have been obtained within the range of 850 to 1100 C. However, at these temperatures the surface roughness increases somehow in comparison with the initial one (Ra ⳱ 0.32 to 0.25 lm). If the annealing temperature is increased from 700 to 1000 C, the aluminum content of the coating outer layer


drops from 43 to 18%. The coating hardness drops respectively (Fig. 3.9) (Ref 9). An important advantage of the method of aluminizing from slurry is the ease of saturation with several elements such as aluminum and chromium, aluminum and silicon, aluminum and boron, and so forth. At multicomponent saturation, an active component of the slurry consists of a mixture of powders of the elements alloying the aluminide coating. The powders with the particle size of 40 lm are thoroughly mixed and added to the amylacetate solution. Further stages of the aluminizing process are similar to those analyzed in previous sections. Chemical composition, thickness, and properties of the coatings formed after annealing depend on the selected ratios of aluminum and the second element of the active component. In this case, the best coatings are obtained if the active component contains a large amount of aluminum powder (50%). In the case of simultaneous saturation with aluminum and silicon (Al-Si) followed by diffusion annealing at 950 C, the thickness of the coatings on nickel-base superalloys reduces with the increase in the amount of silicon in the active component of the slurry. At annealing temperatures above 1050 C, the coating thickness increases (Fig. 3.10) (Ref 2). It is due to intensification of the process of saturation with silicon at high temperatures. Diffusion annealing at temperatures of 950 to 1050 C with the silicon

content of the active component of the slurry of 5 to 50% produces the aluminide coatings alloyed with silicon. If the silicon content of the slurry active component exceeds 50%, the resulting coatings are brittle. The main disadvantage of this method is a need for application of the slurry by hand. It causes uneven coatings on a blade surface. The permissible variation in thickness between the blades of one batch is usually 20 to 50 lm, yet the real variation may be much greater. The variations in the coating thickness may be caused by the presence of oxides and large carbides on a blade surface. They hinder the diffusion processes and result in uncoated areas. The slurries used contain organic components and metal powders of high explosion hazard. That is why the use of well-known electrostatic units for applying the slurry to blade surfaces is a problem. These shortcomings can be overcome by using new physical techniques and special units for application of diffusion coatings.

Application of Diffusion Coatings by Electric Arc Technique under Vacuum The starting considerations at development of new aluminizing techniques are as follows: selecting a proper active medium containing the saturating element; providing a reliable means for transfer of the saturating element to the surface to be saturated; and activating the transfer of the saturating element from its medium to the surface to be saturated. The technological process should ensure controllable conditions under which the transfer of aluminum and the formation of the required diffusion layer occur at the highest possible rate. All these requirements are met by ionized plasma flows, which are produced by the electric arc (EA) technique (Ref 10–13). The unit using EA for material evaporation can: ●

Influence of silicon content of the active component of (Al-Si) of slurry on thickness of resulting coatings on nickel-base superalloys during diffusion annealing at 850 to 1200C for 2 h. Source: Ref 2

Fig. 3.10

Conduct ion cleaning up and activating the surfaces of saturated parts by treating their surfaces with plasma flows with the ion energy that exceeds 200 eV ● Produce active plasma phase from the ions of saturating elements and bring them to the surface under saturation ● Transfer ions from plasma flow to the surface to be saturated and maintain required boundary conditions throughout the entire process of treatment The new technology for forming the diffusion aluminide-type coatings on turbine blades (Ref


14, 15) is based on use of the EA technique. The plasma accelerator cathode is made of an aluminum-base alloy. In the electric arc burning between the cathode and the anode, aluminum and other elements of the cathode are evaporated, ionized, and transferred to the surfaces of the parts to be coated in the form of an ion flow. The nature of surface reactions depends on the energy of the ions in plasma, their number, the physical and chemical characteristics of plasma and the materials that the plasma interacts with, and on the interaction temperature. Depending on the negative potential applied to the part being coated, the energy of the ions deposited may vary within a wide range from tens to hundreds of electron volts. The features of the processes occurring on the surface under saturation depend on the energy of ions, their density, and the temperature of the surface. If the energy of aluminum ions exceeds 200 eV, the saturated surface will be cleaned up from impurities and thin oxide layers. At the ion energy in the range of 10 to 50 eV, aluminum ions will be transferred to the surface to be saturated. The mechanism and kinetics of subsequent processes depend on the speed of aluminum ions transferred and on the rate of their “absorption” by the surface, depending on the surface temperature. The rate of ion transfer to the saturated surface, m, can be calculated from the equation involving arc current in vacuum, Iea, and transfer coefficient, l, depending on the type of material evaporated (Ref 12, 16). For aluminum alloys, the transfer coefficient, l, is equal to (4.5 to 5.0) 10ⳮ5 gram/Coulomb. Under steady working conditions of the unit, when arc current, Iea, does not vary (Iea ⳱ constant) the mass of aluminum alloy, m, transferred to the blade surface within a period of time, t, can be expressed in the following form: M ⳱ K • l • Iea • t in g/m2 where Kⳮratio (1/m2), depending on the type of the unit and on the surface area saturated. The average rate of aluminum transfer from the cathode to the part surface is equal to 1 to 2 g/m2 per minute. Aluminum alloy deposition time required for application of a diffusion coating 25 to 50 lm thick (its thickness is measured after annealing at 1000 C for 4 h) is 20 to 40 min. The mass of aluminum transferred, m, is directly proportional to the process parameters. It allows full automation of the aluminizing process and formation of the coating with required thickness and chemical composition.

Three alternative processes are feasible on the surface being saturated, proceeding from the amount of aluminum, m, transferred to the surface per time unit, and amount of aluminum, “mAl,” that can be absorbed by the surface in diffusion processes. ●

The amount of aluminum transferred to the surface to be saturated considerably exceeds aluminum mass that can form a diffusion coating, that is, m mAl. It is possible at high rates of condensation and low temperatures, which are not high enough to assure proper kinetics of diffusion processes (below 640 C). In this case a coating formed on the surface features the composition of that of the feedstock evaporated. At the surface temperature above 640 C a layer of liquid aluminum alloy will form. The kinetics of its growth is dependent on the rate of aluminum ion condensation. ● The amount of aluminum transferred to the surface to be saturated is equal to the amount of aluminum that can be absorbed through diffusion at a given temperature: m ⳱ mAl. In this case, an aluminum-base diffusion coating forms on the surface. Its aluminum content reaches its possible maximum. The coating composition and kinetics of its thickening depend on the temperature and kinetics of diffusion processes. ● The amount of aluminum transferred to the surface to be saturated is less than the amount that can be absorbed through diffusion processes: m mAl. In this case, the diffusion coating forms on the surface. The kinetics of its growth depends on the rates of aluminum ion transfer to the surface to be coated. Aluminide-type coatings can be alloyed with different elements by making the cathode of aluminum alloys containing various elements. This type of alloying is inapplicable in other currently used aluminizing techniques. The Al-Si system was selected as a base system for development of coating compositions. Its choice was dictated by the research into alloying the diffusion aluminide coatings with silicon that was conducted earlier. Silicon improves casting characteristics of aluminum alloys. It enables cathode to be cast at conventional aluminum foundries. To improve cyclic heat resistance of diffusion coatings and, hence, to prolong their service lives, the aluminum alloys for coatings and the coatings themselves


were alloyed with yttrium, cerium, and some other elements. The technological cycle of forming diffusion coatings consists of the following sequential phases. As-finished blades are subjected to clearing up and degreasing all over their surfaces. Next, the blade airfoil is blasted first with synthetic corundum with the grain size of 30 to 60 lm and then with compressed air. Thus, prepared blades are assembled in special fixtures that are then placed inside the unit chamber. The unit is pumped to vacuum of ⱕ1 ⳯ 10ⳮ2 Pa. The aluminum alloy is evaporated in electric arc discharge struck in the unit between the aluminum cathode and the anode and transferred to the blade airfoil surface. As a rule, the units do not provide for the blade preheating systems. Thus, interaction between plasma and the surface being coated occurs according to the first of the alternative processes. In this case the mass of the material transferred to the surface, m, is much larger than the mass of aluminum, that has entered into forming the diffusion coating, mAl. As a result, a dense layer of the alloy with the same composition as that of the feedstock evaporated is formed on the blade airfoil surface. Usually it takes 20 to 40 min to treat the blade in the plasma flow. After the blades are coated with the aluminum alloy to a required thickness, they are taken out of the unit and annealed under vacuum. The annealing temperature and time depend on kinetics of diffusion and phase transformations. They are chosen to meet the requirements to the coating composition and thickness. After diffusion annealing, the blades will have a dark-gray color. No additional treatment of the surface is required. The diffusion coating thickness and the aluminum content of its outer layer depend on the thickness of the aluminum layer deposited and on the annealing time and temperature. The thickness of the diffusion coating in use is usually 25 to 50 lm. Its aluminum content is 20 to 26%. The main advantages of the technique in question are as follows: automation of the process of aluminum alloy deposition on the blade surfaces; precise reproducibility of the coatings with specified compositions and thickness; the possibility of coating deposition on the certain blade surfaces and areas and, therefore, no problems of preventing coating deposition on the other areas; and ecological friendliness of the technological processes as no powders or halides are used.

Inspection of Diffusion Coatings Diffusion aluminide coatings protect the blade surface against high temperature gas corrosion and erosion and determine to a great extent the blade life and reliability. That is why the quality control of the coatings deposited is a very important stage of the technological process of diffusion saturation. The main parameters determining the coating protective properties and serviceability are its composition and thickness. The chemical composition of the aluminide coatings is characterized mainly by its aluminum content. In the case of alloyed aluminide coatings, aluminum and alloying element contents characterize it. The thickness inspection and the analysis of the aluminum content of the coating (as well as that of the alloying element for multicomponent saturation) are of primary importance for testing coating quality and protective properties. It is noteworthy that as a rule, no composition analysis of the coatings deposited is performed in commercial practice. This is due to the complexity of such an analysis. Also, the methods of the coating formation from powder mixtures with halide activators, which are widely used in industry guarantee stable aluminide coatings with a constant aluminum content. The situation changes when the aluminide coatings alloyed with chromium, silicon, and other elements are formed. To guarantee the presence of the alloying elements, which impart the required properties to aluminum coatings, especially in the cases where the element ratio in the saturating composition may change at application, methods and equipment for direct analysis of the composition of the coating formed should be developed. The coating thickness is determined by a metallographic method or by measuring the increase in the dimensions of the reference specimens saturated together with a batch of blades. For example, at saturating nickel alloys from a mixture containing 98% Al-Fe (50% Al) alloy and 2% NH4CI the side increment B of the specimen and the coating thickness, h, are connected by the following relationship: B 0.6 h. When the slurries are used for aluminizing, a direct connection between the specific mass (g/ m2) of the applied layer and the thickness of the coating formed is observed. In this method, the thickness of the slurry layer is inspected. As the physical properties of the applied slurry layer differ from those of the base alloy, nondestruc-


tive testing may be used. The applied layer thickness inspection is a critical and necessary operation, because if the layers are too thick, it will result in thick and brittle aluminide coatings prone to spalling at operation. In cases where aluminizing is well perfected under the industrial conditions and the technique in use guarantees the coatings with the thickness and composition varying in a narrow known range, indirect methods of quality and thickness inspection may be used. These are testing hardness, measuring specific weight gain (g/m2) at aluminizing, and metallographic structure analysis. For aluminide coatings formed on nickel superalloys at the temperature of 900 to 1100 C the connection between the aluminum content of the outer layer of the diffusion coating and its hardness is observed. The outer-layer hardness is tested on metallographic specimens used for coating thickness measurements. The outerlayer hardness may vary for alloys with different chemical composition. This also depends on the medium used for aluminizing, which contains alloying elements in addition to aluminum. In these cases the calibration curves of “hardness/ aluminum concentration” are plotted for each alloy, saturating medium, and the conditions of aluminizing. For indirect analysis of the aluminum content of the aluminide coatings on nickel superalloys, the metallographic method (comparison of the outer and inner layer thickness) may also be used. At the aluminum content of 33 to 36% in the outer layer, the ratio of the outer and inner layer thickness is about 4. At 24 to 26%, the thickness ratio drops to 2 to 2.5. The value of the specific mass gain (Dm, g/ m2) of the specimen after aluminizing characterizes to a great extent composition, thickness, and protective properties of a diffusion coating. In fact it is a measure of an “aluminum reserve” in the coating; that is, mass of aluminum to be consumed on the formation of protective oxides during the blade service life. The media of certain compositions show a direct connection between the specimen weight gain and the thickness of the diffusion coatings formed. The inadmissible flaws of the diffusion coatings are as follows: uncoated areas of the blade surface to be protected and cracks in a coating. Coating discontinuity can be detected by visual inspection of the blades annealed under oxidizing atmosphere. The uncoated area is covered with an oxide layer of a specific color. The cracks are detected by routine nondestructive testing.

Overlay Coatings The electron-beam (EB) and electric-arc (EA) coating techniques are based on the fundamental physical processes of electron-substance interaction, evaporation, formation of a vapor cloud or plasma flow, and condensation of substance. Since these physical processes have been broadly discussed in the literature (Ref 17–19), to treat them here in detail is not intended.

Some Physical Principles of Material Evaporation by Electron Beam Electron beam evaporation is based on thermal effects that occur as the flow of electrons interacts with the substance. The EB evaporator consists basically of an EB gun generating a directed flow of electrons possessing necessary energy and an evaporation unit containing the substance to be evaporated. The power of the electron beam, Pe, will depend on the gun accelerating voltage, Ue, and on beam current, Je: Pe ⳱ Ue • Je

in kWt. The EB gun accelerating voltage is chosen to meet power and economy requirements. Increasing the voltage will complicate the highvoltage system and make the unit more expensive. Usually the EB coaters used in industry are rated for 18 to 20 kV. As an EB strikes the surface, it penetrates into the substance being evaporated, imparting to it its power, Pe. The depth of electron penetration depends on their energy and the density of the substance evaporated. Normally the depth of penetration varies within 1 to 5 lm. As the electrons interact with the solid substance, their kinetic energy is translated into thermal energy and Roentgen (x-ray) radiation. The thermal energy is spent to heat, melt, and evaporate the substance, but some part of it will be lost through heat radiation of the substance heated and also due to a heat flow entering the evaporator wall. The evaporator thermal efficiency is determined by the relationship between that part of the power consumed for evaporation and the total power of the beam. The efficiency depends on the evaporator design. When evaporated by the EB technique, 99% of the substance evaporated consists of neutral atoms and their complexes. The rate of evaporation from surface area unit is dependent on the physical properties of the


substance evaporated and on the temperature of its heating: V ⳱ ␣ ⳯ 4.4 ⳯ 10ⳮ4 ⳯ po ⳯ 冪M/Tv

in g/s where V is the rate of evaporation from surface area unit; ␣ is the coefficient of evaporation; po is the equilibrium pressure of vapor of substance evaporated at a given temperature; M is the molecular mass of substance evaporated; and Tv is the temperature of evaporation. When an alloy of two and more components is heated and evaporated by the EB technique, the alloy components are evaporated at different rates. The ratio between evaporation rates depends on the activity coefficients of each component, ca and cb, taking into account the interrelation between melted components; on the molar concentration of melted components, Ma and Mb, and on the vapor partial pressures of each component, pa and pb: ␣ab ⳱

cb • pb Ma ca • pa Mb

1/2

冢冣

With ␣ab ⳱ 1, both components evaporate simultaneously, and condensate composition is identical to that of the substance. With ␣ab 1, the component b will have higher concentration in condensate than in original substance. And conversely, when ␣ab 1, component b will have lower concentration in condensate than in original substance. In the process of evaporation, the pool content is continuously changing and, hence, the value of ␣ab also varies. The time of evaporation, taken by condensate to become elementally similar to the original substance, is known as the stabilization time. This time will vary with substance composition, electron gun power, and evaporator design. Difference between the rates of evaporation of various components in complex alloys imposes certain restrictions on the employment of EB technique. Such elements as Hf, W, Ta, Re, and Si do not evaporate simultaneously with Ni, Cr, and Al from the Ni (Co)-Cr-Al-base system. The same can be said about the ceramic materials. As the substance is evaporated on the evaporator surface, the partial pressure of its vapor near the surface reaches 100 to 300 Pa, and a directed flux of vapor is formed whose density can be expressed through spatial distribution U(␣): U(␣) ⳱ U0 • cosn(␣)

Accordingly, condensed substance thickness at various points of the surface running parallel to the evaporator surface and located at a certain distance away from it will correlate as: h(␣) ⳱ h0 • cosnⳭ3(␣)

where h0 is the thickness of condensed substance layer (overlay) on the surface running parallel to the evaporator surface at the 0 point opposite to evaporator; h(␣) is the thickness of a condensed substance layer on the surface running parallel to the evaporator surface, at a point which is r away from the 0 point; and n is assumed to take values in excess of 1 and be dependent on the evaporator characteristics and rate of evaporation. A multi-crucible system of evaporators and facilities presetting the required movement of the components past the evaporator are used to make coating more uniform in thickness.

Some Physical Principles of Material Evaporation by Electric Arc The EA deposition coater operates by creating an ion flux from the material evaporated by an EA burning in vacuum (Ref 19–21). The EA strikes between the two electrodes, which are the cathode and the anode. Depending on arcing conditions, one of the electrodes (the cathode) suffers erosion, so arcing occurs in the plasma consisting of the ions of this electrode. The most commonly used vacuum-arc discharge is that initiated between the integrated cathode and anode. In this case arcing occurs between the cathode spots formed on the cathode surface and the anode, being evenly spread over the surface of the latter. The number of cathode spots is proportional to the arc current, the spots chaotically moving about the cathode surface. High density of the current (106 to 107 A/cm2) and energy flux (107 to 108 W/cm2) in a cathode spot result in intensive evaporation of the material and in the formation of plasma flux with high flux velocity (104 m/s). The average energy of the ions in the plasma flux is high enough for them to break through the difference in potentials between the cathode and the anode and to move in any direction. The consensus is that cathode spots are the sources of plasma with the degree of ionization close to 100%. The high ionic current, consisting of the cathode material ions, can be derived from the plasma of the vacuum arc burning on the cathode surface without affecting the conditions of


arcing. Total ionic current generated by plasma of the arc under vacuum comprises up to 7 to 10% of the discharge current. The amount of material evaporated per second, m, is proportional to electric arc current, Iea, and the cathode material transfer coefficient, l (Ref 10, 20): m ⳱ l • Iea in g/s. The coefficient of transfer in a flux will depend on physical properties of the material evaporated (Table 3.1) (Ref 21, 22). The important parameters of the evaporation process are the temperature of the cathode spot and the pressure of evaporated material vapor in this spot. High pressure in the cathode spot is a must for evaporating the required amount of substance (Table 3.2) (Ref 23). In the cathode spot, the material is evaporated, spattered, and ionized. The amount of vapor, ions, and microdroplets varies with the properties of material evaporated, the design of evaporators, and the electric arc current. The amount Table 3.1 Coefficients of transfer in the flux for various materials Cathode material

Tboiling

Zn Ag Al Cu Cr Ni W

1180 2483 2740 2868 2956 3003 5973

l ⴒ 105, g/s

K K K K K K K

2.15 1.50 4.50 1.15 0.40 1.00 0.62

Source: Ref 22, 23

Table 3.2 materials Cathode material

Cathode spot parameters for various Temperature in cathode spot, K

Tboiling/Tevaporation

2770 5210 9550 2340

1.15 0.90 1.73 0.81

Cu W Ta Sn

Vapor pressure in spot, Pa 4.05 2.02 96 6.07

⳯ ⳯ ⳯ ⳯

104 104 105 104

Source: Ref 23

Table 3.3 Mass-spectrometric analysis data for material evaporation Cathode material Al Ni

Ion content

Degree of ionization 50–60% 60–70%

Source: Ref 24, 25

Ions Ⳮ1

Ⳮ2

Ⳮ3

Al , Al , Al NiⳭ1, NiⳭ2, NiⳭ3

Mⴐ1

Mⴐ2

Mⴐ3

60 65

38 33

2 2

of microdroplet phase can comprise 1 to 30% (by mass); the droplet size can vary within the range of a fraction of a micrometer to tens of micrometers. The degree of ionization of evaporated products also depends on the material evaporated and the design of the cathode. The degree of ionization rises as materials of higher melting points are evaporated (Table 3.3) (Ref 24, 25). The degree of ionization will also be higher when a Hall end-face plasma accelerator is used as an evaporator. Underlying the operating principle of such an accelerator is the presence of a spatial nonuniform external magnetic field that is generated by a solenoid in the simplest case (Ref 26, 27). Under currents of 100 to 500 A, the concentration of charged particles will be 1012 /cm3, so magnetic fields of 10 to 50 G intensity will be necessary for the Hall acceleration to manifest effectively. The Hall plasma accelerator made in the form of a magnetic coil wound around the cathode enhances ionization in the flux, accelerates ions, and collimates ion flow along the accelerator axis. The microdroplet phase is charged with drifting electrons and localized on flux periphery. Ionic energy can be controlled from tens to hundreds of electron volts by using a plasma accelerator and by altering the induction of magnetic current. The degree of plasma flux ionization in the accelerator and its homogeneity will be much dependent on magnetic field magnitude. The greater the magnitude is, the higher is the degree of plasma ionization and the lesser the homogeneity of plasma flux due to a sharper focusing of plasma. Differences in the valence of ions, their atomic masses, and the effects of interaction between the magnetic field of accelerating coil and the electric field of plasma flux bring about separation of ions and change condensate (overlay) composition in comparison with the composition of the evaporated cathode. The evaporator design has an effect on composition similarity between the material evaporated and the coating deposited. For instance, the evaporation of Ni20Cr12AlY alloy will result in aluminum content of the condensate dropping by 10 to 30%. Aluminum ions separate across the plasma flux area (i.e., across the height of the unit working zone). Separation of heavier ions (for example Hf, W) is much less pronounced (5%). The comparison of the chemical compositions of the cathode and the coating deposited in the EA coater are given in Table 3.4 (Ref 28).


The accelerator working conditions must be chosen to meet maximum ionization of the plasma outgoing flux and fall within admissible inhomogeneity of ion current density (in the last analysis of the coating thickness) across the flux area. In choosing the accelerator working conditions, one should consider, first of all, the relationship between the rate of coating deposition and the distance from the workpiece to the cathode surface. These measurements are taken at different radii with respect to the plasma flux axis. After the accelerator working conditions have been chosen, the working conditions of depositing also must be chosen, proceeding from the negative potential applied to the workpiece. Increasing negative potential and, hence, intensifying the energy of ions on the workpiece surface, will bring about re-evaporation and slowdown of deposition buildup. At the voltage above 300 V, workpiece surfaces undergo the processes of cleaning. As the negative potential applied to a workpiece rises, the coating composition diverges from that of cathode. The processes of arc burning, evaporation of cathode material, and condensation of plasma flux must occur in a vacuum not worse than 1 ⳯ 10ⳮ2 Pa; otherwise, high plasma activity will Table 3.4 Chemical composition similarity of Co28Cr5AlY cathode and coating Chemical composition, % Object

Cr

Al

Y

Cathode Coating

27–29 29–32

4.0–6.0 2.0–3.5

0.4–0.5 0.1–0.2

Source: Ref 28

Table 3.5

Specification of the UE-type coater

Parameter Total power of coater Number of guns for evaporating substance Power of one gun for evaporating material Number of guns for heating blades Power of one gun for blade heating Type of guns for heating and evaporation Gun operating voltage Number of cylindrical evaporators, 70 mm in diameter each Distance from evaporator surface to the rotary center of a holder with blades Vacuum required for coating deposition Number of chambers for loading of blades Maximum blade length that can be coated Rate of metal coating deposition Rate of applying ceramic coatings

Value 250 kW 4 60 kW 2 40 kW Flat-beamed 18–20 kV 4 pieces 330 mm 4 ⳯ 10ⳮ2 Pa 2 400 mm 10–12 lm/min 4–5 lm/min

entail uncontrollable pollution of workpiece surfaces with oxides, carbides, and nitrides.

Units for Coating Deposition EB Coater. The design concepts are similar for all EB units. Basic differences are in the design and power of the EB guns in use (flat-beam or axial guns), the number and the shape of the evaporators (the most commonly used ones are cylindrical evaporators with the diameters of 70 mm), and blade position relative to evaporators. Other differences can include technological process monitoring, control systems, and some special facilities available. The units for EB coating (UE-175, UE-187, and UE206) designed at Paton Electric Welding Institute (Table 3.5) are used for depositing overlay coatings on turbine blades by the EB technique (Ref 29, 30). The UE coater consists of three interconnected vacuum chambers (Fig. 3.11). These are the main processing chamber, where deposition takes place, and two other chambers to load the blades. The main processing chamber is divided into two sections, one of which accommodates EB guns and the other contains evaporators. A multi-crucible evaporator of the coater with linearly arranged evaporators enables depositing different alloys and depositing strengthening oxide phases that are uniformly introduced into the matrix, multilayer coatings, thermal barrier ceramic coatings, and so forth. All the crucibles are made of copper and cooled with running water. The evaporators are provided with special mechanisms that continuously feed the materials to be evaporated. Two movable, mounted copper shields in the main processing chamber make it possible to separate blades from evaporators. Connected to the upper wall of the main processing chamber is a chamber with two EB guns to heat the blades in the process of deposition. The main processing chamber communicates with the two chambers for blade loading through side walls. Each loading chamber incorporates a mechanism to secure a holder with blades and a rod to advance the holder with blades into the main processing chamber and make the blades rotate and reciprocate over the evaporators. Blades are heated, and flat-beam EB guns evaporate materials with a linear hot cathode (Ref 29). In operation, the EB is controlled by deviating and scanning along two coordinate axes. Each electron gun has individual packs of power feed, filament current control, and an


electromagnetic system of deviation and scanning. The coater has one source of accelerating voltage with the power levels of 250 kW. Feedstock for evaporation is loaded into the evapo-

rators via a hole in the processing chamber. The evaporators are loaded with ingots of evaporated alloys 70 mm in diameter and up to 400 mm long or with ceramic pellets 70 mm in diameter. EA Coater. The MAP, Duga, and other units

Diagram of central chamber of UE coater for electron-beam deposition. 1, coating deposition chamber; 2, electron-beam guns for material evaporation; 3, electron-beam guns for blade heating; 4, crucibles with material evaporated; 5, chamber for blade loading; 6, facilities for fixing and rotating blades. Source: Ref 30

Fig. 3.11

Diagram of MAP unit for electric-arc deposition. 1, evaporable cathode; 2, anode; 3, vacuum chamber; 4, electromagnetic catch of cathode spots; 5, 6, cathode vertical movement mechanism; 7, turbine blades; 8, cathode shield; 9, cachion electrode; 10, power supply of negative voltage to the blade; 11, magnetic coil; 12, electric arc striking mechanism; 13, unit for protection from microarc striking; 14, source voltage. (B¯-magnetic field, Iea-arc current) Source: Ref 13

Fig. 3.12


(coaters) are designed for depositing metal overlays on turbine blades and other engine components by EA technique (Ref 12, 13, 31). The MAP Unit. (Fig. 3.12 and Table 3.6) designed at All Russian Institute of Aviation Materials (Ref 13) consists of a vacuum chamber (3) accommodating a device for generating plasma flux from evaporated substance. The device is made up of a cylindrical cathode (1), an anode (2), and electric arc striking mechanism (12). By means of a special mechanism (5, 6), the cathode is moved vertically with respect to the cooled interior of a magnetic coil, the latter fixing the burning arc ring zone on the cathode

Table 3.6

Specification of the MAP Unit

Parameter

Value

Diameter of coating deposition chamber Chamber volume Vacuum in coating deposition chamber Total power Cylindrical cathode dimensions Arc voltage Arc current Voltage on blades Number of stations for holders with blades Maximum height of components coated Maximum diameter of components coated Number of blades coated at a time Rate of MCrAlY alloy deposition onto revolving blades Unit overall dimensions

1000 mm 0.6 m3 5 ⳯ 10ⳮ3 Pa 75 kW Height, 340 mm; OD, 180 mm; ID, 140 mm 35 V From 400 to 1200 A From 10 to 300 V 24

surface. Components (7) to be coated are arranged around the cathode on special holders, the latter revolving about their own axes. The unit is handled from above through the vacuum chamber cover. Duga-90 Unit. Another type of the unit for coating deposition by EA technique is the Duga90 (Ref 31–33) (Table 3.7) designed at Machine Research Problems Institute, Russian Academy of Science. Eight end-plasma evaporators are used in the unit (Fig. 3.13). They are arranged in pairs on the opposite walls of the chamber for coating deposition. The dimensions of the chamber for coating deposition are 1150 mm in height and 1000 mm in diameter. The rate of coating deposition is 0.1 to 0.2 lm/min when one pair of evaporators is used, arc discharge current is 100 A, and planetary workpiece rotation takes place. The flat cathodes used in the evaporators have diameters of 100 mm. The main feature of the Duga-90 unit is a high level of ion flow control. It allows even coating deposition along the height of the working zone of 800 mm. The difference in the thickness of

200 mm 130 mm From 12 to 96 0.3–0.6 lm/min 3000 ⳯ 4000 ⳯ 2800 mm

OD, outside diameter; ID, inside diameter. Source: Ref 12, 13

Table 3.7

Specification of the Duga-90 Unit

Parameter Diameter of coating deposition chamber Vacuum in chamber Total power Flat cathode diameter Number of cathodes Arc current Voltage on blades Number of stations for holders with blades Maximum height of blades coated Maximum diameter of blades coated Number of blades coated at a time Rate of MCrAlY-type alloy deposition on to revolving blades Unit overall dimensions Source: Ref 32, 33

Value 1000 mm 5 ⳯ 10ⳮ4 Pa 55 kW 100 mm 8 From 100 to 200 A From 10 to 1500 V 8 800 mm 340 mm From 1 to 8 0.1–0.2 lm/min 2500 ⳯ 2000 ⳯ 2100 mm

Chamber for electric-arc coating deposition of Duga-90 unit from inside. In the center there are blades 500 mm high; electric-arc evaporators are in the side walls. Source: Ref 33

Fig. 3.13


this zone is not more than 20%. This unit may be used for coating deposition on the blades with the length of 800 mm. The blades are arranged in a planetary mechanism that ensures their rotation in the plasma flow. The number of the blades in the operating zone depends on the blade dimensions. To control the process parameters, the unit is equipped with facilities for measuring rates of coating deposition using quartz resonators and for process-temperature monitoring using infrared pyrometers. The obvious advantage of the unit is the use of ion gas sources for blade surface clearing before coating deposition. The unit features the possibility of applying high voltage (up to 1500 V) to the processed blades. At this voltage, effective clearing of

Table 3.8 Variation of contents of the alloying elements in the process of Ni20Co20Cr12AlY alloy melting Chemical composition, wt.% Melting technique Charge VI VIⳭEBR

Ni

Co

Cr

Al

Y

Base Base Base

20.0 19.7 19.8

22.5 22.7 22.5

11.0 11.1 10.2

0.50 0.25 0.18

VI, vacuum induction; EBR, electron beam remelting

Table 3.9 Gas contents in various zones of ingot Ni20Co20Cr12AlY produced without extra deoxidation Gas contents, wt.%

Melting technique

O2

N

H2

VI VI VI Ⳮ EBR VI Ⳮ EBR

0.0036 0.0020 0.0024 0.0026

0.0024 0.0027 0.0020 0.0020

0.00030 0.00022 0.00018 0.00038

Sample cutout location Ingot top Ingot middle Ingot top Ingot middle

VI, vacuum induction; EBR, electron-beam remelting

Table 3.10 Gas contents in ingots produced by VI and subjected to extra deoxidation with carbon and cerium Gas content, wt.% Melting stage Melt Introducing 0.02% C Introducing 0.02% Ce Introducing 11% Al Finished ingot VI, vacuum induction

O2

N

0.0025 0.0014 0.0011 0.0011 0.0010

0.0023 0.0012 0.0012 0.0012 0.0010

blade surfaces with plasma ions takes place. It allows eliminating the preliminary surface preparation and ensures high adhesive strength at the interface of the substrate and the deposited coating.

Basic Elements of the Overlay Coating Technique This technique of coating turbine blades includes the following basic processes: producing ingots and cathodes for coatings; preparing blade surfaces for coating; depositing coatings by the EB or EA techniques; peening and heat treating coated blades; and inspecting the quality of coatings. Producing Ingots and Cathodes for Coatings. The ingots and cathodes produced for evaporation should have sound structure. After being turned, surfaces must be free of poor melted spots, discontinuities, nonmetallic inclusions, cracks, sinks, and pits. There must be no spray of metal from the liquid pool when evaporating the feedstock ingots and cathodes in compliance with the requirements set forth for deposition coatings. The alloys for coatings deposited by the EB technique are melted by a duplex process, including vacuum-induction (VI) melting and electron beam remelting (EBR). Only the VI melting is used for making cathodes. Instead of EBR, vacuum arc remelting (VAR) can be used, meeting all the requirements to gas content of the ingots and to their density. Only VI melting is used for producing the cathodes for electric arc evaporation. During VI melting, the alloying elements are introduced in the following succession: nickel, cobalt, chromium, and aluminum. As the melting process approaches its completion, yttrium (NiY, master alloy) is introduced into the melt. Analysis of variations in alloying element contents during VI melting and subsequent processes has shown that they are within the limits presented in Table 3.8. Such variation of yttrium concentration should be attributed to its oxidation during melting. The oxidation of yttrium in the melt can, however, be reduced by additional deoxidation of the melt before introducing yttrium into it. Deoxidation can be effected by introducing carbon and cerium in the amount of 0.01 to 0.1% either into the NiCr melt. Such a working cycle enables obtaining the materials with minimum content of gases and nonmetallic intrusions (Tables 3.9 and 3.10). Preliminary deoxidization of


the melt will reduce yttrium losses in melting by 20 to 25%. Homogeneity of ingot chemical composition over the ingot height (i.e., its macrosegregation value) is undoubtedly very important for ensuring coating process stability. As studies have shown, applying a two-stage process (VI Ⳮ EBR) reduces macrosegregation to a low level (Table 3.11).

Table 3.11

Macrosegregation in ingot Chemical composition (a), wt%

Cutout spot Ingot, top

Center Edge

Ingot, bottom

Center Edge

Cr

Co

Al

Y

14.45 13.22 14.42 12.42 14.58 14.12 14.42 14.25

29.84 25.93 29.84 26.27 30.33 28.36 30.33 28.28

8.19 7.84 8.12 7.80 8.15 8.16 8.22 8.38

0.08 0.035 0.12 0.036 0.12 0.041 0.08 0.038

(a) Top figure, vacuum-induction (VI) melting; bottom figure, VI melting Ⳮ electron-beam remelting (VI Ⳮ EBR)

Influence of surface preparation methods on the number of intercrystalline defects (“leaders”) in coating: (a) sand blasting (Ra 5–7 lm); (b) vibropolishing (Ra 1.25 lm). Source: Ref 36

Fig. 3.14

Preparing Blade Surfaces for Coating. Preparing operations are conducted to make blade surfaces activated, uniform in outer appearance, and clear of organic and nonorganic impurities. The quality of a finished coating to a great extent depends on how carefully the surface has been prepared. Preferable surface preparing treatments before coating include hydroabrasive blasting, vibropolishing, rinsing, and degreasing in aqueous and organic solutions. The EB technique imposes stricter requirements to the preparedness of surfaces onto which the evaporated material will be deposited. This is dictated by intercrystalline defects of structure (known as leaders) that tend to form on areas running at greater than 20 to 30 degrees with respect to the direction on which the vapor flux comes. Such intercrystalline defects should be attributed to the effects of shadowing microareas of surface with surface microirregularities (traces of machining, spalling of carbides, etc.) and to a sharp decrease of overlay buildup in such areas (Ref 34, 35). The occurrence of intercrystalline defects is assisted by rotation of blades in the vapor flux when each part of surface at a certain moment is presented to the vapor flux heading at an angle greater than 20 to 30. Intercrystalline defects formed in surface depressions are elongated in parallel with the axis about which the workpiece rotates in the vapor flux. Especially amenable to intercrystalline defects are convex surfaces (blade suction surfaces); conversely, concave surfaces (blade pressure surfaces) are much less vulnerable to the occurrence of such defects. Investigating how the surface preparation techniques influence the formation of intercrystalline defects revealed that the number of such defects and their depth penetration into the coating correlate with roughness parameters. Particularly coarse defects occur in sandblasted areas whose roughness, Ra, is 5 to 10 lm and on ground surfaces having rough notches elongated in parallel with the workpiece rotary axis. The defects grow through the full thickness of coating. The smaller the roughness, the fewer is the number of defects per unit area and the lesser the depth of the coating the defects penetrate. When applying the machining methods, which keep surface roughness within Ra ⱕ 1.25 lm, the number of intercrystalline defects in coatings up to 100 lm thick will not be great, and the defects will penetrate only through approximately 1⁄3 of coating thickness (Fig. 3.14) (Ref 36). Thus, any machining method applied to machined blade surfaces to which coatings will be


subsequently applied by the EB technique must ensure surface finish not worse than Ra ⱕ 1.25 lm. Various types of mechanical processing can be applied, the choice being dictated by the available equipment, such as hydroabrasive processing, vibropolishing, and so forth. Substantial rise of labor-intensity in further machining aimed at greater decrease of surface roughness is justifiable only if the coating working characteristics (thermal fatigue resistance in particular) are improved in this way. To lower the risk of intercrystalline defect occurrence in blades with high thermal stresses, the surfaces of such blades should be machined to R ⱕ 0.63 lm. Such surface finish can be obtained by vibropolishing, as a result of which the surface will acquire a smooth-bright homogeneous surface. Special bodies or granules are used to conduct the vibropolishing process. The vibropolishing procedure is followed by thorough rinsing and degreasing of blades. The surfaces of cast blades made of modern superalloys have quite a number of voids and

Influence of blade surface radius and heating temperature on the number of intercrystalline defects (“leaders”) in the coating over 400 lm length. Source: Ref 36

Fig. 3.15

Table 3.12 Effect of surface preparation on adhesion strength of Ni20Cr12AlY coating deposited by electric arc technique Method of surface preparation Manual polishing Vibropolishing Hydro-abrasive blasting Abrasive blasting with corundum (50–60 lm) Source: Ref 37

Adhesion strength, MPa 6 9 24 36

pipes whose size in cross section may reach 20 lm and coarse carbide inclusions. In the course of the machining process, these defects are partly preserved, the carbide inclusions having a knack of spalling off. It is at these voids, pipes, and spalled-off carbides where intercrystalline defects form during coating deposition. The defects can be minimized and surface roughness simultaneously reduced down to Ra ⳱ 0.32 lm with the following method of blade surface preparation. First, the blade airfoil is machined as described earlier. Then, on an appropriately prepared surface, a layer of Ni20Cr6AlY coating 20 to 40 lm thick is deposited. After annealing in vacuum at 1000 to 1050 C, the blades are vibropolished. Thanks to a high structural homogeneity and good machineability of the coating the vibropolishing will make the surface uniformly rough and not coarser than Ra ⳱ 0.32 lm throughout. It results in a defectless coating possessing a good combination of characteristics when tested for endurance and thermal fatigue. The number of intercrystalline defects in EBdeposited coatings will depend not only on how carefully the surface has been prepared, but also on the temperature of blade heating and on the radii of curvature of the surfaces coated. Heating blades above 1000 C will markedly reduce the number of coating defects (Fig. 3.15) (Ref 36). No defects occur, in fact, in the coatings deposited on concave surfaces of blades (pressure surfaces), yet their number grows with the decrease of curvature radii on the leading and trailing edges of blades. The EA technique imposes less stringent requirements on the preparation of blade surfaces on which coatings must be deposited. Though the demands of clearing the surface of organic and inorganic impurities remain the same, surface roughness may constitute Ra ⳱ 5 to 7 lm. Such surface finish can be obtained by processing the surface with synthetic corundum (with grain size of 30 to 60 lm). When depositing coating by this method, the turbine blade is not heated excessively, the chief aim being to ensure sufficiently high adhesion of coating and to prevent its spalling before high-temperature annealing. The relationship between the adhesion strength of EA-deposited coating and surface preparation technique is illustrated in Table 3.12 (Ref 37). All the measurements are carried out before high-temperature annealing. The basic preparatory operations include rinsing and degreasing of blade surfaces, such as


washing outside and inside surfaces with hot water, and rinsing outside and inside surfaces of blades in a degreasing aqueous solution. After washing and drying, the blades are weighed on scales loaded into a clean and closed container and conveyed to the coating deposition area. Note that the admissible period during which workpieces should be kept in a clean and closed metal container between the last preparatory operation and their placement into chamber for depositing and vacuuming must not exceed 10 h. Base Stages of the Technological Cycle for Coating Deposition by the EB Technique. The cycle of coating deposition includes the following steps: loading the unit with the blades; vacuum pumping to the working condition; blade heating to the required temperature; coating deposition; and bleeding-in and unloading the unit. Once the unit is loaded with the blades and the vacuum in the unit reaches 5 ⳯ 10ⳮ2 Pa, the electron guns of the evaporator come into operation. The number of evaporators working depends on blade sizes, blade holder design, and demands on the evenness of coating thickness. The coating alloy, while evaporating, must form near the workpiece surface a homogeneous and uniformly dense cloud of vapor, ensuring the required parameters of condensation (overlaying). An evenly thick coating with uniformly distributed alloying elements can be ensured by certain practical methods. One of them is ensuring continuous evaporation by continuously feeding the feedstock into a liquid pool. Meeting this requirement makes the evaporation process steady in time. Such time of stabilization depends on the chemical composition of the substance evaporated. It is also dependent on the conditions of the evaporation process, such as specific power input, size of liquid pool, and intensity of convection mixing processes going in the pool. The time necessary to stabilize the vapor is determined by experiment, that is, by analyzing the changes of condensate (coating) composition in time. Research into the evaporation of Ni18.6%Cr-12.5%Al-0.3%Y alloy in a watercooled crucible with the diameter of 70 mm in the UE unit (the dimensions are the same for all the units of this type) has shown that chromium and aluminum contents of the vapor stabilize within 15 to 20 min. Further on, the stability of condensate composition depends principally on the stability of EB gun operation coupled with

uniform feed rate of the ingots into the evaporation zone, position of a liquid pool, and some other factors. It takes much more time to stabilize the yttrium content of vapor. During the first 40 to 60 minutes, its content of the vapor remains within 0.01 to 0.05% and only gradually, after 3 to 4 h of evaporation, it will equalize the ingot level. By that moment, yttrium content of the liquid pool reaches 10 to 15% (Fig. 3.16). Introducing yttrium into the pool at the initial stage of evaporation can accelerate vapor stabilization of the yttrium content (or of that of some other refractory elements). The required quantity of yttrium is determined from the volume of liquid pool and yttrium content of it (the latter must be 10 to 15%), since such amount of yttrium will ensure its content of vapor within 0.1 to 0.3%. The same goal can be achieved with the aid of a special start ingot whose composition includes 10 to 15% content of the alloy. It should be noted, however, that increasing yttrium content of liquid pool would initiate the formation of the slag consisting of yttrium oxides on the pool surface. This decelerates the evaporation process and causes splashing. Raising the EB specific power brought to the pool surface gives rise to higher evaporation rates and speeds up stabilization of vapor elemental content. On the other hand, intensified evaporation results in overheating and splashing of liquid pool and in producing microdroplets that drop on workpiece surfaces and degrade coating quality. Lowering EB specific power lengthens the time of vapor stabilization and produces coatings of variable chemical composition within their thickness. The UE-type unit, having acceleration voltage of 18 kV and EB current of 1.8 to 2.2 A, will evaporate Ni20Cr12AlY and Ni20Co20Cr12AlY alloys at the rates of 30 to 50 g/min. When an EB coating technique is used, vapor condenses on heated blade surfaces. Structural features of the coating to be obtained dictate the choice of blade heating temperature in coating. The metallic coating must be as solid as possible and have a minimum number of intercrystalline defects, whereas the ceramic coating must have columnar structure whose porosity at crystallite height is kept at the minimum level. Its crystallite size must be 0.5 to 1.0 lm. There are two boundary temperatures, T1 and T2, at which condensate structure undergoes changes (Fig. 3.17) (Ref 38). In the first lowtemperature zone, at temperature below T1 (T1


Fig. 3.16

(a) Microstructure of Ni20Cr12AlY ingot bath after 20 deposition processes (ingot diameter, 70 mm) (500⳯). (b) Distribution of elements in bath after 20 deposition processes


Fig. 3.17

Influence of temperature on the structure of overlay coatings. Source: Ref 38

Table 3.13 Liquidus-solidus temperatures and crystallization intervals of Ni-Cr-Al and Ni-CoCr-Al alloys Alloy Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni30Co30Cr14Al Ni10Co22Cr11AlY Ni30Co22Cr11AlY Ni20Co15Cr11AlY Ni20Co20Cr11AlY Ni22Cr11AlY Ni14Cr3AlY Ni18Cr5AlY Ni16Cr9AlY Ni21Cr9AlY Ni18Cr16AlY Ni19Cr12AlY

Tliquidus, oC

Tsolidus, oC

Dcryst, oC

1357 1368 1320 1356 1452 1481 1463 1457 1422 1390 1379 1370 1360 1424 1407 1363 1346 1482 1400

1280 1284 1240 1278 1256 1279 1320 1304 1275 1255 1274 1275 1293 1384 1336 1305 1293 1275 1283

77 84 80 78 196 202 143 153 147 135 105 95 67 40 71 58 53 207 117

Distribution of Ni20Cr12AlY coating thickness across middle section of turbine blade. The blades are stationary, fixed in a cassette.

Fig. 3.18

⳱ 0.3 Tliquidus), the coating will embody separate cone-shaped crystallites and quite a number of micropores. The other temperature zone lies between T1 and T2 (T2 ⳱ 0.45 to 0.50 Tliquidus). The coatings in this zone will have columnar structure and be free of microporosity. Above T2, the coatings will have equiaxed structure, the formation associated with diffusion processes that occur between separate crystallites. Clearcut division into structural zones is typical of thick overlays of solid solutions of refined metals. In the two-phase systems, temperature T2 will be offset to a high-value range when highmelting phases (oxides, carbines, intermetallides, etc.) are involved. When depositing metallic coating on turbine blades, the temperature of blade heating should exceed T2 by 200 to 250 C. This is necessary to intensify the diffusion processes between the coating being formed and the blade surface. Liquidus and solidus temperatures for certain alloys can be found in Table 3.13. The regression equation, showing the relationship between liquidus and solidus temperature on the one hand and the alloying of the NiCo-Cr-Al system alloys on the other, is given in the Appendix. The time necessary to heat blades up to 850 to 950 C will vary with blade size and design of holders on which blades are secured. In the coaters of UE-type it usually takes 10 to 15 min. Blade movement over the evaporators is programmed to bring about the required spreading of the coating over the blade surface. The unit feature of great importance is the distance from the surface of metal evaporated to the blades. The amount of metal overlaid on blades will vary as the inverse square of the distance from the evaporator. Hence, the thickness of overlay on the blade airfoil will vary as a function of blade motion characteristics in the vapor cloud. Thus, coating will inevitably be thicker on those blade areas whose surfaces follow the circumference of maximum radius, approaching most closely to evaporator. Figure 3.18 shows a typical coating spread over the turbine blades. The coating has been deposited with the blades positively secured in relation to the holder rotary axis. Coating thickness on the pressure and suction surfaces can be varied by altering the angle at which the blade is secured on the holder and by using various masking. To guarantee coating evenness over the full blade airfoil surface, it is necessary to use the holders, which allow blade rotation about its own axis.


Maintaining the required temperature of blade surfaces in the process of coating deposition is undoubtedly of great importance when designing blade holders and their rotation parameters, since the rotation of blades in the coater affect temperature distribution over blade surfaces. Due to breaking of vacuum, breakdowns in high-voltage circuit, and so forth, the evaporation of material may cease when depositing coating on the blades. Ceasing of deposition will result in discontinuities in overlays and coating buildups. If this is the case when depositing a metallic coating, then, after the cause of failure is eliminated, the deposition process can be resumed and completed. In the event of a ceramic coating failure, the process must be stopped and the part rejected. The process of deposition being over, the holder with blades is handled to the loading chamber. The blades cool down there to 150 to 200 C. After the cooling is over, the chamber is opened and the holder with blades is taken out. The duration of the coater continuous operation will depend on the consumption of material evaporated. With 100 lm thick metallic coatings, the continuous operation time of the UEtype coater will be 10 to 15 h. Base Stages of Technological Cycle for Coatings Deposition by the Electric Arc Technique. The cathode material should be sound, without any cavities or other metallurgical defects. No cracks or other metallurgical defects are permissible on the cathode surface. When arrangements for fixing of the blades are designed, the possible nonuniformity of coating deposition along the height of the operating zone should be considered. In MAP unit for the blade with the height of 200 mm, nonuniformity of the coating thickness may be such that in the central zone the coating is up to twice as thick as in the peripheral zone. As a rule, all electric arc units are of batch operation, as the cathode assembly and the blades are situated in the same chamber. The coating deposition cycle includes loading of the parts into the unit; pumping to operating vacuum; ion clearing of the blades; coating deposition; bleeding-in; and unloading of the unit. After loading of the blades into the unit, the unit is covered with a cap and the unit chamber is pumped to the vacuum of 1 to 5 ⳯ 10ⳮ2 Pa. Pumping time is usually 30 to 40 min. After the required vacuum is reached, the process of ion clearing of the blades in the cathode material

plasma is carried out. To carry out this process of clearing, the electric arc is drawn between the cathode and the anode. The negative voltage of 150 to 1000 V is applied to the parts. Under these conditions ion current is 20 to 30 A. The process of ion clearing of blade surfaces takes 3 to 5 min; its end is determined from ion current stabilization on the blades. The clearing process is accompanied by uncontrolled part heating. It should be taken into consideration for thin-wall parts, as their heating beyond the temperature of 500 C is not permissible. When the stage of ion clearing is finished, the negative potential on the parts is lowered to 5 to 15 V. The operating arc current and its stable arcing is set by adjustment of electric parameters. No special blade heating typical of EB technique is carried out. When the unit electric parameters are stable, the cathode material evaporation rate and the coating thickness are proportional to electric arc current and time of coating deposition. In the electric-arc unit the material utilization factor depends on the number of blades and their arrangement in the chamber. It is usually 10 to 15%. After the process of coating deposition is over, bleeding-in is allowed and the holders with blades are removed from the unit and passed to the further procedures of the technological process. Peening Finished Overlay Coatings. Blades with overlay coatings undergo some (thermal and mechanical) finish treatment and processing. To obtain a higher coating-to-surface adhesion strength, the blades are annealed for 2 to 4 h under (1 to 4) ⳯ 10ⳮ2 Pa vacuum at 950 to 1050 C. The choice of annealing temperature is dictated by the need to preserve the mechanical characteristics of a superalloy and also by the formation of a diffusion zone of the alloy-coating boundary, thus assuring necessary strength of that boundary. The annealing operation is mandatory for EAdeposited coatings, as these have low adhesion immediately after coating deposition. The EBdeposited coatings demonstrate fairly high adhesion, provided all relevant technological parameters are strictly adhered to. This is why such coatings may be subjected to peening, obviating the annealing stage (Table 3.14). Peening EB-deposited coatings with 150 to 300 lm diameter micro-balls made of steel, superalloy, or glass is the next operation. Coatings deposited by EA technique feature dense structure without intercrystalline defects. They do not


require peening with micro-balls. In the course of peening, the coating is compacted due to high compression stresses (up to 1100 MPa). During peening, the intercrystalline defects, formed on the blade surface irregularities, are “squeezed.” Of utmost importance in peening is uniform cold working all over the blade surface coated. In this case, the micro-balls must hit the surface at an incidence angle close to 90. Generally it takes 3 to 4 min to blast around a blade up to 120 mm high. The surface roughness of the blades peened with micro-balls must not be more than 1.25 lm. Strain-hardening processing can involve other techniques, such as vibropolishing, glassball blasting and so forth. The chief requirement to any technique chosen is to obtain uniform cold working on all coated surfaces. Peening produces appreciable compression stresses in the coatings. In the event of poor adhesion of the coating, the latter may spall. After peening is completed, the blades are degreased, washed, and sent to recrystallization annealing. The temperature and time of such annealing are chosen so the superalloy preserves its mechanical properties and recrystallization processes in the coating are unimpeded. As a rule, the annealing is carried out in (1 to 5) ⳯ 10ⳮ2 Pa vacuum at 1000 to 1050 C for 2 to 4 h. Having passed peening and heat treatment, the coatings will have dense, equiaxed, fine-grained structure. When necessary, the coated blades may be subjected to additional heat treatment of aging if it is required to guarantee the specified mechanical properties. All blades are subject to mandatory checkcontrol operations to see whether the blades with heat-resistant overlay coatings meet the requirements of the process and design documents (Ref 39).

Table 3.14 coating Deposition technique EB EB EA EA

Adhesion strength of Ni20Cr12AlY Method of surface preparation, operation

Adhesion strength, MPa

Vibro-polishing, after deposition Vibro-polishing, annealing 1050 o C, 2 h Abrasive blasting with corundum, after deposition Abrasive blasting with corundum, annealing 1050 oC, 2 h

110 (a)

(a) Determined at tear-apart testing of brazed joints

130 (a) 36 130 (a)

Inspecting of Overlay Coatings. The quality of overlay is dependent on the stability of technique used and on the following characteristics. ● Thickness of coating and its even spreading over the blade airfoil must be consistent with preset specifications. The difference in coating thickness between batches of blades is tolerable within 10% or less. ● Deposited coatings must be uniform in composition through a full depth. Batch-to-batch elemental variation must be within the tolerances specified. ● After being deposited and annealed, the coating must be strong enough to prevent spalling in operation (for metallic coatings, radhesion 100 MPa; for ceramic ones, radhesion 70 MPa). ● The surface of EB-deposited coating must have surface roughness within Ra 1.25 lm after all processes have been completed. ● Voids, intercrystalline defects, and droplets are objectionable both on the coating surface and in its structure. No cracks whatsoever is tolerable. The coating quality will depend on the quality of superalloy ingots and ceramic briquettes delivered, on how thoroughly the conditions of blade preparation are met, and on how coated blades are processed. To meet all the abovementioned requirements, mandatory controlling operations are performed at all stages of the process (Ref 27). All as-delivered ingots of superalloys must have marks and a certificate with data on their chemical composition and the results of testing for splashing of the bath delivered. The ceramic briquettes are checked for quality by chemical and phase analyses of powder samples. Prior to depositing coatings on blades with prepared surfaces, the latter should be inspected visually for consistency with standard reference samples. Before securing blades to their holders, the blades should be checked by weighing to Ⳳ0.1 g accuracy. Flat blank-samples of wrought superalloy are loaded together with working blades into the holder. Such blank samples must have their surfaces prepared like those of the working blades. The EB UE-type unit controls the following parameters while depositing coatings on the blades: ● ●

Vacuum in the main process chamber Voltage and current of the electron beams of evaporating and preheating guns and geom-


etry of beam area on the surface of substance evaporated and parameters of its scanning ● Temperature of blades and their motion over the evaporator ● Level of pool of evaporated substance and rate of its feed to evaporators ● Rate of condensation (overlaying) of evaporated substance ● Time of coating deposition In the EA unit, the following parameters are controlled in work: ● ●

Vacuum in the main process chamber Voltage, arc current on cathode, and stability of arc burning ● Ionic current on blades ● Coating deposition cycle time All of these parameters should meet the requirements specified in the process documentation based on the results of test-operation during coating development. After the deposition process is over and the blades are taken out of the coater, the coating is inspected visually by comparing it with the reference sample. No droplets or any other defects are tolerable. The adhesion of coating is tested prior to (for EB technique) or after (for EA technique) annealing of the blades. Flat blank samples coated simultaneously with a batch of blades are tested. The samples are bent over a mandrel of 3 mm radius to the angle of 90. There must be no spalling during their bending. Flaws in the coating should not be regarded as grounds for rejection. The same requirements hold for ceramic coatings. If the blank-samples show poor adhesion while tested for bending, then all the blades from the same batch are rejected and sent for coating removal. After the surfaces have been prepared anew, they are recoated. Having passed peening, all blades are visually inspected for detection of spalling and swelling. Any blade with such a defect is rejected. Its coating is removed, and its surface is prepared anew and then recoated. Average thickness of blade coating is inferred from weight gain correlation charts plotted for each blade design version, such charts showing the relationship of coating thickness to blade mass gain after coating. Thickness of ceramic coatings can be examined using various thickness gages. Necessary chemical and spectroscopic techniques are used to analyze coating chemical

composition. Coating microstructure is inspected while running-in the working process, introducing some alterations, and selectively for some processes conducted within a month. All process operations being finished, the blades are flow checked.

REFERENCES

1. A.G. Andreeva, V.V. Terekova, and G.D. Fomenko, Heat-Resistance Coatings of Nickel-Base Alloys, High-Temperature Coatings, Leningrad, Nauka, 1967, p 96– 110 2. G.V. Zemskov and R.I. Kogan, Multicomponent Diffusion Saturation of Metals and Alloys, Metalurgiya, 1978, 207 p 3. P.T. Kolomytzev, Heat Resistant Diffusion Coatings, 1979, 272 p 4. N.V. Abraimov, High-Temperature Materials and Coatings for Gas Turbines, Mashinostroene, 1993, 336 p 5. B.N. Arzamasov, Chemical-Heat Treatment of Metals in Activated Gaseous Media, Mashinostroene, 1979, 224 p 6. B.N. Arzamasov, Processes and Equipment for Ecologically Sound Chemical Heat Treatment, Vestnik Machinostroeniy, N 5, 1996, p 26–28 7. B.N. Arzamasov, A.K. Beliavskiy, L.M. Mulakaev, and G.B. Stroganov, Circulating Method of Diffusion Coatings Application, Vestnik Machinostroeniy, N 11, 1991, p 43– 48 8. Yu.A. Tamarin, Heat Resistant Diffusion Coatings for Turbine Blades, Mashinostroene, 1978, 133 p 9. T.V. Levchenko, V.I. Moroz, and L.P. Buiyanova, Protective Coatings on Metals, Naukova Dumka, 4, 1971, p 158–164 10. T.I. Blinov, A.M. Dorodnov, V.Ye Minaychev, S.A. Muboyajan, et al., Vacuum High Current Plasma Devices and their Use in Microelectronics Production Equipment, Obzory po elektronnoy tekhnikie, 7 (268) and 8 (269), Moscow, CNIIElektronika, 1974 11. B.N. Barabanov, I.G. Blinov, and A.M. Dorodnov, High Energy Equipment: Cold Systems for Plasma Generation of Solid Conductors, Fizika I Khimiya Obrabotki Materialov, 1, 1978, p 44–46


12. S.A. Muboyadjan, Vacuum Plasma Unit of MAP for Depositing Protective Coatings on Machine Components, VIMI, NTD 6, 1989, p 7–11 13. E.N. Kablov, Cast Blades of Gas Turbine Engines, Moscow, MISIS, 2001, 632 p 14. E.B. Kachanov and Yu.A. Tamarin, Protective Coatings: Effective Path of Increase of Reliability of the Blades, Aviation Materials on the Eve of the 21st Century, Moscow, VIAM, 1994, p 296–304 15. S.A. Muboyadjan, S.A. Budinovsky, and E.N. Kablov, Turbine Blade Coating Deposition by Ion-Plasma Method, Aircraft Industry, 3–4, 1997, p 65–70 16. S.A. Budinovsky, E.N. Kablov, S.A. Muboyadjan, and Ya.A. Pomelov, High-Power Vacuum Plasma Technique as a Promising Means for Development of New Coatings and Materials, Aviation Materials on the Eve of 21st Century, Moscow, VIAM, 1994, p 314–325 17. L. Holland, Vacuum Deposition of Thin Films, John Wiley, New York, 1956, 135 p 18. R. Bakish, Ed., Introduction to Electron Beam Technology, John Wiley, New York, 1962, 395 p 19. J. Lafferti, Ed., Vacuum Arcs, Moscow, Mir, 1982, 432 p 20. A.M. Dorodnov, Some Areas of Application of Plasma Accelerators in Engineering, Physics and Application of Plasma Accelerators, Minsk, Nauka i Tekhnika, 1974, p 330–365 21. A.M. Dorodnov, Technological Plasma Accelerators, J. Tech. Phys., 1981, Vol 51 (No. 3), p 504–524 22. S.U. Kimblin, Erosion of Electrodes and Ionization Processes in Applied Areas of Vacuum Arcing and under Atmosphere Pressure, Experimental Studies of Plasmatrons, M.F. Jukov, Ed., Novosibirsk, Nauka, 1977, p 226–253 23. Yu.N. Andreev, T.M. Andronova, A.I. Vovsi, et.al., Electric Arc Sputtering of Metals and Alloys under Vacuum, Riga, LatNIINTI, 1982, 45 p 24. V.M. Lunev, V.G. Padalka, and V.M. Khoroshikh, Research into Some Plasma Characteristics of Vacuum Metal Arc, J. Technical Phys., 1977, Vol 47 (No. 7), p 1491– 1495 25. I.I. Aksenov et al., Research into Plasma of Steady Vacuum Arc Discharge, Thermal Physics of High Temperatures, 1983, Vol 21 (No. 4), p 646–651

26. O.N. Mironov, To Determination of Forces Affecting Stationary Hall Plasma Accelerators, J. Technical Phys., Vol 44 (No. 3), 1974, p 525–535 27. S.D. Grishin, L.V. Leskov, and N.P. Kozlov, Plasma Accelerators, Moscow, Mashinostroenie, 1983, 243 p 28. S.A. Leontiev, V.G. Kuznetsov, A.I. Rybnicov, and I.V. Burov, Structure and Properties of Protective Coatings by Vacuum Arc Deposition, Le Vide: Science-Technique et Application, N275, 1995, p 462– 466 29. B.A. Movchan and I.S. Malachenko, Heat Resistant Coatings Deposited in Vacuum, Naukova Dumka, 1983, 232 p 30. B.A. Movchan, EB-PVD Technology in Gas Turbine Industry: Present and Future, JOM, Nov 1996, p 40–45 31. V.G. Kuznetsov, C.A. Kukuchkin, and I.S. Polipanov, State-of-Art and Prospects of Developments in the Field of Vacuum Arc Coating Deposition, preprint IMASh RAN, Sankt-Petersburg, No. 105, 1994, p 7–14 32. V.G. Kuznetzov, “Interaction between Accelerates Metal Plasma Flux and Surface of Solid”, Autoreferat of Dissertation Dr. Sc., Sankt-Petersburg, 1997, 38 p 33. V.G. Kuznetzov, V.P. Bulatov, and S.I. Rybnikov, Unit for Coating Deposition under Vacuum. Patent 2155242, C23C14/16, Bulletin of Inventions No. 24, 27.08.2000 34. D.H. Boon, T.E. Strengman, and L.W. Wilson, Some Effects of Structure and Composition on the Properties of Electron Beam Deposited Coatings for Gas Turbine Superalloys, J. Vac. Sci. and Technol., Vol 11 (No. 4), 1974, p 641–646 35. V.I. Lachkov, A.V. Demchishin, I.S. Vasileva, et al., Structure of Titanium Coatings Deposited by Condensation of Non-Steady Atom Beams, Problems of Special Electric Metallurgy, Kiev, No. 9, 1978, p 67–77 36. P.A. Pap, “Research into Structure and Properties and Design of Electron Beam Technique for Deposition of Protective Coatings on Gas Turbine Blades,” Ph.D. dissertation, scientific consultant B.A. Movchan, IES, 1982 37. A.A. Dalskiy and K.M. Pasukov, Effect of Condition of Substrate Surface Layer of GS6U Alloy on Adhesive Strength of NiCrAlY Coating, Surface Layer and Reli-


ability of Machine and Instrument Components, MDNTP, Moscow, 1989, p 42–44 38. B.A. Movchan and A.V. Demchishin, Research into Structure and Properties of Thick Vacuum Condensate Layers of Titanium, Tungsten, Aluminum Oxide, Physics

of Metals and Metal Science, Vol 28 (No. 4), 1969, p 653–660 39. D.F. Betteridge and R.G. Ubank, Quality Control of High-Temperature Protective Coatings, Mater. Sci. and Technol., Vol 2 (No. 3), 1986, p 232–242



CHAPTER 4

Phase Composition and Structure of Coatings on Superalloys IN COATINGS obtained by aluminizing nickel alloys in media containing aluminum and its compounds, the following aluminides can be formed: NiAl3, Ni2Al3, NiAl, and Ni3Al. Their properties are indicated in Table 4.1 (Ref 1). The phase composition of diffusion coatings formed on superalloys at their saturation with aluminum is determined by the activity of the medium used for aluminizing, the aluminizing temperature, the thermodynamics of aluminide formation reactions, the kinetics of the diffusion process, and the superalloy chemical and phase compositions. The initial structural constitution and phase composition of overlay (condensation) coatings are dependent on the chemical and phase compositions of the alloy evaporated, the temperature of the surface upon which it is condensed, and the state of aggregation of the substance transferred (vapor, plasma).

Phase Composition and Structure of Diffusion Coatings Judging by chemical and phase compositions of the aluminides that are formed in the outer zone of the coating (the latter constituting 60 to 80% of coating thickness), all aluminizing media can be divided conventionally into those of high, moderate, and low activities (Table 4.2). The main phases of the outer zone of the coatings, produced in high-activity media, are NiAl3 or Ni2Al3 aluminides. Concentration of aluminum in the outer zone of such coatings is above 40%. In low-activity media, the outer zone of coatings is made up of NiAl compounds or of a mixture of NiAlⳭNi3Al compounds. In such

coatings, the outer zone contains less than 31% Al. In moderate-activity media, the bulk of the coating outer zone is generally made up of NiAl compound containing 32 to 38% Al. Diffusion Coatings Formed in High-Activity Media. High-activity media include powder blends with high free-aluminum content, slurry layers that contain powders of aluminum and its alloys, and layers based on aluminum and its alloys that are deposited by the electric arc (EA) technique. High-activity media can be employed to deposit diffusion coatings within a wide temperature range, beginning at 450 C. Diffusion Coatings Formed at Temperatures below 640 C. The EA-deposited coatings are applied without special preheating of the blades to be aluminized in EA units. That is why aluminum alloy layers are deposited on relatively cool (300 to 500 C) surfaces. As the aluminum alloy is being evaporated, a layer of that alloy is being formed on these surfaces. The diffusion coating is formed due to processes in the solid phase during the course of subsequent annealing at the temperature below its melting point (640 C). In the temperature range below the melting point of aluminum, the diffusion processes in aluminum-nickel and aluminum-nickel alloy systems can be thought of as the processes proceeding under rigorous boundary conditions, when the aluminum concentration on the surface of the formed diffusion layer is definite and time-constant. To study diffusion processes, thick aluminum layers (1 mm) were applied to nickel specimens and to the specimens of Ni20CrTi alloy (Ni base, 19 to 22% Cr, 0.15 to 0.35% Ti). The specimens were annealed under vacuum at 470, 500, 600, and 615 C. Annealing time was within the range of 2 to 74 h.


Table 4.1

Nickel-aluminum system aluminide properties (Ref 1) Crystalline structure

Aluminide NiAl3 (e-phase), 57.97 wt% Al

NiAl (b-phase), 36–22 wt% Al at 900 oC

Orthorhombic ˚ a ⳱ 6.611 A ˚ b ⳱ 7.367 A ˚ c ⳱ 4.812 A Rhombohedral ˚ a ⳱ 4.036 A ˚ c ⳱ 4.900 A c/a ⳱ 1.214 bcc ˚ a ⳱ 2.886 A

Ni3Al (c⬘-phase), 12.6–14.4 wt% Al at 20 oC

fcc ˚ a ⳱ 3.589 A

Ni2Al3 (d-phase), 44.7–40 wt% Al at 600–1120 oC

Heat of Density, g/m3, Hardness, formation, (theoretical) MPa kJ/mol

Tmelt, C

Homogeneity region

854

...

3.96

6100–7700

ⳮ38.0

1132

44.7–40% Al at 600–1120 oC

4.76

11,200

ⳮ57.0

Violet

1638

36–22% Al at 900 oC

5.35 ⳰ 6.50

5600–6200

ⳮ59.2

1380

12.6–14.4% Al at 20 oC

7.29

4600–5600

ⳮ39.0

Blue at Al content of 31% and yellow at Al content of 31% Pale yellow

o

Color ...

Note: bcc, body-centered cubic; fcc, face-centered cubic

After annealing the nickel-aluminum specimens under all the previously mentioned conditions, a violet-colored diffusion layer is observed at the interface of two metals (Fig. 4.1). Some “islets” of lighter phase are found at the diffusion layer/aluminum interface. The amount of this phase increases with the increase in annealing temperatures. The diffusion layer formed consists of Ni2Al3 compound (Table 4.3). The composition of the phase that forms islets corresponds to that of NiAl3 compound (59% Al). No other intermetallic compounds of the nickel-aluminum system have been detected in the diffusion layer. On annealing the specimens of the Ni20CrTiAl system in the temperature range of 470 to 615 C, the diffusion layer is formed at the interface of two metals. It consists of two zones of different-colored hues (Fig. 4.2). The crystalline

Table 4.2 Conventional classification of saturating media

Activity of saturating media

Aluminides forming outer zone of the coating Aluminum concentration of the outer zone, %

High Moderate Low

NiAl3, Ni2Al3; 40% Al NiAl; 32–38% Al NiAl, NiAlⳭNi3Al; 31% Al

structure of the zone adjacent to the aluminum layer is that of NiAl3 compound, while that of the zone adjacent to the alloy side is Ni2Al3. The patterns of nickel and chromium distribution through the thickness of the diffusion layer are similar. They feature the decrease in nickel concentration at the transition from one zone to another that is accompanied by the proportional decrease in chromium concentration.

Table 4.3 Phase and chemical compositions of diffusion layers formed at annealing of aluminumnickel and Al-Ni20CrTi specimens

Annealing conditions

Specimen

470 oC, 74 h

Al-Ni

600 oC, 9 h

Al-Ni

o

Al-Ni20CrTi

o

Al-Ni20CrTi

470 C, 74 h 600 C, 9 h

Chemical composition(a), wt%

Layer zone

Crystalline structure of the zone

Hardness, MPa

Ni

Cr

Al

Islets Outer Islets Outer Outer Inner Outer Inner

NiAl3 Ni2Al3 NiAl3 Ni2Al3 NiAl3 Ni2Al3 NiAl3 Ni2Al3

... 9500 ... 9500 8000 10,000 8000 10,000

41 55–56 41 55–56 30 40 26–29 37–42

... ... ... ... 6.8 9.3 6.0–6.7 8.3–9.6

59 45–44 59 45–44 63 50 68–64 54–48

(a) In a range, the first digit is the beginning of the zone (from aluminum side); the second one is its end.

Phase Composition and Structure of Coatings on Superalloys / 57

The increase in the thickness of the diffusion layers formed at the aluminum/nickel interface, as well as at the aluminum/NiCrTi alloy interface, fits the general pattern for all diffusion processes (Fig. 4.3a,b). To describe phase composition of the coatings, it is expedient to use special diagrams on the x-y coordinates, where the y-axis equals phase-zone thickness in the coating (lm) and the x-axis is saturation time (s1/2). The value equal to zero on y-axis represents the initial superalloy surface before its saturation with aluminum. The diagrams clearly demonstrate the location of aluminides, the ratios of their thickness as well as the changes in the sizes of the specimens during aluminizing, the kinetics of each aluminide zone growth, and the increase in the total coating thickness. In the cases where the kinetics of growth is determined by diffusion processes, the diagram of phase transformations represents straight lines dividing phase zones of the coating. Figure 4.4 represents phase diagrams typical of nickel and nickel-base superalloys at their sat-

Fig. 4.1

Diffusion layer between aluminum and nickel. Annealing at 600 C for 9 h. 365⳯

Fig. 4.2

Diffusion layer between aluminum and Ni20CrTi alloy. Annealing at 600 C for 9 h. 350⳯

uration from the aluminum layer deposited by the EA method at the temperature below its melting point. As mentioned previously, coating formation proceeds due to diffusion of aluminum through the aluminide layer to the coating/ alloy interface. The kinetics of the increase in coating thickness depend on aluminum diffusive mobility in the aluminides formed. Alloying with chromium retards aluminum diffusion; in turn, it results in the formation of NiAl3 aluminide in the outer zone. Phase compositions of the coatings identical to those mentioned previously are also observed on aluminizing from the mixture of 98% Al powder and 2% NH4Cl at 500 to 600 C. At aluminizing, a violet-colored coating is formed on the nickel surface. In its metallographic structure, the coating is similar to the diffusion layer formed at the annealing of nickel with a deposited aluminum layer. The aluminum content of the coating and its crystalline structure (Table 4.4) evidence that the coating consists of a single intermetallide zone of Ni2Al3 compound. The coatings forming at low-temperature aluminizing Ni20CrTi alloys and Ni10CrWMoCo superalloy (Fig. 4.5) have the same metallographic structure. It consists of two zones of different-colored hues. The outer zone is lighter and the inner one is darker. No other phase precipitates are detected in them. The hardness number of the outer zone for all the alloys is 7500 to 8000 MPa, while that of the inner zone is 10,000 MPa. Crystallographic structure of the outer zone corresponds to NiAl3 compound, while that of the inner zone corresponds to Ni2Al3 compound. Aluminum contents of the outer and inner zones are 69 to 62% and 56 to 49% respectively (Table 4.4). Nickel, chromium, and aluminum distribution patterns in the coating zones on Ni20CrTi alloy are the same as those of the layer formed at the annealing of aluminum-Ni20CrTi specimens. In the coating formed on Ni10CrWMoCo alloy, nickel and all its alloying elements (chromium, molybdenum, tungsten, cobalt) show the same patterns of their distribution through the zones. The proportional decrease in concentrations of these elements is observed in the direction from the base material toward the surface. For low-temperature aluminizing, the mixtures of aluminum-iron alloys with high aluminum content may be used. By saturation of JS6K superalloy from the mixture containing 98% Al-Fe alloy (75% Al) and 2% NH4Cl at 600 C, the coatings have been formed. Their structures and compositions are similar to those


of the coatings on Ni20CrTi and Ni10CrWMoCo alloys discussed previously. The aluminum content of the coating surface zone on JS6K superalloy is 60%, while that of the inner zone is 40 to 50%. The patterns of

Fig. 4.3

aluminum, chromium, molybdenum, and tungsten distribution are the same as at saturation from aluminum-base mixtures. They feature the stepwise decrease in aluminum concentration in the direction toward the base material, accom-

Thickness of diffusion layers between (a) aluminum and nickel and between (b) aluminum and Ni20CrTi alloy vs. annealing time. (A solid line is for the full layer; a dashed line is for its outer zone.)

80

80 Al

40

Thickness, µm

Thickness, µm

Al δ-Ni2Al3 0

40 ε-NiAl3 0

δ-Ni2Al3

40

40

γ-(Ni20CrTi)

γ-(Ni) 80

80 0

1

4

9

16

0

25

4

9

16

25

(b)

(a)

Fig. 4.4

1

Aluminizing time, h

Aluminizing time, h

Diagrams of phase transformations during (a) nickel and (b) Ni20CrTi alloy aluminizing from deposited aluminum layer at 600 C

Table 4.4 Chemical compositions of diffusion coatings formed on nickel and Ni20CrTi and Ni10CrWMoCo superalloys. Aluminizing in the mixture of 98% Al, and 2% NH4Cl at 580 oC for 4 h Alloy Ni Ni20CrTi Ni10CrWMoCo


Layer zone


Hardness, MPa

Ni

Cr

Mo

W

Co

Al

... Outer Inner Outer Inner

Ni2Al3 NiAl3 Ni2Al3 NiAl3 Ni2Al3

9500 8000 10,000 8000 10,000

56–56 25–27 33–40 21–26 31–36

... 5.9–6.4 7.7–9.3 3.0–3.5 4.2–4.8

... ... ... 2.7–3.5 4.0–5.0

... ... ... 1.6–2.0 2.5–2.6

... ... ... 2.0–2.3 2.7–3.0

44 69–66 59–51 69–62 56–49

(a) The first digit is the beginning of the zone; the second one is its end.


panied by the stepwise increase in the alloying element concentrations. Phase analysis allowed the detection of NiAl3, Ni2Al3, and NiAl aluminides in the coating. Thus, independently of the aluminizing technique chosen, the diffusion coatings applied in the high-activity media at the temperatures below aluminum melting point are NiAl3 and/or Ni2Al3 compound-base coatings. The presence of either of the aluminides depends on the saturation process temperature as well as on the alloy additives. Research into the composition and structure of NiAl3 and Ni2Al3 compound-base coatings applied from high-activity media made it possible to describe the general mechanism of their formation on nickel alloys (Ref 2, 3). As mentioned previously, the distribution patterns for nickel, chromium, and other alloying elements through the coating zones in diffusion layers between aluminum and Ni20CrTi alloy are the same in the coatings on Ni20CrTi and Ni10CrWMoCo alloys. Gradual decrease in nickel concentration in the direction toward the surface is accompanied by the decrease in the concentrations of all alloying elements. The formation of NiAl3 and Ni2Al3-base diffusion coating proceeds as follows. Because alu-

Diffusion coating on Ni10CrWMoCo superalloy. Aluminizing in the mixture of 98% Al and 2% NH4Cl at 580 C for 4 h. 300⳯

Fig. 4.5

minum has high mobility in these compounds, at aluminizing it diffuses to the depth equal to the coating thickness. At the same time, the alloying elements (such as nickel, chromium, tungsten, molybdenum, and cobalt) of the saturated alloy remain actually immobile, that is, as if the saturated alloy is diluted with diffusing aluminum. The increase in the coating thickness takes place at the inner interfaces, in fact at the outer zone/inner zone and inner zone/alloy interfaces. Coatings formed in high-activity media at the temperature below aluminum melting point demonstrate high hardness and low ductility. In addition, the melting point of NiAl3 compound is 854 C. That is why blades and other components with these coatings should undergo short-term annealing (for 1 to 2 h) at 950 to 1100 C before putting them into service. Diffusion Coatings Formed at Temperatures above 640 C. When the blades with layers of aluminum alloys deposited on their surfaces are annealed at temperatures above 640 C, the diffusion coatings are formed from the liquid phase. In this case, their structure and phase composition differ from that mentioned previously. If the annealing temperature exceeds that of melting of the deposited aluminum alloy, then an active liquid phase is formed on the saturated surface, in which aluminum content (mAl, g/m2) is limited by the thickness (d) of the primary layer deposited. In these circumstances, a diffusion coating forms a few minutes after the aluminum alloy heats to its melting point. Further elevation of temperature and lengthening of annealing time bring about diffusion processes occurring between the coating formed and the superalloy. As a result, aluminum concentration is lowered and coating thickness is increased. The Al5SiY layer (aluminum-base, 5% Si, 1.5% Y) was deposited by the EA technique onto the blades of JS6U superalloy. This was followed by annealing at 700, 800, 900, and 1000 C for 1 to 4 h. Table 4.5 shows the analysis data of phase and chemical compositions and the structure of the diffusion coatings formed. Initially, a layer is formed on the surface to be saturated, its chemical and phase compositions being similar to those of the aluminum alloy evaporated. Further on, the changes in chemical and phase compositions that take place on the saturated surface at annealing are fully de-


Table 4.5 Phase and chemical compositions of diffusion coatings on JS6U superalloy. Aluminizing by EA method from alloy Al-5%Si-1.5%Y. Specific mass of alloy deposited, mAl 100 g/m2 Chemical composition, wt% Annealing Without annealing 700 oC, 1 h 800 oC, 1 h 900 oC, 1 h 1000 oC, 2 h

Coating thickness, lm 40 45 50 55 65

Phase composition Al, Si Al, NiAl3, Ni2Al3 NiAl3, Ni2Al3 Ni2Al3 NiAl(a) NiAl(b) NiAl(c), carbides

Hardness, MPa

Al

Cr

Co

W

Si

Y

680 10,800 10,800 8900 6300 ... 6830

Base 70 48 40 26 26 13

... 2.5 4.5 5.0 7.0 3.7 13.4

... 3.0 5.0 6.0 ... ... ...

... 3.0 5.0 6.0 ... ... 17

5.0 3.5 2.5 2.0 1.4 0.1 0.5

1.5 1.0 0.7 0.6 0.1 0.1 0.1

(a), Outer zone of coating. (b) Middle zone of coating. (c) Intermediate zone of coating

Diffusion coating on JS6U superalloy. Electric arc deposition of Al-5%Si-1.5%Y alloy and annealing at 700 C for 1 h. 1000⳯

Fig. 4.6

pendent on the kinetics of diffusion processes occurring in aluminides. Annealing of specimens at 700 C results in the formation of a coating on their surfaces, due to dissolution of a thin layer of the superalloy in the liquid Al5SiY alloy. The layer is structurally dissimilar; the top part contains (NiAl3ⳭAl) eutectic while its inner layers are fully constituted of NiAl3, with Ni2Al3 compound occurring at the coating/alloy interface (Fig. 4.6). Annealing at 800 C leads to the formation of a Ni2Al3-base coating on the surface. Such an annealing temperature makes chromium, silicon, and tungsten spread evenly throughout the coating thickness. Elevating the annealing temperature to 900 C does not bring about any radical changes in the structure of the coating formed; its basis is still made up of Ni2Al3 compounds. After annealing at 1000 C, the outer zone of coating on the JS6U superalloy, as well as on other superalloys, consists of two parts whose basic component is NiAl. The first part lies on the coating surface and includes phase precipi-

tates based on refractory elements and silicon. The second part of the outer zone, bordering the inner zone, is free of precipitates and contains refractory elements in solid solution. Such an outer zone structure is formed due to the changes of diffusion processes taking place when the NiAl layer emerges in the coating. The phenomenon is typical of aluminide coatings produced from aluminizing media of high activity and subjected to diffusion annealing after saturation. The inner zone of coating, also based on NiAl, contains many carbide phases of M23C6 and M6C type (where M represents Cr, W). After being annealed at 1000 C, the structure and composition of coatings formed with aluminumion energy in excess of 100 eV and 15 to 20 eV are different (Fig. 4.7). The high energy of aluminum ions heats up the surface on which condensation (overlaying) occurs and simultaneously makes some part of aluminum reevaporate. This leads to decreasing the mass of aluminum deposited on the surface and to forming NiAl3, Ni2Al3 aluminide coatings at deposition. When diffusion annealing is applied, no liquid phase is formed on the surface, and, in this case, a coating is formed due to gradual lowering of aluminum content in growing aluminides. With annealing temperature elevated to 1100 to 1200 C and the annealing process taking 1 to 2 h, aluminum concentration in the coating outer zone lowers to 16 to 18%, and Ni3Al compound emerges in it. The higher the temperature of annealing, the lower the amount of phases precipitating in the outer zone and based on chromium, tungsten, and molybdenum, which gradually dissolve in aluminides. Elevating the annealing temperature to 1100 to 1200 C leads to coagulation of carbide phases in the innerzone layer. The latter tends to form a columnar-


Fig. 4.7

Diffusion coating on JS6U superalloy. Electric arc deposition of Al-5%Si-1.5%Y alloy and annealing at 1000 C for 2 h: (a) Aluminum-ion energy 100 eV. (b) Aluminum-ion energy 20 eV. 1000⳯, reflected electrons

like structure. After annealing at 1200 C, the inner zone is actually free of carbide phases. Typical of the previously described operational procedure, the basic characteristics— composition, structure, thickness—of the resultant diffusion coatings are fully dependent on the temperature and time of annealing as well as on the specific mass of aluminum layer deposited from plasma onto the surface (mAl). The specific mass of the deposited layer is related with the

composition and thickness of the diffusion coating formed from it by the following relationship: m

mAl ⳱

兺

di • Xi • qi

i⳱1

in g/m2, where di is the thickness of a certain (i) zone of coating, Xi is aluminum mass fraction in that zone, and qi is the density of that zone. At certain stages of its forming in the course of annealing, the diffusion coating is actually constituted of zones of successively arranged aluminides: NiAl3-Ni2Al3 NiAl(32 to 38% Al)NiAl(20 to 26% Al)-Ni3Al. From aluminum content in the coating aluminide zones, and from density of the zones, the following approximation can be written: mAl ⳱ 2.3 • dNiAl3 Ⳮ 2.1 • dNi2Al3 Ⳮ 1.9 • dNiAl(36%Al) Ⳮ 1.6 • dNiAl(25%Al) Ⳮ 1.1 • dNi3Al

Mass of deposited Al-5%Si-1.5%Y alloy vs. thickness of diffusion coating with the given phase composition theoretical relationship

Fig. 4.8

From this relationship, one may (judging by the specific mass of aluminum layer deposited on the specimen surface—mAl, g/m2) calculate the thickness of a diffusion coating formed from


it and having definite phase composition. One may also deduce how coating thickness varies under certain phase transformations. The results of such calculations are presented in Fig. 4.8. They correlate with the experimental data obtained when the relationships between the mass of deposited aluminum alloy and the thickness

of coatings and their compositions were investigated. On annealing at 1000 C for 2 to 4 h, the thickness (h) of the diffusion coating formed bears the following relationship (derived experimentally) to the specific mass of aluminum alloy deposited (50 to 70 g/m2): h mAl • 0.85.

Table 4.6 Phase and chemical compositions of outer zone of diffusion coatings on JS6K superalloy formed at aluminizing from slurries of different compositions Active component of slurry; annealing conditions



Ni

Cr

W

Mo

Co

Ti

Al

Other elements(b)

NiAl NiAl NiAl NiAl NiAl

Base Base Base Base Base

4.6 4.6 4.7 5.6 5.9

2.5 2.6 2.6 1.1 0.5

1.8 1.7 1.8 1.7 0.8

3.8 3.6 3.8 4.0 4.1

1.2 1.2 1.4 1.0 1.4

25 25 25 22 21

... 3.0–0.3% Si 3.7–0.3% Si 1.0–0.4% Nb 1.0–0.4% Ta

100% Al; 950 oC, 4 h (97.5% Al, 2.5% Si); 950 oC, 4 h (92.5% Al, 7.5% Si); 950 oC, 4 h (40% Al, 60% Nb); 1050 oC, 4 h (30% Al, 70% Ta); 1050 oC, 4 h

(a) Average content of the zone. (b) The first digit is the beginning of the zone; the second one is its end.

External surface of a coating

Phases thickness in a coating

Initial alloy boundary

NiAl (36% Al)

NiAl (24% Al)

Al Si Y

NiAl3

Ni3Al3

Ni3Al

Surface of aluminides formation in a coating

Superalloy (γ + γ′) I 600

II 700

III 800

IV 900

V 1000

1100

Temperature, °C

Fig. 4.9

Diagram of phase transformations during annealing of superalloy with deposited Al-Si-Y alloy layer

Table 4.7 Characteristics of diffusion coatings formed on nickel. Aluminizing in the mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 800 to 1100 oC Aluminizing conditions 950 oC, 36 h 1000 oC, 8 h 1100 oC, 2 h

Ordinal number of the zone from the surface

Zone thickness, lm

1 2 3 1 2 3 1 2 3

105 35 4 72 19 5 90 24 8

Zone color

Aluminum concentration(a), %

Hardness, MPa

Pale blue Yellow Pale yellow Pale blue Yellow Pale yellow Pale blue Yellow Pale yellow

36–36 28–22 15 36 24 15 36 24 15

8400–7800 5000 4000 8000 5000 4000 8000 5000 4000

(a) In a range, the first digit is the beginning of the zone; the second one is its end.

Aluminide; its crystal lattice parameter ˚ NiAl; 2.86 A ˚ NiAl; 2.86 A ˚ Ni3Al; 3.56 A ˚ NiAl; 2.86 A ˚ NiAl; 2.86 A ˚ Ni3Al; 3.56 A ˚ NiAl; 2.86 A ˚ NiAl; 2.86 A ˚ Ni3Al; 3.56 A


The aluminum content of the outer zone of such a coating is 22 to 24%. The behavior of elements (silicon, yttrium, etc.) introduced into the aluminum alloy for alloying of the coating formed depends on the thermodynamic activity of these elements as related to the elements in the aluminized superalloy as well as on their solubility in the aluminides formed. At annealing temperatures up to 700 C, silicon is uniformly distributed throughout the (AlⳭNiAl3) eutectic layer and then throughout the NiAl3 layer. Its redistribution occurs at temperatures above 800 C, at the moment when phases are formed on the basis of refractory elements. Silicon is included into these phases, locating in the outer-zone surface layer and in the inner zone of the coating. Yttrium content of the coating does not exceed its solubility level in aluminides (0.1%). The mechanism of diffusion-coating formation, when employing aluminum powder containing slurries as well as aluminum Ⳮ silicon powder containing slurries, is similar to that mentioned previously. Coatings are formed through a liquid phase of aluminum alloy followed by diffusion processes evolving between the aluminides and superalloy. Phase and chem-

Diffusion coating on nickel. Aluminizing in mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 950 C for 36 h. 300⳯

Fig. 4.10

ical compositions of diffusion coatings, obtained from slurry processes, are similar to those mentioned previously. The introduction of additional elements (silicon, niobium, tantalum, etc.) into the active component of the slurry makes it possible to produce diffusion coatings alloyed with these elements (Table 4.6). The process taking place upon formation of diffusion coatings at the temperatures above the melting point of the deposited aluminum alloy is shown in Fig. 4.9. At the first stage (AlⳭNiAl3), eutectic formation takes place. Then, NiAl3 aluminide-base zones form. The subsequent process kinetics depend on aluminum diffusive mobility in the growing aluminides at the annealing temperature and on the amount of aluminum deposited. At 800 C, the Ni3Al zone transforms completely into Ni2Al3 within 1 h period. After it, a thin zone with the structure of NiAl aluminide forms at a coating/superalloy interface. Under these conditions, all diffusion processes are retarded. At the annealing temperature above 1000 C, Ni3Al phase forms at the interface of the alloy and the zone with the NiAl structure. Thus, diffusion processes during annealing of the coating formed in the high-activity media involve gradual changes of aluminide types in the following succession: NiAl3→Ni2Al3→NiAl→Ni3Al. When annealing temperature exceeds 800 C, the outer and inner zones undergo precipitation of refractory element-base phases. These elements (molybdenum, tungsten, chromium, etc.) have low solubility in aluminides. This effect is useful, because it imparts higher hot corrosion resistance to diffusion coatings. Diffusion Coatings Formed in Moderateand Low-Activity Media. When aluminizing superalloys in moderately active media, diffusion coatings are formed. Such coatings are based on NiAl and contain 32 to 38% Al. A representative mixture of the kind that is broadly used consists of 98% Al-Fe (50% Al) alloy and 2% NH4Cl. After aluminizing nickel in the mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 950, 1000, and 1100 C, the diffusion coatings formed consist of three intermetallic zones (Table 4.7, Fig. 4.10). The surface zone is NiAl compound with aluminum content of 36%. The second zone also consists of NiAl compound with aluminum content of 28 to 22%. Although the crystalline structures and the lattice parameters of these zones are the same, there is an interface between them, with a stepwise change


80

40 NiAl (36% Al) 0

NiAl (25% Al)

40 Ni 80 0

1

Ni3Al

4 9 16 Aluminizing time, h

External zone NiAl (36% Al)

40

Internal zone NiAl (25% Al)

0

Superalloy

40

Ni3Al + phases based on refractory elements

80 25

(a)

Fig. 4.11

Thickness, µm

Thickness, µm

80

0

1

4 9 16 Aluminizing time, h

25

(b)

Diagram of phase transformations on (a) nickel and (b) JS6U superalloy surfaces during aluminizing at 1000 C

in the aluminum content through it. Chemical etching develops the third (deepest) zone. Both aluminum content and crystalline structure of this zone are typical of Ni3Al compound. Figure 4.11(a) shows a diagram of phase transformations developing on a nickel surface at its aluminizing in the mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 1000 C. Within the full aluminizing temperature range, it is observed that iron diffuses into the coating to a depth of not more than 10 lm. Iron content of the coating surface layer is below 1.0%. The relative positions of Ni2Al3, NiAl, and Ni3Al compounds in diffusion coatings on a nickel surface corroborate the heterophase diffusion theory statement concerning phase positions in the phase diagram. Yet, the patterns of aluminum and nickel distribution through each coating zone differ substantially from those expected (Ref 4). The main distinction is that within the region of NiAl compound homogeneity, two diffusion zones of different chemical compositions form. Those are NiAl (36% Al) and NiAl (28 to 22% Al). The zones demonstrate different colors and hardness numbers. A concentration “step” is observed at their interface. Such an abnormality is due to crystalline structure features of NiAl compound and some specialties of diffusion in this compound (Ref 5). Experimental data allow estimation of the relative rate of nickel-atom diffusion toward the specimen surface as well as that of aluminumatom diffusion from the surface inside the bulk specimen. As calculations have shown, after alu-

minizing, almost all aluminum atoms are in intermetallic zones. The amount of aluminum that has diffused through these zones and dissolved in nickel is negligible. If the zone composition and thickness are known, it is possible to calculate the amount of nickel and aluminum taking part in opposing diffusion through the sec-

Diffusion coating on JS-type superalloy. Aluminizing in mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 1100 C for 8 h. 300⳯

Fig. 4.12


the size of less than 1 lm are observed in this zone. The inner zone consists of the mixture of a few phases. There are no sharp boundaries between the inner and outer zones nor between the inner zone and the base material. The compound with the NiAl structure and the aluminum content of 34 to 36% is the base of the outer zone of the diffusion coating on superalloys after they aluminize. Some chromium-, tungsten-, molybdenum-, and titanium-base phases precipitate in the zone, but their fraction is relatively little. The inner-coating zone consists of the mixture of dispersed phases. Total nickel, chromium, tungsten, and cobalt content of the inner zone is 84 to 87%. Phase analysis of the diffusion coating has revealed that its inner-zone ma-

ond zone to form the first and third zones. Relative rates of nickel- and aluminum-atom diffusion through the second zone are equal to 41 to 1, respectively. Thus, at aluminizing, the directed flow of nickel atoms moves through both the second and third zones toward the first zone, while the opposite flow of aluminum atoms is less by a factor of 41 (Ref 4). After aluminizing nickel alloys in the mixture consisting of 98% Fe-Al alloy (50% Al) at 950 to 1100 C for 2 to 8 h, the diffusion coatings on their surfaces show two structural zones different in their metallography (Fig. 4.12, Table 4.8). The outer zone that is 70 to 80% of the coating thickness may have different-colored hues, depending on aluminizing temperature. Some isolated inclusions of a lighter phase with

Table 4.8 Characteristics of diffusion coatings formed on nickel alloys. Aluminizing in the mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 950 oC for 4 h Chemical composition, wt% Alloy Ni20CrTi Ni10CrWMoCo JS6K

Coating zone

Ni

Cr

W

Mo

Co

Ti

Fe

Al


Hardness, MPa

Outer Inner Outer Inner Outer Inner

Base Base Base Base Base Base

1.8 49 3.0 14 5.3 14.7

... ... 2.2 7.0 2.1 5.9

... ... 4.7 18 2.2 6.0

... ... 3.0 4.0 2.8 3.4

... ... ... ... 0.5 4.5

... 1.0 1.0 0.3 0.8 0.2

35 16 35 14 36 23

NiAl ... NiAl ... NiAl ...

8200 820 8400 8000 8200 9500

Table 4.9 Phase and chemical compositions of diffusion coating formed on JS6K superalloy in lowactivity powder mixtures Chemical composition, wt% Saturating mixture and conditions 98% Al-Fe alloy (40% Al), 2% NH4Cl; 950 oC, 4 h Powder mixture (5% Al, 45% Cr, 50% Al2O3); vacuum, 1000 oC, 6 h

Table 4.10

Coating zone

Ni

Cr

Outer Inner Outer Inner

Base Base Base Base

1.2 15 5.0 10.0

Ti

Al

Phase composition

0.1 0.1 4.0 0.3 6.0 6.2 3.8 4.8 Was not analyzed Was not analyzed

W

Mo

Co

24 10 24 15

NiAl NiAl, Ni3Al, carbides NiAl NiAl, Ni3Al, carbides

Chemical and phase composition of Ni-Cr-Al alloys Chemical composition, wt%

Alloy Ni22Cr11AlY Ni14Cr3AlY Ni18Cr5AlY Ni16Cr9AlY Ni21Cr9AlY Ni18Cr16AlY Ni19Cr12AlY Ni8Cr18AlY Ni8Cr15AlY Ni7Cr12AlY Ni7Cr10AlY Ni20Cr6AlY Ni28Cr11AlY Ni5Cr25Al

Ni

Cr

Al

Y

Phase composition at 20 oC

b, %

Base Base Base Base Base Base Base Base Base Base Base Base Base Base

22.3 14.7 17.7 16.0 20.8 18.3 19.5 8.0 8.0 7.0 7.2 20.0 28.5 5.0

11.2 2.8 5.5 9.1 8.8 15.7 12.6 18.0 15.0 12.0 10.0 6.0 11.0 25.0

0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 ...

c, c⬘, b, ␣ c c, c⬘ c, c⬘, b c, c⬘, b, ␣ c⬘, b, ␣ c, c⬘, b, ␣ c⬘, b c⬘, b c⬘, b c⬘ c, c⬘ c, c⬘, b, ␣ b

50 ... ... 15 20 80 60 95 70 15 ... ... 45 100


trix consists of NiAl and Ni3Al compounds. The precipitation of the phases on the base of refractory elements (chromium, molybdenum, tungsten, etc.), which alloy superalloys, has also been detected. Growth kinetics of diffusion coatings on nickel alloys, within the accuracy of measurement, is the same as growth kinetics of the coatings on nickel. The changes in aluminide phase size shown in Fig. 4.11(b) are brought about by the processes taking place in the coatings on su-

Fig. 4.13

peralloys. The thickness of outer-zone diffusion coating on superalloys coincides with the zone of NiAl compound (36% Al). The inner zone includes NiAl (24% Al) and Ni3Al aluminide layers, in which refractory element-base phases precipitate; the crystallographic structure and morphology of such phases are dependent on alloy composition and aluminizing temperature. The dependence of superalloy diffusion-coating thickness on the temperature and the time of the aluminizing process is similar to the same

Microstructure of the alloys (a) Ni14Cr3AlY. (b) Ni16Cr9AlY. (c) Ni22Cr11AlY. (d) Ni7Cr12AlY. (e) Ni8Cr15AlY. (f) Ni8Cr18AlY. 500⳯


dependence for nickel. The coincidence should be attributed to the fact that refractory elements are difficult to solubilize in aluminides and actually do not affect the activation energy of aluminum and nickel diffusion in that compound. Mixtures of moderate activity occupy intermediate position in a conventional gradation of aluminizing media. This is why, under certain conditions, the coatings obtained from such mixtures may have structures and compositions close to those of coatings obtained from highor low-activity mixtures. Thus, at 800 and 900 C, a thin layer of Ni2Al3 compound constituting the substrate of the outer-zone coating is formed on the coating surface, such coatings being deposited from highly active mixtures. On the other hand, repetitive applications of a moderately active mixture, which is inevitable under manufacturing conditions, lead to lowering aluminum concentration in the outer-zone coating and forming coatings whose composition and

Thermal expansion coefficients vs. temperature. (a) 1, Ni5Cr25Al alloy; 2, Ni7Cr12AlY alloy; 3, Ni8Cr15AlY alloy. (b) 1, Ni22Cr11AlY alloy; 2, Ni18Cr5AlY alloy; 3, Ni16Cr9AlY alloy; 4, Ni14Cr3AlY alloy

Fig. 4.14

structure are typical of those produced from lowactivity mixtures. Low-activity mixtures may be needed to obtain diffusion-aluminide coatings possessing higher ductility and also when depositing coatings onto the insides of cooled blades. Chemical compositions of the coatings formed on JS6K superalloy in low-activity mixtures are presented in Table 4.9. The distinguishing feature of the process is that the NiAl compound (constituting the substrate of the outer zone of coating) does not contain refractory elements. The inner zone of the coating has a multiphase structure, comprised of (NiAl, Ni3Al) matrix and refractory-element based phases. The same structure and phase composition have a diffusion coating forming under vacuum on the inside of cooled turbine blades at 1000 C from a pouredin mixture of aluminum (5%), chromium (45%), and Al2O3 (50%) powders. The kinetics of forming a coating from lowactivity mixtures depend on the type of mixture used, the ratio between the amount of aluminum in the mixture and the quantity spent for coating buildup, and finally on the kinetics of reaction of halide or vapor formation. If the aluminizing process is conducted in a vacuum using a powder mixture of aluminum, chromium, and aluminum oxide, the kinetics of the process slow down with time, which is due to rapid mixture depletion and a corresponding drop of aluminum and chromium vapor pressures. As revealed by the results of research into diffusion coatings applied at 950 to 1000 C in the moderately active mixtures, the outer zone of the coatings on nickel and nickel alloys consists of NiAl compound with aluminum content of 34 to 36%. The inner zone of the coating on nickel consists of NiAl and Ni3Al compounds going one after another. Some phases on the base of refractory elements (such as chromium, molybdenum, tungsten, etc.) precipitate in the innerzone matrix that consists of NiAl and Ni3Al compounds in the coatings on superalloys. The amount, fineness, and crystalline structure of these phases depend on the alloy composition and the aluminizing temperature conditions. The distribution patterns of alloying elements in the diffusion coatings formed on nickel alloys in moderate- and low-activity mixtures in the temperature range of 950 to 1100 C are basically different from those typical of the coatings formed in highly active mixtures. Chromium, tungsten, and molybdenum contents of the outer zone are relatively low. At the same time, the


concentration of these elements in the inner zone is higher than their concentration in the superalloy. The mechanism of coating formation in the moderately active media can be represented as follows (Ref 2, 6). Almost all aluminum taken up from the saturating medium is spent on the formation of the outer zone of the NiAl compound-base coating. This compound is formed due to aluminum flowing from the outer boundary and nickel atoms diffusing through the inner zone from the alloy. Chromium-, molybdenum-, and tungsten-atom diffusion rates are much less than nickel-atom diffusion rate in the same direction. The coating inner zone is formed due to partial depletion of deeper alloy layers of nickel diffusing into the outer zone. Chromium, molybdenum, and tungsten are practically immobile under these conditions. Due to their low solubility in the growing aluminides, they form their own phases. When coating outer-zone formation proceeds by this mechanism, it can take place at either of two interfaces. In the case of high-aluminum-diffusive mobility, it may be at the interface between the inner and outer zones, whereas in the cases where nickel-diffusive mobility in the aluminide of the outer zone is high,

Table 4.11

it may take place at the outer-coating boundary. The first of the previously mentioned cases is typical of the process of forming the NiAl compound-base outer-zone coating with aluminum content beyond its stoichiometric concentration (31.5%) in the media of moderate activity. The second one occurs at forming the NiAl compound-base outer zone with aluminum content below its stoichiometric concentration in the media of low activity. When NiAl formation takes place at the coating surface, low diffusive mobility of refractory elements in it results in considerable reduction of their concentration in the outer zone. No chromium-, molybdenum-, and tungsten-base phases are detected in this zone. This mechanism of forming diffusion coatings on nickel superalloys accounts for the experimental fact that at aluminizing under the conditions discussed, the outer zones of the coatings on nickel and nickel superalloys demonstrate the same size, structure, and aluminum contents. Calculated specific-mass gain of the specimens of nickel alloy aluminized (mAl) at the expense of aluminum spent on coating formation

Chemical and phase composition of Ni-Co-Cr-Al alloys Chemical composition, wt%

No.

Alloy(a)

Ni

Co

Cr

Al

Y


Qb(b), %

Qb(c), %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni30Co30Cr14Al Ni10Co15Cr8AlY Ni30Co15Cr8AlY Ni10Co30Cr8AlY Ni30Co30Cr8AlY Ni10Co15Cr14AlY Ni30Co15Cr14AlY Ni10Co30Cr14AlY Ni30Co30Cr14AlY Ni10Co22Cr11AlY Ni30Co22Cr11AlY Ni20Co15Cr11AlY Ni20Co30Cr11AlY Ni20Co22Cr8AlY Ni20Co22Cr14AlY Ni20Co22Cr11Al Ni20Co22Cr11AlY Ni20Co22Cr11AlY Ni20Co22Cr11AlY

Base Base Base Base Base Base Base Base Base Base base Base Base Base Base Base Base Base Base Base Base Base Base base base Base

11.3 27.5 10.1 29.4 10.5 27.5 10.6 29.7 11.2 30.2 11.1 28.8 10.4 29.5 10.5 29.9 10.1 29.8 19.6 19.7 19.7 19.8 20.3 19.7 19.6 22.5

15.5 14.4 28.5 29.6 15.1 15.2 30.0 29.7 15.2 14.7 29.3 29.5 14.6 15.9 28.8 29.7 22.2 22.2 14.6 29.3 22.2 22.7 21.7 22.7 22.5 22.5

8.9 9.3 7.8 7.8 13.5 13.6 14.1 14.1 8.2 8.1 8.0 8.1 13.6 13.8 13.8 14.1 10.6 10.9 11.2 11.1 8.3 13.7 11.1 11.1 11.3 11.2

0.04 0.05 0.11 0.10 0.07 0.08 0.06 0.08 0.04 0.45 0.20 0.33 0.04 0.25 0.32 0.46 0.30 0.23 0.35 0.33 0.40 0.24 0.04 0.28 0.19 0.39

c, c⬘, b c, c⬘, b c, c⬘, b, ␣ c, c⬘, b, r c, c⬘, b, ␣ b, ␣ b, ␣ b, r c, c⬘, b c, c⬘, b c, c⬘, b, ␣ c, c⬘, b, r c, c⬘, b, ␣ b, ␣ b, ␣ b, r c, c⬘, b, ␣ c, c⬘, b, r c, c⬘, b b, ␣ c, c⬘, b, ␣ c, c⬘, b, ␣ c, c⬘, b, ␣ c, c⬘, b, ␣ c, c⬘, b, ␣ c, c⬘, b, ␣

36 33 60 68 71 78 78 91 ... ... ... ... ... ... ... ... 63 76 53 72 35 79 60 60 60 60

38 33 42 48 87 79 76 68 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

(a) The alloys (except No. 26) include: 1.5% Hf, 1.0%W, 0.5% Nb, and 0.5% Si. (b) Phase quantity (Qb, vol%) in original state. (c) Phase quantity (Qb, vol%) after annealing at 1200 oC for 200 h


Fig. 4.15

Microstructure of the alloys. (a) Ni10Co15Cr8AlY. (b) Ni30Co15Cr8AlY. (c) Ni10Co30Cr8AlY. (d) Ni30Co30Cr8AlY. (e) Ni10Co22Cr11AlY. (f) Ni30Co22Cr11AlY. (g) Ni20Co15Cr11AlY. (h) Ni20Co30Cr11AlY. 335⳯

Table 4.12 Chemical composition of structural constituents in Ni-Co-Cr-Al alloys Chemical composition, wt% b-phase(a)

c-cⴕ

Alloy

Al

Cr

Co

Al

Cr

Co

Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni30Co30Cr14Al

17 18 14 ... 18 18 20 21

10 7.5 26 ... 13 14 11 19

7.5 20 8.0 ... 10 27 11 31

6.5 6.0 4.5 4.5 4.5 4.5 ... ...

21 18 36 36 32 30 ... ...

13 33 11 33 17 38 ... ...

(a) The quoted average chemical composition includes ␣-Cr: 70% Cr, 3% Al, and 10 to 30% Co.

actually coincides with the experimental value of specimen mass-specific change after aluminizing. Thus, when determined after aluminizing, the specific change of specimen mass interrelates two basic parameters of aluminide coating: concentration of aluminum in a coating

and its thickness. The postaluminizing specific change of specimen mass defines the “aluminum reserve” in aluminide coating and is a vital characteristic applicable both to control of coatings and to comparison of their protection properties.

Phase Composition and Structure of NiCr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al Alloys and Overlay Coatings Made of Them After being deposited according to established technology, the overlay coatings of the Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al systems have chemical and phase compositions very close to those of the alloy evaporated. The research into Ni-Cr-Al-, Ni-Co-Cr-Al-, and Co-Ni-Cr-Al-system alloys was arranged to determine the relationship between variations in chemical composition and the resultant changes


of structure and properties of the alloys. The range of composition the overlay coatings may have after they have been deposited and subjected to various high-temperature tests is investigated. Ni-Cr-Al Alloys. The alloy group under consideration and the coatings made of them consist of the following main phases: c-Ni-base solid solution, c⬘-Ni3Al phase, b-NiAl phase, and ␣Cr-base solid solution (Ref 7, 8). The number of phases and temperature fields of their existence depend on alloy chemical composition. Table 4.10 gives chemical composition and phase constitution of investigated alloys after the latter had been annealed at 1100 C for 4 h; typical structures of alloys are shown in Fig. 4.13. Phase transformations in Ni-Cr-Al alloys were examined by dilatometric analysis (thermal expansion coefficient values are given in Chapter 7, “The Effect of Protective Coatings on the Mechanical Properties of Superalloys”). Specimens for this were made by extrusion from cast blanks. Their microstructure was close to overlay-coating structure. Prior to testing, the specimens had been annealed at 1100 C for 2 h. Alloys whose compositions fall within the phase fields of c, c⬘-(Ni7Cr12AlY), (bⳭc⬘)(Ni8Cr15AlY), and b-(Ni5Cr25Al) have no phase transformations throughout the full range of temperatures up to 1000 C (Fig. 4.14 a,b). In the Ni18Cr6AlY alloy, whose composition falls within a two-phase field (cⳭc⬘), marked enlargement was noted at 700 to 1000 C, attributable to c⬘-phase solution. During cooling, the reversed transformation, c→c⬘, evolves without hysteresis in the same interval of temperatures. The transformation c↔c⬘ is typical of all the Ni-Cr-Al system alloys under consideration that have chromium content above 17%. In contrast to c↔c⬘ transformation, the phase reaction bⳭc↔c⬘Ⳮ␣ that takes place at 1020 C involves considerable diffusion motion of aluminum, nickel, and chromium. It proceeds with time hysteresis, which depends on temperature, phase sizes, and forms. During cooling, the bⳭc→c⬘Ⳮ␣ process occurs with lessening of volume. Thus, within the range of compositions under consideration, a diversity of phase transformations and, hence, volumetric changes has been noted, with all these processes taking place within the interval of temperatures at which turbine blades run. Ni-Co-Cr-Al Alloys. Experiments were planned statistically to enable research into alloy

properties for the Ni-Co-Cr-Al-system coating deposition. The compositions of the alloys produced are given in Table 4.11. In this plan of experiments, cobalt (x1), chromium (x2), aluminum (x3), and yttrium (x4) are variables. A regression equation, obtained by statistical processing the experimental data, correlates the property under study with the content of alloying elements (Ref 9). Figure 4.15 shows typical structures of the alloys studied (Ref 10), while phase compositions of the alloys determined from the data of x-ray structural and metallographic analyses are given in Table 4.11. The main phase constituents of these alloys are c-Ni-base solid solution, b(Ni,Co)Al phase, c⬘-(Ni,Co)3Al phase, and ␣Cr-base solid solution. (For the sake of simplicity, these phases are subsequently written as phases of the Ni-Cr-Al system.) In addition to these listed phases, still another r-phase of CoCr type evolves in high-chromium- and cobalt-content field. Chemical compositions of some alloy phases can be found in Table 4.12. It is the amount of NiAl phase in the coating structure that is most essential for the coatingprotection capacity and physical-mechanical properties. The amount was determined metallographically using the Quantimet instrument for two conditions: as-cast and postannealing at 1200 C for 200 h. Regression equations, interrelating the amount of b-phase (Qb) with the alloying of alloys, have been formulated while statistically processing the results of metallographic measurements. These results are presented in the Appendix. Aluminum is the main element controlling the amount of b-phase in alloy and coating structures (Fig. 4.16). With aluminum concentration rising from 8 to 14% and with chromium and cobalt being minimum (15 and 10%, respectively), the amount of b-phase in cast alloys is doubled, from 35 to 70%. Cobalt does not appreciably affect the amount of b-phase; chromium augments the phase content. In alloys containing minimum values of aluminum (8%) and cobalt (10%) and with chromium concentration rising from 15 to 30%, the amount of b-phase increases from 40 to 60%. In addition to b, c⬘, and c-solid solution, the structure of the studied alloys contains ␣-Cr and r-phases. As for ␣-Cr in b-phase, it should be attributed to its oversaturation with chromium and to precipitation at cooling of ␣-Cr particles coherent with the matrix. The structure of the


alloys also has eutectic (bⳭ␣-Cr) when chromium and aluminum are present in their maximum contents (30 and 14%, respectively) in the structure. The r-phase occurrences are typical of alloys with maximum content of cobalt and chromium (Ni30Co30Cr14AlY). Platelike rphase precipitation dramatically degrades the ductility of alloys and, hence, the coatings thereof. Phase transformations, which occur in Ni-CoCr-Al alloys during heating and cooling, were studied dilatometrically on sample pieces extruded from cast blanks and annealed at 1100 C for 2 h. (TEC values are given in Chapter 6, “Turbine-Blade-Coating Protective Properties.”) The volumetric changes taking place when an alloy with minimum values of all alloying elements (Ni10Co15Cr8AlY) is heated in a temperature range in excess of 600 C are due to c⬘→c transformations (Fig. 4.17). Comparison of volumetric effects in the system under consideration with those of the Ni-Cr-Al system reveals that c⬘-phase starts to dissolve at a lower temperature and that the effect of such transformation is less. It is the evidence of less thermal stability of c⬘-phase when alloyed with cobalt. Increasing cobalt content up to 30% does not affect the phase transformations in progress. In alloys with medium contents of all alloying elements, phase transformations differ from that previously mentioned. The difference consists of the decrease in the total amount of (cⳭc⬘)phases from 65 to 40%, and, ensuing from that, in a lesser volume effect of (c⬘→c) transformation. As heating continues, this transformation fully completes at 750 C. Subsequently, the alloy is in a (bⳭc)-double-phase field. In the absence of cobalt, the same transformation (c⬘→c) takes place at a temperature higher by 250 C, that is, at 1000 C. In alloys with high contents of aluminum, cobalt, and chromium, the volume effects caused by (c↔c⬘) transformation are not great. Typical of these alloys is uniform thermal expansion associated with thermal expansion of b-phase and ␣-Cr solution. Thus, alloying the Ni-Cr-Al system with cobalt reduces thermal stability of c⬘-phase, decreases its quantity, and converts the Ni-Co-CrAl system into the state based on (bⳭc). It is this phase condition that makes coatings so highly ductile when received in the system considered. Co-Ni-Cr-Al Alloys. All alloys and coatings based on the Co-Cr-Al system were thoroughly studied earlier (Ref 11). When used for protec-

tion of nickel-base superalloys, the gradual rise of nickel content in the coating has been noted. Adding nickel to the Co-Cr-Al system can possibly reduce the processes of interaction between the coating and the superalloy. Table 4.13 gives chemical and phase compositions of the Co-NiCr-Al alloys produced. (Note that the main phases in the Co-Ni-Cr-Al and Ni-Co-Cr-Al systems have similar notation.) The main alloy phase constituents include cCo-base solid solution, b(CoAl)-phase, and r(CoCr) phase. No low-temperature modification of cobalt solid solution with hexagonal close-packed lattice was discovered in phase composition. This is due to introducing nickel into the Co-Cr-Al-system alloys, which repays in stabilizing the high-temperature modification of face-centered cubic solid solution. The bphase in the Co-Ni-Cr-Al system is similar to the b-phase in nickel-base systems. It has bodycentered cubic lattice. No c⬘-phase (one of the most important phases in the Ni-Cr-Al system) was discovered in the alloys studied. Microstructure of the alloys is illustrated in Fig. 4.18. Low-aluminum alloys consist of solid solution with a small quantity of b-phase. Increased contents of aluminum and chromium cause the emergence of r-phase in the alloys. Alpha-chromium precipitates in b-phase of alloys containing 12% Al and 30% Cr. Dilatometric investigations were conducted on sample specimens cut out of as-cast blanks. The specimens were annealed at 1000 C for 2 h. (TEC values are given in Chapter 6, “TurbineBlade-Coating Protective Properties.”) With regard to the thermal expansion coefficient, no appreciable alloying effects have been noted in the temperature ranges of 20 to 500 C. In the 500 to 600 C range, the alloys experience magnetic transformation (Fig. 4.19). At higher temperatures, b- and r-phases begin to dissolve. All of the previously mentioned phase transformations result in variations in the specimen coefficients of linear expansion.

Phase Composition and Structure of Overlay Coatings The structure and phase composition of coatings produced by the electron beam (EB)-physical vapor deposition and EA techniques depend on the chemical and phase composition of the alloy evaporated and on a number of process factors. The most important of these factors are:


Fig. 4.16

b-phase contents of (a) as-cast alloys and (b) those annealed at 1200 C vs. element content

Thermal expansion coefficients vs. temperature. (a) 1, Ni10Co15Cr8AlY alloy; 2, Ni30Co15Cr8AlY alloy; 3, Ni10Co30Cr8AlY alloy; 4, Ni30Co30Cr8AlY alloy. (b) 1, Ni10Co15Cr14AlY alloy; 2, Ni30Co15Cr14AlY alloy; 3, Ni10Co30Cr14AlY alloy; 4, Ni30Co30Cr14AlY alloy

Fig. 4.17

Table 4.13

Chemical and phase compositions of Co-Ni-Cr-Al alloys

No.

Alloy

Co

Ni

Cr

Al

Y

Qb, %


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Co10Ni15Cr6AlY Co30Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni15Cr12AlY Co30Ni15Cr12AlY Co10Ni30Cr12AlY Co30Ni30Cr12AlY Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY Co22Cr9AlY

Base Base Base Base Base Base Base Base Base Base base base Base Base Base Base Base Base

10.6 30.8 10.8 30.6 10.6 30.6 10.7 29.7 12.6 31.0 22.4 22.2 17.2 21.2 20.4 20.6 17.8 ...

15.1 14.9 29.8 30.2 15.0 15.1 29.8 29.2 22.5 22.4 14.3 30.0 23.5 22.6 22.6 22.5 22.6 22.6

6.0 5.9 6.0 6.2 11.8 11.8 12.0 11.8 9.3 9.2 9.1 9.2 6.9 11.8 9.3 9.0 9.3 9.2

0.3 0.3 0.3 0.3 0.3 0.3 0.2 0.6 0.7 0.2 0.2 0.1 0.1 0.3 0.3 0.1 0.5 0.3

5 ... ... ... ... 68 ... ... ... 42 ... ... 14 ... ... 39 ... ...

c, b c, b c, b c, b c, b c, b, r c, b, r c, b, r c, b, r c, b, r c, b c, b, r c, b, r c, b, r c, b, r c, b, r c, b, r c, b, r

Chemical composition, wt%


Microstructure of the alloys. (a) Co30Ni15Cr6AlY. (b) Co10Ni30Cr6AlY. (c) Co30Ni15Cr12AlY. (d) Co30Ni30Cr12AlY. (e) Co10Ni22Cr9AlY. (f) Co30Ni22Cr9AlY. (g) Co20Ni15Cr9AlY. (h) Co20Ni30Cr9AlY. (i) Co20Ni22Cr9AlY. (j) Co22Cr9AlY. 500⳯

Fig. 4.18


● ● ●

Temperature of heating (the blades) Stability of evaporation process Conditions of blade revolution in vapor flow; the radius of curvature and roughness of the blade surface ● Peening and heat treatment conditions of coated blades

Thermal expansion coefficients vs.temperature (a) 1, Co10Ni15Cr6AlY alloy; 2, Co30Ni15Cr6AlY alloy; 3, Co10Ni30Cr6AlY alloy; 4, Co30Ni30Cr6AlY alloy. (b) 1, Co10Ni15Cr12AlY alloy; 2, Co30Ni15Cr12AlY alloy; 3, Co10Ni30Cr12AlY alloy

Fig. 4.19

Fig. 4.20

Any instability of the process involved in coating deposition results in coatings whose chemical and phase compositions vary and whose properties are not stable. One of the main parameters controlling the coating structure is the temperature of a substrate on which the vapor is condensed. The diagram in Fig. 3.17 shows the relationship between coating structure and substrate temperature. The diagram is plotted for pure metals that have been condensed on nonrevolving substrates. The most favorable structure of metallic coatings is formed in the zone above T2. In reality, however, heating to the required temperature above T2 can be complicated by a risk of overheating the superalloy structure. Figure 4.20 shows condensate structures that have been deposited by the EB technique on substrates having different temperatures (Ref 12). As a rule, these are fine-grained structures oriented normally to the substrate surface. The orientation is especially pronounced in coatings with highchromium content or in cobalt-base coatings.

Microstructure of condensate from Ni22Co22Cr12AlY alloy. Substrate temperatures of (a) 850 C, (b) 950 C, and (c) 1150 C. 400⳯. Source: Ref 12


The increase in the substrate temperature improves structure stability but reduces its hardness. The structure and phase composition of overlay coatings change notably in the course of peening and heat treatment that follows the condensation process. Microball peening that brings about mechanical hardening of the coating surface and subsequent annealing, causing recrystallization of the coating structure, strikes a balance between structure and phase composition of coating. The orientation of structure with respect to the substrate disappears. Retaining the

Fig. 4.21

initial structural orientation in EB-produced coatings unfavorably affects the coating thermal fatigue. Figure 4.21 shows typical microstructures of overlay coatings obtained from Ni20Cr12AlY and Ni20Co20Cr12AlY alloys, as well as those from Ni8Co12Cr7Al alloy (4.5% Ta, 0.3% Re, 0.4% Hf), after a complete cycle of peening and heat-treating operations. After depostion of the coating from the Ni8Co12Cr7Al alloy by EA technique, the diffusion coating of the Al5%Si1.5%Y alloy was formed on its surface by the technique just described. Although phase

Microstructure of overlay coatings. (a) Ni20Cr12AlY (EB), 200⳯. (b) Ni20Co20Cr12AlY (EB), 300⳯. (c) Ni20Co20Cr12AlY (EA), 300⳯. (d) Ni7Co12Cr17Al (Ni8Co12Cr7Al [EA] Ⳮ Al5% Si1.5%Y, [EA]). 480⳯


compositions of the coating and feedstock alloy principally coincide, there are still some differences caused by instability of the evaporation process. In the original zone of condensation, the phase composition was shifted to cⳭ␣-Cr field, which can be attributed to increased chromiumcontent and lowered aluminum-content at the first stage of feedstock evaporation. The Ni-Cr-Al system is unique because at 1020 C, it has a point of bⳭc↔c⬘Ⳮ␣ transformation. This is why the structure and phase composition of coatings of this system depend

Fig. 4.22

on the annealing temperature and, subsequently, on conditions of work. The results of x-ray structural research are given in Fig. 4.22 and in Table 4.14. The coating obtained by the EB technique from the Ni20Cr12AlY alloy is subject to a heat treatment operation conducted at 1050 C for 2 to 4 h at the final stage of the whole processing cycle. As research has shown, such heat treatment brings about the structure based on NiAl (60 to 65%) phases and c-solid solution. Formation of such structure involves considerable

X-ray pattern of overlay coating surface from Ni20Cr12AlY alloy. 1, initial; 2, annealing 1050 C, 2 h; 3, 1050 C, 4 h; 4, 1100 C 2 h; 5, 900 C, 4 h


Table 4.14 Heat treatment effect on b-phase quantity in Ni20Cr12AlY coating Coating technique applied Electron beam

Electric arc

Annealing condition

Quantity of b-phase, %

After deposition 900 oC, 4 h 1050 oC, 2 h 1050 oC, 4 h 1100 oC, 2 h After deposition 900 oC, 4 h 1050 oC, 2 h 1050 oC, 4 h 1100 oC, 2 h

55 40 60 65 70 (a) 45 50 60 60

(a) Cannot be determined by x-ray analysis

redistribution of aluminum and chromium between the previously mentioned phases. During cooling, as the bⳭc↔c⬘Ⳮ␣ transformation point is passed through, the previously mentioned redistribution does not have enough time to occur, and, as cooling goes on, Ni3Al phase precipitates from c-solid solution in the temperature range of 1000 to 750 C. Thus, a nonequilibrium state bⳭ(cⳭc⬘) is fixed at room temperature. With chromium in the coating exceeding 20%, coherent ␣-Cr precipitates in the NiAl phase. The fixed structure is unstable, and it undergoes changes as blades and vanes are heated to working temperatures. Because coating service life covers thousands of hours, the coating phase composition eventually comes to a state that is equilibrium for a particular temperature. (The changes caused by oxidation and the reactions between coating and alloy are discussed in Chapter 5, “Phase and Structural Changes in Coatings during High-Temperature Tests.”) At temperatures below 1000 C, c⬘Ⳮ␣ or c’Ⳮ␣Ⳮb (for the top limit of aluminum concentration) comes out as an equilibrium phase composition. bⳭc→c⬘Ⳮ␣ phase transformation is accompanied with a decrease of volume and, hence, brings about growing tensile stresses in the coating. In a coating annealed at 950 C for 4 h, the NiAl phase lessened quantitatively to 40%. High-temperature state of bⳭc is notable for high ductility, whereas in c⬘Ⳮ␣ state, the ductility of the coating lowers. It follows, therefore, that high-temperature heat treatments can be used to good advantage to lengthen bⳭc→c⬘Ⳮ␣ transformation in time, thereby retaining the ductility. Such heat treatments augment the sizes of phases that interact and, hence,

decrease the rate of their interaction at temperatures below 1000 C. After being deposited by the EA technique, the Ni-Cr-Al coating is amorphous in terms of x-ray picture; it has no reflections from phase constituents of the coating. This is due to the conditions of condensation on a cold substrate (that is typical of the technique), 500 C, and also owing to fineness of coating structure. After annealing, the phase composition of EA coatings is similar to that of EB coatings (Table 4.14). In the structure of the coatings deposited by EA technique, some banding can be detected. It makes their structure similar to the structure of the coatings deposited from powders by plasma technique. The presence of a droplet phase that behaves similar to melted powder particles is responsible for banding development. The analysis of results from investigating phase composition of alloys and coatings based on Ni-Cr-Al, Ni-Co-Cr-Al, and Co-Ni-Cr-Al systems revealed an important difference between these coatings and their diffusion counterparts. Phase composition of diffusion coatings remains stable when briefly heated in the temperature range of 20 to 1100 C; however, in the same range of temperatures, the alloys used for overlay (condensed) coating deposition and the coatings deposited show diverse phase transformations followed by considerable volume effects. REFERENCES

1. V.S. Sinelnikova, V.A. Podergin, and V.I. Rechin, Aluminides, Kiev, Naukova Dumka, 1965, 240 p 2. G.W. Goward, D.H. Boon, and C.S. Giggens, Formation and Degradation Mechanism of Aluminide Coatings on NickelBase Superalloy, Trans. ASME, Vol 60, 1967, p 228–241 3. A.G. Andreeva, V.A. Konstantinov, Y.A. Tamarin, and V.V. Terekhova, Structure and Composition of Aluminized Layer on Nickel and Its Alloys, Met. Sci. Heat Treat., No. 8, 1967, p 27–30 4. A. Konstantinov, V.V. Terekhova, and Y.A. Tamarin, Nature of Aluminized Layers on Nickel, Met. Prot., No. 2, 1970, p 213–216 5. A. Teylor, X-Ray Metallography, Moscow, Metallurgy, 1970, p 633 6. V.A. Konstantinov, Y.A. Tamarin, and V.V. Terekhova, Structure and Composition Changes in Aluminized Layer on Nickel and Its Alloys, Temperature Resis-


tant Protective Coatings, Moscow, Nauka, 1968, p 151–157 7. A. Taylor and R.W. Floyd, The Constitution of Nickel-Rich Alloys of Nickel-Chromium-Aluminum System, J. Inst. Met., Vol 81, 1952/1953, p 25–32, 451–464 8. S.M. Merchant and M.R. Natis, A Review: Constitution of the Al-Cr-Ni System, Mater. Sci. Eng., Vol 66, 1984, p 47–60 9. Y.M. Doljansky, G.B. Stroganov, and R.E. Shalin, Optimization of Properties of Machine-Building Materials, Moscow, Voenizdat, 1980, 240 p 10. Y.A. Tamarin, V.I. Nikitin, and N.V. Zabrodina, Research into Structure and Physi-

cal-Chemical Characteristics of NiCoCrAl System Alloys, Deposition and Application of Protective Coatings, Leningrad, Nauka, 1987, p 67–78 11. G.W. Goward, Coatings and Coating Processing for Gas Turbine Airfoils Operating in a Marine Environment, Gas Turbine Materials in the Marine Environment Conf., J.W. Fairbanks and I. Machlin, Ed., Columbus, 1975, p 277–296V 12. P.A. Pap, “Research into Structure and Properties and Design of Electron Beam Technique for Deposition of Protective Coatings on Gas Turbine Blades,” Ph.D. dissertation, scientific consultant B.A. Morchan, IES, 1982



CHAPTER 6

Turbine-Blade-Coating Protective Properties THE COATING should protect blade surfaces against damage caused by a gas flow consisting of a mixture of air and fuel combustion products. Different materials based on chromium, aluminides, a number of carbides, and other compounds have high resistance to gas corrosion (oxidation and/or hot corrosion). Extensive studies disclosed that aluminides and the alloys based on them are the best materials for forming protective coatings on nickel superalloys. This is explained by two major characteristics of aluminides: their thermal stability on the surfaces of superalloys strengthened by c⬘-phase, and their high resistance to oxidation. The results of research into high-temperature corrosion of aluminides and the protective coatings based on them are dealt with in numerous publications, such as Ref 1.

Protective-Properties Evaluation Methods Test for Heat Resistance. Superalloy heat resistance (resistance to oxidation) investigation results depend on test methods. Alloy heat resistance is determined by exposing the samples in a furnace in air to a constant or cycling temperature for a certain period. To describe the oxidation process, the following parameters were used: sample specific weight loss qⳮ; and specific weight gain qⳭ; oxidation depth, h, for a certain period, s: qⳮ ⳱

m0 ⳮ mk in g/m2 S

qⳭ ⳱

mox ⳮ m0 in g/m2 S

hox ⳱

qⳮ in lm q

where m0 is sample initial weight, mk is sample weight after corrosion product removal, mox is sample weight with all its oxidation products, S is sample surface area, and q is sample material density. For a majority of the studied aluminide materials, specific weight change, q, on testing for heat resistance shows its parabolic relationship with oxidation time, s, at the specified temperature: qn ⳱ Kp • s

where n is the parabola exponent (for the studied spectrum of the materials n 2) and Kp is the oxidation constant in g2 /cm4 • s. Testing is performed within the time allowed for determining the oxidation general features. Heat resistance test results depend on many factors, such as furnace capacity and air circulation conditions, sample surface roughness, materials and dimensions of the crucibles in which the samples are placed, sample heating and cooling rates, and the number of thermal cycles during testing. The aircraft turbine blades work under heavy thermal-cycling conditions. That is why the evaluation of aluminide alloy and coating protective properties should be made both under isothermal conditions and at the temperature cyclically changing according to the heating-holding-cooling cycle. Cyclic testing reveals oxide cohesive strength on the sample surface. Under cyclic testing conditions, oxidation speeds up and test time reduces. Cyclic heat resistance testing results depend on the cycle temperature


range, DT, and time of holding the sample at the cycle maximum temperature, Tmax. Alloy heat resistance investigation methods are based on the evaluation of oxidation under the condition when the sample diameter is substantially bigger than the oxidation penetration depth. This condition leads to steady-state oxidation when oxidation constant Kp and parabola exponent n do not alter. The approach to research into heat resistance of coated alloys imposes limits on the test procedures and interpretation of the test results. In this case, heat resistance test peculiarities are caused by specific coating thickness and thus, limited amounts of the elements in the coating, which determine the oxidation general features. During heat resistance testing, diffusion, together with oxidation, occurred between the protective coating and superalloy. This results in decreasing the content of the elements that form protective oxides. That is why the general features of coated sample oxidation processes, determined at the initial stage of testing, substantially change in time. Heat resistance testing of the coated superalloys is carried out similar to that for uncoated superalloys. The coated specimen heat resistance was determined by the previously mentioned characteristics, such as coating weight change and oxidation penetration depth. In addition to those characteristics, a period of protective effect of the oxides formed on the coating surface (oxide lifetime) or a period of protective

Diagram of coated bar specific mass change during testing for heat resistance. 1, T1; 2, T2. T1 T2. sq, oxide lifetime

Fig. 6.1

effect of the coating (coating protective lifetime) can be determined. Figure 6.1 shows the specific weight change of the specimens with the coating (qⳭ) in the form of a diagram versus the test time. At the initial stage of oxidation, the specimen specific weight change bears a parabolic relationship to test time. At a certain moment, the coating composition substantially alters. This is due to the consumption of the elements providing protective properties for oxidation and diffusion into the superalloy. The coating composition change results in changing its general oxidation behavior. The transitional period is characterized by the change in composition and structures of the oxides formed on the surface. On completion of this period, sudden acceleration of the oxidation processes occurs. It results in oxide spalling. Oxide spalling from the coating surface is not considered to be a criterion of the coating damage. As testing goes on, new oxide is forming on its surface. It can have protective properties, too. However, oxide spalling is the evidence of a transition to a new stage of oxidation, with a higher rate that ultimately results in coating oxidation through its full thickness. A period of oxide protective effect (oxide lifetime) is determined by time, sq, which, in turn, determines the moment of the loss of protective effect of the oxides formed on the coating surface. For the coating having a specific composition and thickness, sq depends on the test temperature and environment. Within the time interval of up to sq, the coating oxidation behavior is determined by the same characteristics that were used for the evaluation of superalloy heat resistance. Heat resistance can be determined metallographically by measuring the coating oxidation penetration depth for the specified period of time. The time required for oxidation penetration to the depth equal to the coating thickness also determines its lifetime. To speed up testing, the moment of the coating chemical and phase composition change can be determined instead. This change results in deterioration of the protective oxides properties. For overlay coatings, it may be the time until the formation of the nickel-base c-solid solution through the full coating thickness, with the aluminum and chromium content lower than is required for the formation of protective oxides. For the diffusion coatings, this is the time for complete disappearance of the NiAl phase in the coating outer zone.

Turbine-Blade-Coating Protective Properties / 99

The initial coating thickness affects the heat resistance test results. As a rule, coating lifetime is directly proportional to its thickness (Ref 2). To compare the heat resistance of the diffusion coatings with different initial thickness values and aluminum contents, it is convenient to introduce a parameter of “aluminum reserve” in the coating. Tests for Resistance to Hot Corrosion. When studying aircraft gas-turbine engine turbine blades after their long service life, deposits containing sodium, sulfur, vanadium, chlorine, iron, and potassium, which substantially speed up high-temperature oxidation, are detected on the blade surfaces. These elements get into a gas flow in the form of various compounds from the environment as well as from the combustion of fuel that contains small amounts of sulfur, vanadium compounds, and other elements. The mechanism of hot corrosion caused by the presence of Na2SO4, NaCl, V2O5, and other aggressive compounds depends on the thermodynamic characteristics of these compounds as well as temperature, velocity, and pressure of a gas flow. The difference in the temperatures of the gas flow and the blade surface contributes much to corrosion development. The simulations of high-temperature corrosion caused by sulfur-base compounds are covered in numerous works (Ref 3–8). Various methods can be used for hot corrosion testing. Dipping the specimens into crucibles with aggressive media is widely practiced as the simplest method. The crucibles are held at a certain temperature and time in the furnace. Corrosion progress is determined by inspecting the specimen appearance and the corrosion penetration depth. Under such testing conditions, corrosion behavior depends substantially on the depth the specimen is dipped into the aggressive media and, as a rule, to the utmost extent, takes place at the aggressive mixture/air interface. This method lacks quantitative criteria for comparison of protective properties and shows poor agreement of the test results with the data on

Table 6.1 (Ref 9)

Hot corrosion test conditions

Fuel at service

Synthetic ash composition (GZT)

Gas-turbine fuel

66.2% Na2SO4; 1.8% V2O5; 20.4% Fe2O3; 8.3% NiO; 3.3% CaO

Gas medium composition at testing Air, flow rate of 20 L/min

actual coating properties under service conditions. This drawback is rectified by applying special synthetic ash that imitates carbon deposit formed on blade surfaces. The synthetic ash is deposited from aqueous or alcohol solutions onto the surface of the specimens. It evenly covers the specimen surface, and its amount is determined by weighing. In the course of hightemperature testing, the synthetic ash can be applied repeatedly to the specimen at the specified time intervals. The test results are analyzed and presented as a function of the specimen specific weight change and corrosion penetration depth versus time. It is done similarly to that specified previously for oxidation. The test conditions run on gas-flow test benches (gas rigs), which provide the gas-flow parameters similar to the aircraft-engine turbine parameters and most closely approximate the actual operating conditions. Various combinations of sodium salt, vanadium oxides, and more complex systems of compounds are used as aggressive mixtures. The aggressive environment composition should provide for the propagation of corrosion according to the mechanism similar to that existing under blade service conditions. The drawback of this method is its high cost and environmental pollution. Shorter test periods, due to additional aggressive components in the gas flow, can change corrosion mechanism and cause inaccuracy of test results. The proper selection of the testing methods is a crucial contribution to the successful selection of the materials and coatings that feature high resistance to hot corrosion. For this purpose, it is advisable to employ the well-proven lab methods corroborated by the practice of blade testing in engines. To assess the resistance of different materials and coatings to hot corrosion, which is presented subsequently, the test method designed by the authors of Ref 9 was used. Synthetic ash of gasturbine fuel (GZT) was applied to the specimen surfaces (Table 6.1). Depending on the gas-turbine engine or the gas-turbine unit service conditions, the fuel in use, and the analysis of ash deposit formed on the blades, the use of specially selected synthetic ash compositions is required. Gas-turbine fuel composition is chosen on the basis of numerous studies of ash deposits on turbine blades operating under different conditions (Ref 2). The initial components of synthetic ash GZT are thoroughly mixed and milled. Then, an


ethyl-alcohol-base slurry is made of this mixture. The slurry is applied uniformly to the specimen surface in the amount weighed out as 120 g/m2. The initial surface roughness of the specimens should be Ra ⳱ 0.6–1.5 lm. The samples with GZT ash applied to their surfaces are placed in alundum crucibles, which, in turn, are arranged in a closed container with air forced through it at the rate of 20 L/min. In specified intervals, the specimens are taken from the container for inspection and weighing, and, depending on test conditions, specific weight loss or gain (g/m2) of the specimens is determined. During testing, slurry application can be repeated more than once. The tests are carried out at a constant temperature in the range of 700 to 900 C. Test time depends on the corrosive activity of the synthetic ash in use, the gas medium, the protective coating properties, as well as on test temperatures and predicted coated-blade service lives. After every test cycle, specimen surfaces are examined by visual inspection and by weighing, both before and after removal of corrosion products. One specimen from the test batch is randomly taken for metallographic examination. In the case of testing coated specimens, a few techniques for determining their lifetime are available: ●

Corrosion products are not removed from the specimen surface. After every stage of testing, specimens are weighed. A sudden change in kinetics of specimen specific weight gain or loss is considered to be the moment of coating failure. ● Corrosion products are periodically removed from the specimen surfaces, and their mass loss is determined. Sudden specific weight change is evidence of the change in corrosion mechanism and kinetics caused by coating failure. Scale is removed by methods that do not result in damage to the remaining coating. The suitability of the scale removal method is veri-

fied by special check tests of the as-coated specimens. Corrosion products can be removed by their reduction in metallic sodium with ammonia bubbling through it (Ref 9). The degree of corrosion damage is determined from specimen mass loss, qⳮ, and corrosion penetration depth, hk. The parameter chosen to characterize corrosion lifetime of the coating is the time interval from the start of specimen testing to the moment of coating damage, even though damage is in one location on the surface. In all cases, the fact of coating damage should be supported by the data of metallographic examination.

Heat Resistance of Aluminides and Alloys for Overlay Coatings Heat Resistance of Aluminides. NiAl compounds or a combination of NiAl and Ni3Al compounds are an outer-zone base for as-deposited and as-annealed aluminide diffusion coatings on nickel and superalloys. These very compounds determine the protective properties of diffusion coatings at high temperatures. To study heat resistance, the samples of Ni2Al3, NiAl, and Ni3Al aluminides as well as NiAl alloyed with chromium and silicon were cast in vacuum induction furnaces. The compositions of the compounds are given in Table 6.2. Heat resistance testing was carried out under air atmosphere at 1000 C. During testing, specimen weight changes and the structure of oxides formed on the specimens were examined. The test results are shown in Fig. 6.2. Table 6.2 contains the x-ray diffraction analysis data of the oxides after 500 h of testing. Heat resistance of Ni2Al3 and NiAl compounds with different aluminum contents is actually the same. A protective, dense oxide layer with ␣-Al2O3 structure is formed on the surface. Heat resistance of Ni3Al compound on whose

Table 6.2 Analysis of tested aluminides and crystal structure of oxides formed on their surface after testing at 1000 C Aluminide Ni2Al3 NiAl NiAl Ni3Al NiAl NiAl

Chemical composition, wt%

Oxidation time, h

Oxide structure

58% Ni, 41.6% Al 64% Ni, 36% Al 78% Ni, 22% Al 86% Ni, 14% Al 74% Ni, 23% Al, 3% Si 74% Ni, 23% Al, 3% Cr

75 500 500 500 500 500

␣-Al2O3 ␣-Al2O3 ␣-Al2O3 NiO, NiAl2O4, ␣-Al2O3 ␣-Al2O3 ␣-Al2O3


surface a mixture of ␣-Al2O3, NiAl2O4, and NiO oxides was found is a bit lower than that of NiAl compound. Alloying NiAl with silicon has a beneficial effect on heat resistance. During in-testing inspection of the NiAl specimens alloyed with silicon, a denser, less-spalling oxide is seen. However, the electron spectroscopy and x-ray diffraction did not reveal structural differences between the oxides formed on the NiAl specimens with and without silicon. Alloying NiAl with such elements as zirconium, molybdenum, and tungsten, which, due to their low solubility in this compound, form their own phases of ␣-Mo and ␣-W types, results in reducing its heat resistance. In this case, oxidation first penetrates the zones containing ␣-Mo and ␣-W phases. It results in the penetration of oxidation into the aluminide and the damaging of the ␣-Al2O3 protective oxide. Carbides and other compounds of the previously mentioned elements have a similar effect. The mechanism of the NiAl oxidation is selective oxidation of aluminum at the Al2O3/NiAl interface, due to the diffusion of oxygen to this interface through the formed oxide layer. In this case, the kinetics of oxidation is defined by diffusive mobility of oxygen within the Al2O3 lattice. However, no aluminide oxidation penetration depth is observed, because the oxide layer is formed on the NiAl surface and does not penetrate the bulk compound. At a certain stage of oxidation, the consumption of aluminum for forming Al2O3 and simultaneous accumulation

Heat resistance of aluminides at 1000 C. 1, NiAl (23% Al, 3% Si); 2, NiAl (24 to 26% Al); 3, NiAl (23% Al, 3% Cr); 4, Ni3Al (14% Al)

Fig. 6.2

of nickel result in forming Ni3Al compound beneath the oxide layer. In this case, the mechanism of oxidation changes, because the formation of NiAl2O4 spinel and, hereafter, NiO oxide results in oxidation penetration into the aluminide. Heat resistance of Ni-Cr-Al alloys was investigated for the compositions simulating overlay coatings. Alloy compositions are given in Tables 4.10 and 6.3. Testing was carried out isothermally and cyclically at 900 C. In all the cases, sample specific weight gain (qⳭ, g/m2) was measured. During cyclic testing, the temperature varied up and down the range of 20 to 900 C; the holding time at the maximum temperature was 1 h. The total holding time in the vicinity of 900 C was determined graphically by reference to the temperature measurement chart. The longest isothermal test lasted 10,000 h; cyclic testing took 4500 h (4500 cycles). All the Ni-Cr-Al-Y alloys under study have high heat resistance at 900 C (Tables 6.4, 6.5). During 4500 h of cyclic testing, no oxide scales on any of the tested alloys of this system, except for those containing hafnium, were damaged (Table 6.6). This is explained by the formation of ␣-Al2O3•Cr2O3 oxide scale on their surface. Replacement of yttrium with hafnium impairs the protective properties of the oxides formed. Alloying Ni20Cr11Al alloy with 1% Hf results in reducing the service life of the protective oxides to 5000 h under isothermal testing conditions and to 2000 h under cyclic testing conditions. The increase up to 2% Hf content of the alloy reduces oxide protective life at isothermal testing and increases it a little at cyclic testing. Cyclic testing at 1200 C has revealed the unfavorable effect of hafnium and silicon on heat resistance of Ni-Cr-Al alloys (Fig. 6.3). X-ray diffraction of the oxides formed on the surface of the alloys with hafnium revealed that HfO2 was observed to appear in addition to the crystal structure similar to that of ␣-Al2O3•Cr2O3. Heat Resistance of Ni-Co-Cr-Al Alloys. Alloy compositions are given in Tables 4.11 and 6.3. Heat resistance testing of this alloy system was carried out at 900, 1100, and 1200 C. The isothermal and cyclic test results at 900 C are given in Tables 6.7, 6.8, and 6.9. The Ni-Co-Cr-Al-system alloys alloyed with yttrium, silicon, and hafnium have high heat resistance. The Ni20Co20Cr12Al alloy alloyed with 2% Hf and 0.5% Si has an oxidation con-


stant less than that for the same alloy alloyed with yttrium, and the oxide layer life under isothermal and cyclic testing conditions is more than 10,000 and 5000 h, respectively. The same Table 6.3

effect is produced when alloyed with 2% Si and 0.5% Hf. However, the life of the oxide layer on this alloy is reduced a little, to about 5000 h under isothermal testing conditions.

Chemical composition of alloys with hafnium and silicon additions Chemical composition, wt%

Alloy Ni22Cr11Al2Hf Ni22Cr11Al1Hf Ni22Cr11Al1Si Ni22Cr11Al2Si Ni20Co22Cr11AlHfSi Ni20Co22Cr11Al2HfSi Ni20Co22Cr11AlHf2Si Ni20Co22Cr11Al2Hf2Si Ni20Co22Cr11Al1Hf1Si Ni20Co22Cr11AlY2HfSi

Table 6.4

Ni

Co

Cr

Al

Y

Hf

Si

Base Base Base Base Base Base Base Base Base Base

… … … … 20.0 20.0 20.0 20.0 20.0 20.0

22.0 22.0 22.0 22.0 22.5 22.5 22.5 22.5 22.5 22.5

11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0 11.0

… … … … … … … … … 0.3

2.0 1.0 … … 0.5 2.0 0.5 2.0 1.2 1.5

… … 1.0 1.5 0.5 0.5 2.0 2.0 1.2 0.5

Heat resistance of Ni-Cr-Al alloys. Isothermal tests at 900 C Weight gain (qⴐ), g/m2

Alloy

100 h

500 h

1000 h

2500 h

5000 h

10,000 h

1.4 1.6 3.0 2.5 3.4 1.8 2.9

2.1 2.8 3.8 3.5 4.1 3.6 4.6

2.8 3.6 4.6 4.5 4.9 4.5 5.9

3.5 4.3 5.2 4.8 5.4 4.5 6.5

4.9 5.8 6.5 6.4 5.9 3.9 6.9

5.9 5.1 7.3 7.1 7.0 1.6 1.4

Ni22Cr11AlY Ni16Cr9AlY Ni21Cr9AlY Ni19Cr12AlY Ni28Cr11AlY Ni22Cr11Al2Hf Ni22Cr11Al1Hf

Table 6.5 Heat resistance of Ni-Cr-Al alloys. Cyclic tests under 20 to 900 C (1 h), 20 C conditions Weight gain (qⴐ), g/m2 Alloy Ni22Cr11AlY Ni16Cr9AlY Ni21Cr9AlY Ni19Cr12AlY Ni28Cr11AlY Ni22Cr11Al2Hf Ni22Cr11Al1Hf

Table 6.6 900 C Alloy Ni16Cr9AlY Ni18Cr5AlY Ni21Cr9AlY Ni19Cr12AlY Ni28Cr11AlY Ni20Cr11Al1Hf Ni20Cr11Al2Hf

100 h

500 h

1000 h

2500 h

4500 h

4.9 2.6 1.2 3.8 2.8 1.9 4.0

8.2 4.6 2.5 5.3 4.5 3.0 8.0

9.5 4.7 3.4 5.5 5.5 3.5 10.0

12.0 6.2 4.2 6.8 6.5 5.2 18.0

14.0 6.0 4.7 6.9 6.5 3.3 59.0

Life of Ni-Cr-Al alloys oxides at Oxide life, h Kp • 1014, Isothermal tests Cyclic tests g2/cm4 • s 10,000 10,000 10,000 10,000 10,000 5000 2000

5000 5000 5000 5000 5000 2000 5000

3.6 5.6 5.8 2.2 6.6 9.6 5.6

In all cases, indicates that lifetime is longer than sample test time.

Heat resistance of alloys during cyclic testing in the temperature range of 20 ⇔ 1200 C (1 h). 1, Ni22Cr11AlY; 2, Ni22Cr11Al1Hf; 3, Ni22Cr11Al2Hf; 4, Ni22Cr11Al1Si; 5, Ni22Cr11Al2Si

Fig. 6.3


Table 6.7

Heat resistance of Ni-Co-Cr-Al alloys. Isothermal tests at 900 C Weight gain (qⴐ), g/m2

Alloy Ni20Co22Cr11AlY Ni20Co22Cr11AlHfSi Ni20Co22Cr11Al2HfSi Ni20Co22Cr11AlHf2Si Ni20Co22Cr11AlY2HfSi

Table 6.8

100 h

500 h

1000 h

5000 h

10,000 h

2.9 1.0 1.0 2.3 1.9

4.4 1.9 2.3 3.2 3.9

6.3 2.8 3.0 4.3 5.6

17.0 4.4 7.2 4.8 11.0

19.0 4.8 9.6 2.9 15.0

Heat resistance of Ni-Co-Cr-Al alloys. Tests under 20 to 900 C (1 h), 20 C conditions Weight gain (qⴐ), g/m2


100 h

500 h

1000 h

2500 h

4500 h

4.1 1.1 3.4 2.5 3.2

5.8 2.4 5.3 3.6 5.6

6.3 2.7 5.8 4.4 7.0

8.4 6.5 6.4 5.2 11.0

9.4 35.0 6.1 5.7 15.0

Comparison of oxidation constants of Ni20Cr11AlY and Ni20Co20CrAlY alloys shows that cobalt reduces heat resistance and speeds up the oxide-layer formation and growth. At the test temperatures of 1100 and 1200 C, heat resistance of the alloys was determined from specimen weight change, from the formed oxide layer structures, and from the phase composition and structure of the subscale zones. The results of the measurements of the specimen specific weight change during testing are presented in Fig. 6.4, 6.5, and 6.6. As a result of statistics-based processing of specific weight measurements of the Ni-Co-CrAl alloys after testing at 1200 C for 100 h, the regression equations were derived. They relate specific weight gain, qⳭ, with the cobalt, chromium, aluminum, and yttrium contents of the alloys. Common for the whole spectrum of the investigated compounds is the fact that they have high heat resistance at 1100 and 1200 C, which hardly varies with the addition of alloying elements. This fact is explained by the similar Table 6.9 at 900 C

Life of Ni-Co-Cr-Al alloys oxides Oxide life, h


Isothermal tests

Cyclic tests

Kp • 1014, g2/cm4 • s

10,000 10,000 10,000 5000 10,000

5000 2000 5000 5000 5000

11.0 2.1 2.5 5.1 1.3

In all cases, indicates that lifetime is longer than sample test time.

crystal structure of the oxides, which, in all the alloys, consists of ␣-Al2O3•Cr2O3 and (NiCo) (CrAl)2O4 spinel. In the alloys with high cobalt content, the spinel proportion increases. If the cobalt content is about 30%, spinel, by its lattice parameters, approaches CoAl2O4. In all the alloys alloyed with hafnium, the formation of HfO2 is observed. The oxide weight gain is observed in the alloys of a higher yttrium content, due to their better adherence to the surface as well as to the extended zone of internal oxidation. More detailed information on the effect of alloying elements was obtained from the analysis of b-(NiAl)-depleted outer-zone dimensions, which are a measure of aluminum consumption for protective oxide formation (Fig. 6.7). The regression equation relating the depth of bphase-depleted subscale in NiCoCrAlY alloys with their alloying is presented in the Appendix. Aluminum has the main effect on the dimensions of the zone free from b-phase. This is explained by different aluminum contents of the alloy outer zones. The calculated amounts of aluminum in the 100 lm-thick outer zone of the alloys, which contain 8, 11, and 14% Al, are 8, 30, and 55 g/m2, respectively. During calculations, it was assumed that the lower limit of the aluminum content of the alloy that contains b-phase is about 7%. The aluminum consumption rates for the formation of oxides for all the alloys during testing are rather similar and are 10 to 15 g/m2. Thus, all the aluminum accumulated in the b-phase of the alloys with 8% Al is consumed for the formation of the oxides, while approximately only 1⁄4 is consumed in the


Fig. 6.4

Heat resistance of alloys under isothermal testing conditions at 1100 C. 1, Ni10Co15Cr8Al; 2, Ni30Co15Cr8Al; 3, Ni10Co30Cr8Al; 4, Ni10Co15Cr8AlY; 5, Ni10Co15Cr14Al; 6, Ni30Co15Cr14Al

Heat resistance of alloys under isothermal testing conditions at 1200 C. (a) 1, Ni10Co15Cr8Al; 2, Ni30Co15Cr8Al; 3, Ni10Co30Cr8Al; 4, Ni30Co30Cr8Al; 5, Ni10Co15Cr8AlY and (b) 5, Ni10Co15Cr14Al; 6, Ni30Co15Cr14Al; 7, Ni10Co30Cr14Al; 8, Ni30Co30Cr14Al; 13, Ni10Co15Cr14AlY

Fig. 6.5

alloy with 14% Al. From the obtained results, it follows that when comparing heat resistance versus b-phase-free zone dimensions, the alloys of high-aluminum contents have much higher resistance than the alloys of lower-aluminum content. Addition of chromium to the studied alloys results in decreasing the dimensions of the zone depleted of b-phase. However, chromium has a different effect on the alloys with varying aluminum contents. With 8% Al, and especially with the lower-cobalt content, more chromium reduces the zone depleted with b-phase and increases heat resistance, respectively. This effect of chromium at 1200 C is explained by the fact that chromium is active in oxidation, accompanied by the partial substitution of chromium in Al2O3•Cr2O3 oxide. With the high-aluminum content, the effect of chromium on oxidation is insignificant. The effect of cobalt is most conspicuous in the alloys with low-aluminum content. Increasing the cobalt content of the 8% Al alloy to 20 to 25% results in a slowdown of the oxidation processes. The zone depleted of b-phase diminishes from 150 lm at 10% Co to 90 lm at 20% Co. Further increase in the cobalt content results in the gradual progress of oxidation processes and a larger b-phase-depleted zone. The effect of cobalt is associated with the changes in the protective properties of (NiCo)(CrAl)2O4 spinel, whose amount increases as the cobalt content


Heat resistance of the alloys under isothermal testing conditions at 1200 C. (a) 1, Ni20Co15Cr11AlY; 2, Ni10Co22Cr11AlY; 3, Ni30Co22Cr11AlY; 4, Ni20Co30Cr11AlY and (b) 1, Ni20Co22Cr14AlY; 2, Ni20Co22Cr11AlY; 3, Ni20Co22Cr8AlY; 4, Ni22Co22Cr11AlY

Fig. 6.6

increases. At a certain stage, it transforms into cobalt-base spinel of Co(CrAl)2O4, which has worse protective properties. With the high-aluminum content, the effect of cobalt is insignificant. The effect of yttrium on the formation of the b-phase-free zone is insignificant. Cyclic heat resistance testing in the temperature range of 20 ↔ 1200 C, with the holding time of 1 h at the cycle maximum temperature 1200 C, confirmed the relationships generally observed during the other types of heat resistance testing (Fig. 6.8). Hafnium and silicon in the specified amounts (1.5% Hf, 0.5% Si) increase cyclic heat resistance, especially if they are used together with yttrium. Complete sub-

stitution of yttrium with hafnium and silicon in the Ni-Co-Cr-Al system results in lower cyclic heat resistance. Heat Resistance of Co-Ni-Cr-Al Alloys. Alloy compositions are given in Table 4.13. The results of research into heat resistance of Co-NiCr-Al-Y-system alloys at 900 C are presented in Tables 6.10 and 6.11. The regression equations derived from statistics-based processing of

Heat resistance of the alloys during cyclic testing under condition of 20 ⇔ 1200 C cycling (1 h). 1, Ni20Co22Cr11AlY; 2, Ni20Co22Cr11Al2Hf2Si; 3, Ni20Co22Cr11AlY2HfSi; 4, Ni20Co22Cr11Al2HfSi; 5, Ni20Co22Cr11Al1Hf1Si; 6, Ni20Co22Cr11AlHf2Si; 7, Ni20Co22Cr11AlHfSi

Fig. 6.8 Effect of alloying on thickness of zone free from nickel-aluminum phase in Ni-Co-Cr-Al alloys. Tested at 1200 C for 200 h

Fig. 6.7


Table 6.10 ditions

Heat resistance of Co-Ni-Cr-Al alloys. Cyclic tests under 20 to 900 C (1 h), 20 C conWeight gain (qⴐ), g/m2

Alloy

100 h

500 h

1000 h

2500 h

5000 h

10,000 h

Co10Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni30Cr12AlY Co10Ni22Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY

12 1.1 1.2 4.3 5.4 1.5 2.6 2.3 1.9 5.0

13 1.9 2.9 4.9 6.5 2.9 2.7 2.8 2.4 7.0

16 4 3.9 7.3 7.7 3.1 3.7 3.7 3.2 8.5

19 5.4 5.8 9 9.4 4.5 5.3 4.8 4.7 9.5

22 5.4 7.4 8.7 10 5.3 6.4 4.9 5.3 12

ⳮ29 6.6 11 7.4 11 4.9 5.4 5.6 4.9 13

the test results are given in the Appendix. The total test time was 8000 h. All the alloys studied have high heat resistance. During testing, the oxide layers on the alloys were not damaged, except for the alloy of low-aluminum and -chromium contents. Increasing nickel, aluminum, and yttrium contents of this system results in a slight rise in the alloy oxidation rates. The oxidation rates for alloys with the maximum content of the previously mentioned elements increase 1.5 to 2 times. Chromium exerts an opposite effect. Increasing its content from 10 to 30% decreases the oxidation rate during cyclic heat resistance testing 5 times. The favorable effect occurs in the chromium content range of 15 to 22.5%. These general features of the alloying elements do not depend on the test time. Despite alloying, the oxides of all the tested alloys are highly stable.

Heat Resistance of Coated Superalloys During heat resistance testing and other hightemperature tests, the following reactions beTable 6.11 Life of Co-Ni-Cr-Al alloys oxides. Cyclic testing under 20 to 900 C (1 h), 20 C conditions Alloy Co10Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni30Cr12AlY Co10Ni22Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY JS6K

Oxide life, h

Kp • 1014(a), g2/cm4 • s

7000 8000 8000 8000 8000 8000 8000 8000 8000 8000 2000

71 4.4 4.2 15 16 2.7 3.8 3.8 2.8 20 …

(a) Designed for 1000 h of oxidation

tween the oxidizing environment, coating, and superalloy take place: ●

Formation of a protective oxide on the coating surface ● Diffusion of aluminum, chromium, and other components of the coating into the oxide scale ● Diffusion of aluminum and elements added for alloying the coating into the superalloy ● Diffusion of the alloy components into the coating These processes result in the decrease in aluminum and chromium coating contents down to the level when protective oxides no longer form on the coating surface. Oxidation processes speed up suddenly, and the protective effect of the coating ends. For full and proper estimation of coating protective properties during testing for heat resistance, research into each of the previously mentioned processes is necessary. Heat Resistance of Superalloys with Diffusion Coatings. The analysis of the results of heat resistance testing of the superalloy samples (e.g., Ni20CrTi, Ni10CrMoWCo, JS6K, JS6U, and JS30) with various diffusion aluminide coatings shows that the distinctions in the kinetics of their specific weight change (qⳭ, g/m2) are insignificant. This is explained by the fact that the oxides with the same or similar properties are formed on the surface of the diffusion coatings. X-ray diffraction of the diffusion coatings after 100 and 1000 h of testing at 950 C allowed the detection of the oxides with ␣-Al2O3 structure on their surfaces. The same crystalline structure and protective properties of the oxides that are formed on the surface of diffusion aluminide coatings deposited on different nickel alloys are the cause of their similar heat resistance. During testing, the aluminide-coating oxidation


rate is considerably reduced; if, in the first 2 h of holding at 950 C, it is approximately 0.3 g/m2 • h, on a lapse of 100 h, it is about 0.025 g/m2 • h. The sample specific weight change after 100 h of testing corresponds to the formation of an Al2O3 oxide layer 2 to 4 lm thick. The increase in protective oxide layer thickness at testing reduces layer adherence to the coating surface and causes cracking and spalling. The oxide layer defects are rectified by aluminum diffusing from the coating. Aluminum depletion of the coating outer zone at testing for heat resistance results in altering the protective oxide structure and speeding up oxidation. However, as long as aluminides remain in the coating outer zone, including Ni3Al, the oxides provide protection. They protect the coating and alloy against accelerated oxidation. Heat resistance testing of nickel superalloys with aluminide diffusion coatings was carried out by determining the sample weights during isothermal oxidation in the range of 900 to 1000 C. The diffusion coating was deposited on the samples by aluminizing in the mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl; initially it was 40 to 50 lm thick, with 34 to 36% Al

Fig. 6.9

Parametric diagram of heat resistance of aluminized coatings on superalloys at 950 C

content of the outer zone. During the analysis of the test results for a period of more than 1000 h, the alteration of the oxidation mechanism related to the formation of NiAl2O4 spinel on the coating and its partial spalling was observed. The parametric heat resistance diagram shown in Fig. 6.9 presents numerous results of 1000 h testing of the superalloys with diffusion coatings. Plotting such diagrams is proposed elsewhere (Ref 10). A parametric diagram allows integration of test results measured at different temperatures for different periods: lgq ⳱

1 (lg k0 ⳮ p) n

where q is specific weight gain (g/m2), n and k0 are coefficients depending on materials properties and test medium, and p is the heat resistance parameter. The heat resistance parameter is calculated from the equation: p ⳱

0.218 • Q ⳮ lg t T

where Q is oxidation activation energy (in the temperature range of 900 to 1000 C; Q ⳱ 267,000 cal/mol); T is test temperature, in K; and t is test time, in hours. All the tested superalloys with aluminide diffusion coatings have high heat resistance practically independent of the alloy composition. The test temperature rise to 1100 C results in decreasing heat resistance of the diffusion coating. However, even at this temperature, the coating provides protection against oxidation for 200 to 400 h for the alloys of the JS-type (e.g., JS6K, JS6U, and JS30). The heat resistance test results for the JS6U superalloy with different versions of diffusion coatings deposited by the electric arc (EA) method are presented in Fig. 6.10 and in Tables 6.12 and 6.13. The oxidation constant of the coatings alloyed with silicon and yttrium is approximately onethird of that of the diffusion coating formed in the powder mixture with the composition of 98% Al-Fe alloy (50% Al) and 2% NH4Cl. If the aluminum alloy that contains more than 10% Si is used for saturation, oxide-layer spallation occurs. High-silicon content reduces layer lifetime to 2000 to 3000 h, while the life of the layers on the coatings with less silicon content exceeds 10,000 h.


Alloying the aluminide coatings with silicon and yttrium increases their heat resistance. High heat resistance of aluminide coatings alloyed with silicon is explained by the fact that silicon

Fig. 6.10

partially substitutes ions of aluminum and promotes a rise in the ␣-Al2O3 oxide density in it. Heat Resistance of Superalloys with Overlay Coatings. The characteristics of the coatings

Heat resistance of JS6U alloy with diffusion coatings at 900 C. (a) Isothermal test conditions, and (b) cyclic test conditions, 20 ⇔ 900 C (1 h). 1, Aluminized coating; 2, Al5Si1.5Y electric arc coating; 3, Al10Si0.7Y electric arc coating

Table 6.12 Oxidation constants of JS6U superalloy with diffusion aluminized coatings. Isothermal tests Oxidation constant, Kp • 1014, g2/cm4 • s Aluminizing technique

Alloy

900 C

950 C

1000 C

Aluminizing in mixture of Al-Fe alloy (50% Al), 2% NH4Cl EA (Al5Si1.5Y) EA (Al10Si0.7Y) EA (Al15Si0.5Y)

Ni10CrMoWCo JS6K JS6U JS6U JS6U JS6U

6.7 7.2 6.9 2.3 4.0 3.0

10.0 12.0 … … … …

17.0 20.0 … … … …

Table 6.13 Life of oxides on coatings on JS6U superalloy. Cyclic testing under 20 to 900 C, 20 C conditions Coating Co20Cr11AlY Co20Cr11AlY/ZrO2-15%Y2O3 Ni20Co20Cr11AlY/ZrO215%Y2O3 Fe12Ni20Cr12Al1SiY Al AlPd Al10Si0.7Y Al15Si0.5Y Uncoated JS6U

Oxide life, h

Kp • 1014(a), g2/cm4 • s

4000 4000(b) 4000(b)

4.7 1.7 0.71

4000 1000 6000 4000 2500 2000

10.0 5.1 1.4 4.0 3.0 22.0

(a) Designed for 1000 h of oxidation. (b) Ceramic layer lifetime


studied are presented in Table 6.14. During testing in the temperature range up to 1000 C, an oxide with the ␣-Al2O3•Cr2O3 structure is formed on the surface of the superalloys with overlay coatings in the same manner as on the alloys with diffusion coatings. This explains why the specific weight change of the superalloy samples with overlay coatings is similar to the specific weight change of the superalloy samples with diffusion coatings (Table 6.13 and Fig. 6.11). The use of the iron-base coatings for protection of nickel-base superalloys cannot provide the desired heat resistance. The life of the protective oxide layers on the Fe12Ni20Cr12Al1SiY coating at 900 C does not exceed 4000 h (4000 cycles). It is caused by instability of the coating composition during testing. The formation of a thin ceramic layer (15 lm thick) on the coating surface by the electron beam (EB) method allows the improvement of coating heat resistance and retardation of its oxidation. After 400 h of cyclic testing, the ZrO215% Y2O3 ceramic spalls. However, the metallic coating still preserves its protective properties. At temperatures lower than 1000 C, interaction between nickel-base coatings and superalloys may be neglected. A temperature rise to 1100 C increases aluminum, chromium, and co-

Table 6.14 JS6U alloy

Characteristics of coatings on

Coating Co20Cr11AlY Ni20Co20Cr11AlY Ni23Co28Cr9AlY Ni28Cr6AlY Fe12Ni20Cr12Al1.5SiY Ni20Cr12AlY/Cr carbide Co20Cr11AlY/ZrO215%Y2O3(a) Ni20Co20Cr11AlY/ZrO215%Y2O3(a) Al(b) AlPd(b) Al10Si0.7Y(b) Al15Si0.5Y(b) Ni8Co12Cr7Al(c) Ni7Co12Cr17Al(c,d)

Deposition method

Thickness of layers, lm

EB EB EB EB EB EB EB

100 Ⳳ 20 120 Ⳳ 20 110 Ⳳ 20 120 Ⳳ 20 180 Ⳳ 20 50 Ⳳ 10(a) 100 Ⳳ 20

EB

100 Ⳳ 20

Aluminizing Aluminizing EA EA EA EA

50 60 50 60 80 80

Ⳳ Ⳳ Ⳳ Ⳳ Ⳳ Ⳳ

5 5 5 5 5 5

EB, electron beam; EA, electron arc. (a) Ceramic coating is 15 lm thick. (b) Diffusion coatings. Al, aluminizing in powder mixture of 98% Al-Fe (40% Al) alloy, 2% NH4Cl; AlPd, palladium layer (10 lm), aluminizing in powder mixture of 98% Al-Fe (40% Al) alloy, 2% NH4Cl. Chromium carbide sublayer 10 lm thick. (c) Coating additionally contains 4.5% Ta, 0.3% Re, and 0.4% Hf. (d) Two-layer (overlay Ⳮ diffusion) coating (Fig. 4.21d)

balt diffusive mobility. The amounts of these elements diffusing from the coating to the interaction zone become comparable with their amounts spent on oxidation. During heat resistance testing of the JS6U alloy with the Ni20Cr12AlY coating, which is about 120 lm thick at 1100 C, the composition and structure of oxide layers gradually change. This change is caused by oxidation and diffusion of chromium and aluminum into the alloy. At the first oxidation stage, ␣-Al2O3•Cr2O3 is the base of the oxide layer (Fig. 6.12). After a lapse of 50 h, Ni(Al,Cr)2O4 spinel appears in the oxide layer. The amount of the spinel increases gradually in the course of testing. The life of Ni20Cr12AlY and Ni20Co20Cr12AlY coatings about 100 lm thick at 1100 C is 600 to 800 h. Long testing is required to reveal the differences in heat resistance of the coatings. Testing duration should exceed coating lifetime, sq, shown in Fig. 6.1. At temperatures less than 1000 C, the test time required to assess coating lifetime is several thousand hours. Thus, shorttime heat resistance testing does not always permit one to define the coating protective properties. Under certain testing conditions, the similar structures of the oxides formed and the kinetics of their growth do not allow revealing differences in the coating protective properties, even in the case of rather long test time. To determine and compare the protective properties of the coatings on nickel-base superalloys, a combination of different criteria should be used. These may be the data on inspection of the structures of the oxides, which are formed on the surface and phase composition variations during testing. The presence of spinels, the reduction of the aluminum content to 12%, and the disappearance of NiAl phase from the outer zone indicate that the diffusion coating protective properties are exhausted. For the overlay coatings, the absence of NiAl phase does not mean that the protective properties are exhausted. The coating ceases to be protective when the aluminum content reduces to 4 to 6% and the chromium content falls to 14 to 16%.

Resistance of Alloys for Overlay Coatings to Hot Corrosion The alloy resistance to hot corrosion was studied on the samples tested in the synthetic ash of GZT (120 g/m2) at temperatures of 600, 750,


Fig. 6.11

Heat resistance of JS6U alloy with Ni20Cr12AlY overlay coating (100 lm thick) under isothermal test conditions

850, and 900 C for 200 h using the previously mentioned technique. The corrosion rate was determined by measuring the loss of the sample specific weight (qⳮ, g/m2) and the uniform corrosion depth (hhc, lm). Hot Corrosion of Ni-Cr-Al Alloys. The results of testing Ni-Cr-Al alloys are presented in Table 6.15 and Fig. 6.13. The corrosion damage rate depends on the test temperature and the chemical and phase compositions of the alloys. From the results of testing in GZT ash, all the Ni-Cr-Al alloys can be conventionally divided into three groups: high, moderate, and low re-

Fig. 6.12

Kinetics of oxide structural change in overlay coatings on JS6U alloy at 1100 C

Table 6.15 Hot corrosion of Ni-Cr-Al alloys. Test time in GZT ash, 200 h Weight loss (qⴑ), g/m2 Alloy Ni22Cr11AlY Ni16Cr9AlY Ni21Cr9AlY Ni18Cr16AlY Ni19Cr12AlY Ni28Cr11AlY Ni8Cr18AlY Ni7Cr12AlY Ni22Cr11Al2Hf Ni22Cr11Al1Hf Ni22Cr11Al1Si Ni22Cr11AlSi

600 C

750 C

850 C

900 C

2.3 0.6 … … 3.8 1.9 … … 2.7 2.1 3.6 3.6

13 14 … 52 23 32 53 1100 31 31 24 26

1500 6400 2000 280 860 150 1500 1600 130 910 910 1200

260 4900 … … 180 140 … … 100 2800 180 1300

Uniform corrosion penetration at testing in GZT ash for 200 h. 1, Ni18Cr5AlY; 2, Ni28Cr11AlY; 3, Ni18Cr16AlY; 4, Ni19Cr12AlY; 5, Ni22Cr11AlY; 6, Ni8Cr18AlY; 7, Ni7Cr12AlY; 8, Ni16Cr9AlY

Fig. 6.13


sistance to corrosion. The specified groups have the following depths of corrosion attack: less than 50 lm, 50 to 200 lm, and more than 200 lm, respectively. In the temperature range of 600 to 750 C, all the studied alloys, except for Ni7Cr12AlY, have high resistance to hot corrosion (Ref 11). The depth of corrosion attack does not exceed approximately 7 lm. The maximum depth of corrosion attack against the alloys was observed at 850 C. Classification of the alloys with high, moderate, and low resistance to corrosion at this temperature is based on their chemical and phase compositions. The analysis of the corrosion test results has revealed that, in the system in question, chromium has the maximum effect on hot corrosion. The effect of chromium depends on its amount and its distribution between certain phases. This, in turn, is closely related with the aluminum content of the alloy. The effect of aluminum on resistance to hot corrosion is exhibited mainly through the effect on the phase composition of the alloys. The aluminum content governs the presence of heterogeneous or homogeneous structures based on the phases having various chromium solubility. At the temperature of 850 C, chromium solubility in the main phases of the Ni-Cr-Al-system alloys is about 30 to 35% in c-solid solution, 10 to 15% in Ni3Al, and 10 to 13% in NiAl. A c-solid-solution-based Ni20Cr6AlY alloy that is homogeneous at 850 C and with a chro-

Uniform corrosion penetration at testing in GZT ash for 200 h. 1, Ni22Cr11Al2Hf; 2, Ni22Cr11Al1Si; 3, Ni22Cr11AlSi; 4, Ni22Cr11Al1Hf

Fig. 6.14

mium-aluminum content ratio of 3.3 has the maximum resistance to hot corrosion. The alloys (e.g., Ni7Cr12AlY, Ni3Al-base Ni16Cr9AlY) were observed to have the minimal corrosion resistance. The NiAl-base alloys (e.g., Ni8Cr18AlY) with the same chromium content as Ni3Al-base alloys demonstrate higher resistance to hot corrosion. The Ni28Cr11AlY alloy with 28% Cr and a chromium-aluminum content ratio of 2.7 has high resistance to hot corrosion. Apart from phase heterogeneity of this alloy, the previously mentioned chromium content provides the alloy with high resistance to hot corrosion within the entire temperature range of 600 to 900 C. The effect of hafnium on hot corrosion resistance of NiCrAl alloys depends on its amount (Table 6.15, Fig. 6.14). The addition of 1% Hf does not change corrosion resistance in the temperature range of 600 to 850 C. At the same time, an expansion of the zone that is attacked by corrosion is observed at 900 C. This expansion is caused by the shorter life of the formed oxide layers that are noticed at heat resistance testing of this alloy. Alloying the Ni22Cr11Al alloy with 2% Hf transfers it to the group of alloys with high resistance to corrosion in the entire temperature range of 600 to 900 C. The depth of uniform corrosion attack for this alloy at 850 C does not exceed the depth of corrosion attack for the alloy containing 28% Cr. The effect of 0.5 to 1% Si on resistance of the Ni22Cr11Al alloy to hot corrosion is insignificant. At 850 C, the addition of silicon to the alloy reduces the corrosion penetration depth approximately 1.5 times compared with the Ni22Cr11AlY alloy. Hot Corrosion of Ni-Co-Cr-Al Alloys. Testing of Ni-Co-Cr-Al alloys for resistance to hot corrosion was carried out at 850 and 900 C, that is, under maximum corrosion-temperature conditions. The statistics-based processing of the results of the test run in GZT ash at 850 and 900 C allowed for deriving the regression equations relating alloy addition contents to the sample specific weight change. (The equations are given in the Appendix.) Similar to the Ni-Cr-Al system, the effect of alloying elements on the resistance to hot corrosion is combined from two factors: the effect on the phase composition of the alloys and, thus, on their homogeneity and distribution of the elements between phases; and the direct effect of the alloying element on corrosion characteristics (Table 6.16, Fig. 6.15) (Ref 12, 13).


Chromium increases resistance of the lowaluminum content NiCoCrAlY alloys to hot corrosion. In this case, it is uniformly distributed mainly in c-solid solution. In the alloys of medium-aluminum content (10 to 12%), the increase in chromium content does not result in better resistance to corrosion. It is due to phase heterogeneity of the structure that contains a

Table 6.16 Hot corrosion testing of Ni-CoCr-Al-Y alloys. Test time in GZT ash, 200 h Weight loss (qⴑ), g/m2 Alloy Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni30Co30Cr14Al Ni30Co15Cr8AlY Ni10Co30Cr8AlY Ni30Co30Cr8AlY Ni10Co15Cr14AlY Ni30Co15Cr14AlY Ni10Co30Cr14AlY Ni30Co30Cr14AlY Ni10Co22Cr11AlY Ni30Co22Cr11AlY Ni20Co15Cr11AlY Ni20Co30Cr11AlY Ni20Co22Cr8AlY Ni20Co22Cr14AlY Ni20Co22Cr11Al Ni20Co22Cr11AlY Ni20Co22Cr11AlY

Fig. 6.15 C for 200 h

850 C

900 C

460 210 110 130 63 87 230 280 150 130 140 71 56 170 81 430 72 66 170 9 140 110 120 120

46 1300 68 86 58 42 74 … … … … … … … … 42 33 3800 40 85 120 110 120 100

Effect of alloying on specific mass loss (qⳮ, g/m2) of Ni-Co-Cr-Al alloys. Tested in GZT ash at 900

considerable amount of Ni3Al and NiAl-like phases featuring limited solubility of chromium. With high-aluminum content (13 to 14%), the increase in chromium content to more than 15 to 20% reduces alloy resistance to hot corrosion because of the formation of a great number of ␣-Cr-like phases and r-phase. The effect of aluminum on the resistance of the NiCoCrAlY alloys to hot corrosion is similar to that described for the Ni-Cr-Al system. It exhibits itself in the change of the phase composition, the alloy structure, and the distribution of alloying elements between phases. When phases, which differ greatly by their composition, are present in the alloy structure, phase heterogeneity reduces the resistance to hot corrosion. There are two types of compositions that have maximum resistance to corrosion. They feature high- and low-aluminum contents. Cobalt increases resistance to hot corrosion. Unlike chromium, cobalt is uniformly distributed between the phases of the alloys and, due to its positive effect, exhibits itself in the zones with both low- and high-aluminum contents. Cobalt reduces the amount of Ni3Al phase, thus promoting more uniform distribution of chromium within the structure of the alloys. The effect of cobalt is extremely evident in the alloys with 18 to 28% Co content. Further increase in cobalt content promotes the formation of rphase, which bonds chromium with cobalt, creates structural heterogeneity, and reduces alloy resistance to hot corrosion. Yttrium in the content range of 0.04 to 0.08% had no effect on the alloys tested for resistance to hot corrosion. The effect of hafnium and silicon on the resistance of the NiCoCrAlY alloys to hot corrosion is much weaker than for the NiCrAl system (Table 6.17). This is due to the high resistance of this system alloy to corrosion. At the same time, alloying the alloys with hafnium and silicon causes the depth of corrosion attack to be reduced a little. Hot Corrosion of Co-Ni-Cr-Al Alloys. The test results for hot corrosion resistance of the Co-Ni-Cr-Al-system alloys are given in Table 6.18. The derived regression equations relating sample specific weight loss to alloying are given in the Appendix. The chromium and aluminum contents and their ratio have a crucial effect on this system as well as on the previous one. To analyze the effect of alloying on resistance to hot corrosion, the alloys were divided into three groups with different ratios of chromium and aluminum con-


Table 6.17

Hot corrosion testing of Ni-Co-Cr-Al alloys. Test time in GZT ash, 200 h Weight loss (qⴑ), g/m2

Alloy Ni20Co22Cr11AlY Ni20Co22Cr11AlHfSi Ni20Co22Cr11Al2HfSi Ni20Co22Cr11AlHf2Si Ni20Co22Cr11Al2Hf2Si Ni20Co22Cr11Al1Hf1Si Ni20Co22Cr11AlY2HfSi Ni6Co22Cr11AlY

600 C

700 C

750 C

800 C

850 C

900 C

4.3 2.1 2.0 0.8 … … … …

… … … … … … … 1.0

22.0 16.0 3.5 2.5 12.0 … … …

… … … … … … … 7.3

110 120 130 96 30 130 98 63

79 110 83 86 … … … 796

tents, CCr /CAl. The first group was characterized by 3.4 to 4.9, the second by 2.0 to 2.5, and the third by 1.3 to 2. Such a classification is explained by different compositions of the oxides formed on the surface. If the first group is the case, Cr2O3 oxides are formed mainly on the surface of the alloys; if the third group is the case, Al2O3 oxides are formed; and if the second group is the case, mixed oxides are formed. Alloying with nickel is unfavorable for hot corrosion resistance of the alloys, independent of their chromium and aluminum contents. The alloys with a CCr /CAl ratio of 3.4 to 4.9 are less impaired by nickel than the others. Higher-chromium content results in better resistance to hot corrosion. The effect of chromium is the strongest if the nickel and aluminum contents are high. In all the alloys studied, aluminum content increase leads to a higher hot corrosion rate. This Table 6.18 Hot corrosion testing of Co-NiCr-Al-Y-system alloys. Test time in GZT ash, 200 h Weight loss (qⴑ), g/m2 Alloy

700 C

800 C

850 C

900 C

0.34 0.55 1.5 1.3 1.3 2.7 0.4 0.29 1.0 1.0 0.6 1.1 0.81 0.55 0.82 1.2 0.53 1.3 … …

4.7 5.3 6.0 25 9.2 11 7.5 3.2 9.1 14 3.1 19 24 2.5 2.0 22 3.1 3.1 47(a) 5.7(a)

32 15 41 33 21 17 25 5.3 26 55 15 50 33 41 31 30 23 27 45 18

72 27 29 35 333 86 28 7.7 36 149 278 57 170 37 26 31 29 31 … …

Co10Ni15Cr6AlY Co30Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni15Cr12AlY Co30Ni15Cr12AlY Co10Ni30Cr12AlY Co30Ni30Cr12AlY Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY Co22Cr9AlY Co22Cr11AlY Fe22Cr11AlY (a) Test temperature 750 C

effect of aluminum is explained by the moreevident phase heterogeneity of the alloy structure, the worse solubility of chromium, and the presence of r-phase. The effect of aluminum on resistance to hot corrosion is similar to that observed in nickel alloys. Yttrium exerts a positive effect on resistance to hot corrosion, if the alloys have high and medium ratios of chromium and aluminum contents. For the other alloys, the effect of yttrium is detrimental. This is due to the formation of yttrium oxides, which are easily damaged by sulfur-base compounds. Studies of the structure of the sample surface zones after testing show that corrosion uniformly penetrates the alloys. Table 6.18 contains the results of iron-base alloy hot corrosion resistance tests for comparison. This system features high resistance to corrosion. However, when it is used for coatings, there is a problem of interaction between the coating and the superalloy.

Resistance of Coated Superalloys to Hot Corrosion Hot corrosion resistance tests of turbine-blade coatings were described in numerous papers. However, it is rather difficult to compare their results because of the differences between the testing procedures, the temperature conditions, as well as some uncertainty about the compositions and thickness of the coatings tested. The hot corrosion resistance test methods used differ from the alloy sample test methods. The difference is that for the coated specimens, the time to coating damage is determined instead of sample weight loss. In both cases, test results are comparable. By knowing the kinetics of corrosion in the alloy, one can calculate the time required for damage of the alloy top coat of a certain thickness, which is equal to the coating thickness.


During testing of coated specimens, a layer of GZT ash, weighed out as 120 g/m2, is applied to the surface. The specimens with the ash application are held in a furnace at the temperatures of 750, 800, 850, and 900 C. After a lapse of the specified time period, the specimens are taken from the furnace, weighed, and inspected. The corrosion products are left untouched on the specimen surfaces. After inspection, a layer of ash is applied once again, and testing is resumed. Until coating damage takes place, the specimen gains weight. As it does so, a dense film of corrosion products is formed on its surface. Before the coating is damaged, the specimen weight-gain rate increases. Coating damage is accompanied by an abrupt specimen weight loss due to spallation of corrosion products. From this moment, the specimen starts losing weight, and a thick, loose layer of corrosion products is formed on its surface. Coating life is evaluated by the time, up to the last holding, when the

Fig. 6.16

abrupt specimen weight loss occurs. If prior to this, only weight gain had been observed, then the moment when specimen weight loss occurs may be considered as the moment of damage. The choice of time to inspect the specimens is dictated by the test temperature. When the temperature is raised, holding time to inspection is shortened. By way of example, Fig. 6.16 presents kinetic curves of specific weight change at 800 and 900 C for the specimens with Co22Cr11AlY/ZrO215%Y2O3 coating deposited by the EB method. The presented results are obtained on specimens 7 mm in diameter and 30 mm high. When depositing overlay coatings on the specimens, a problem of specimen fastening in the unit for coating deposition arises. The areas where the specimens are attached to a fixture are free from the coating. This is why Co22Cr11AlY coating was deposited on the specimens under study in two steps. First, the coating was applied to one

Kinetics of hot corrosion of Co22Cr11AlY/ZrO2-15%Y203 coating tested in GZT ash. (a) 800 C, (b) 900 C


side of the specimen, then the specimen was fastened from the other side, and the coating was applied again. In the middle, the specimen was coated twice. This resulted in forming a zone with a coating of double thickness. As-deposited coating was peened with microballs, and specimens were annealed in vacuum at 1050 C for 2 h. The ceramic coating does not allow overlapping and so should be deposited only once. As a result, a small coating-free zone appeared on the specimen. The presence of such a zone reduces the protective effect of the ceramic layer. During testing, coating life depends on many technological factors. These factors dictate the similar composition of the evaporated alloy and

Table 6.19

the coating formed as well as the coating structure and the presence of various defects in it. When continuously deposited, the coatings demonstrate chemical composition homogeneity through the whole thickness and no directional structure, porosity, and microdots. The resistance of these coatings to hot corrosion is no less than that of the studied alloys. The presence of different defects reduces coating lifetime. The characteristics of the coatings studied are given in Table 6.14. The data on coating lifetime at hot corrosion testing are given in Table 6.19 and in Fig. 6.17 and 6.18. Test results are affected by the coating quality. This is why the table presents the data for the best and the worst specimens.

Coating life. Hot corrosion tests in GZT ash Coating life, h

Coating

750 C

800 C

850 C

900 C

Co22Cr11AlY Co22Cr11AlY/ZrO2-15%Y2O3 Ni20Co20Cr11AlY Ni20Co20Cr11AlY/ZrO2-15%Y2O3 Ni23Co28Cr9AlY Ni28Cr6AlY Ni20Cr12AlY with chromium carbide sublayer Fe12Ni20Cr12Al1.5SiY AlSiCrY AlSiBY AlPd Ni8Co12Cr7Al Ni7Co12Cr17Al (Ni8Co12Cr7Al/Al5Si1.5Y)

… 9000 … 9000 8500 … … 9000 … … 4000 … …

4500 4500 4500 4500 3000 4500 3500 4500 1000 600 7000 4500 4500

1500 3200 3000 4000 1500 3000 700 2500 400 400 2500 3000 3000

2600 2000 800 900 500 1000 200 5500 200 150 100 1500 1500

Life of diffusion coatings on JS6U alloy on testing in GZT ash (mAl 80 g/m2). 1, aluminized coating; 2, Al5Si1.5%Y electric arc coating; 3, Al14Si0.7Y electric arc coating; 4, AlPd coating

Fig. 6.17

Coating life; tested in GZT ash. 1, Co22Cr11AlY; 3, 2, Co22Cr11AlY/ZrO2-15%Y2O3; Ni20Co20Cr11AlY; 4, Ni20Co20Cr11AlY/ZrO2-15Y2O3; 5, Ni23Co28Cr9AlY; 6, Ni28Cr6AlY; 7, Ni20Cr12AlY with chromium carbide sublayer; 8, Fe12Ni20Cr12Al1.5SiY

Fig. 6.18


The common feature for all the coatings is an accelerated hot corrosion at raising the test temperature from 750 to 900 C. At 900 C, only the lifetime of iron-base coatings is higher than that at lower temperatures. Figure 6.17 presents the data on the corrosion life of the coatings alloyed with silicon, yttrium, and palladium and on the aluminized coatings formed in the powder mixture of 98% of the aluminum-iron alloy (50% Al) and 2% NH4Cl. The results of testing were based on the assumption that the coatings had the same aluminum reserve of 80 g/m2. Alloying diffusion coating with silicon makes its life 3 to 4 times longer than that of the siliconfree coating. All modifications of the coatings deposited by the EA method have longer life, in comparison with conventional aluminized coatings. The life of a diffusion coating, even if it is alloyed with chromium, silicon, and boron, is shorter than that of an overlay coating. The diffusion coating alloyed with palladium is distinguished by its protective properties. At 800 and 850 C, its properties are highly competitive with those of cobalt-base overlay coatings, which are 2 times as thick. At 900 C, protective properties of the aluminum-palladium coating

Table 6.20 Life coefficient of coatings. Hot corrosion tests in GZT ash Life coefficient Coating Co22Cr11AlY Co22Cr11AlY/ZrO2-15%Y2O3 Ni20Co20Cr11AlY Ni20Co20Cr11AlY/ZrO2-15%Y2O3 Ni23Co28Cr9AlY Ni28Cr6AlY Ni20Cr12AlY/chromium carbide Fe12Ni20Cr12Al1.5SiY AlPd

800 C

850 C

11.1 11.1 11.1 11.1 6.7 8.9 7.8 13.3 8.9

22.7 22.7 20.0 20.0 11.8 18.2 7.3 31.8 15.4

Table 6.21 Life of coatings on turbine blades of JS6U superalloys operating at 800 C Life, h Coating Co22Cr11AlY Co22Cr11AlY/ZrO2-15%Y2O3 Ni20Co20Cr11AlY Ni20Co20Cr11AlY/ZrO2-15%Y2O3 Fe12Ni20Cr12Al1.5SiY AlPd

Liquid fuel

Gas

20,000 20,000 20,000 20,000 25,000 …

70,000 70,000 70,000 70,000 80,000 55,000

worsen suddenly (Ref 14, 15). The effect of palladium is described in detail elsewhere (Ref 16). At 750 C, all overlay coatings feature high protective properties. A 900 h testing did not reveal any distinctions between their properties. At 800 C, most overlay coatings have lives of more than 4500 h. The exception is the Ni23Co28Cr9AlY coating. At this temperature, none of the systems has an obvious advantage. At 850 C, the differences in the coating lives are more substantial. At this test temperature, the cobalt-base coatings do not have advantages in life. The Ni28Cr6AlY coating has longer life than the widely used Co22Cr11AlY coating. The ceramic coat applied to the metallic coating protects it against damage and increases its life 1.5 times. This effect is achieved despite the previously mentioned defects, such as the presence of ceramic-free zones, thin ceramic layer, and imperfection of the ceramic deposition technique used. The effect caused by the ceramic coating can be more substantial (Ref 17–19). The blade ceramic layer not only protects the surface from aggressive compounds but also changes the surface temperature, thus promoting the reduction in condensation of aggressive compounds from gas flow. The cobalt-base coating advantages fully manifest themselves at 900 C. Lives of these coatings are 2 to 3 times longer than those of the nickel- and nickel-cobalt-base coatings. Ironbase coatings have high resistance to hot corrosion. The Ni7Co12Cr17Al coatings alloyed with tantalum, hafnium, and rhenium showed good protective properties at this temperature. Table 6.20 contains the calculated relative life values of the coatings tested. The coefficient of the coating lifetime is calculated as a ratio of coating lifetime and lifetime of the uncoated alloy. The time-to-damage of the alloy to the depth of 250 lm during its testing for hot corrosion was taken as an alloy lifetime characteristic. This corresponds to the specific weight loss of qⳮ ⳱ 0.2 g/cm2. This corrosion depth is considered to be allowable for stationary turbine blades (Ref 2). Some modifications of the coating allow for prolonging sample lives 20 to 30 times at 800 and 850 C. Table 6.21 contains the calculated data on the life of JS6U alloy blades used in stationary turbines at 800 C. The calculations were made on the basis of the studies of uncoated blade lives and relative lifetime coefficients given in Table 6.20.


REFERENCES

1. High-Temperature Corrosion and Protection of Materials, Proc. International Symposiums on High-Temperature Corrosion and Protection of Materials, R. Streiff, J. Stringer, R.C. Krutenat, and M. Caillet, Ed., (Les Embiez, France), 1986, 1989, 1992, 1996, 2000 2. V.I. Nikitin, Corrosion and Protection of Gas Turbine Blades, Leningrad, Machinostroenie, 1987, 272 p 3. E.L. Simoms, G.V. Browning, and H.A. Liebhaisky, Sodium Sulfate in Gas Turbines, Corrosion, (No. 2), 1955, p 505–514 4. J.A. Goebel, F.S. Petit, and G.W. Goward, Mechanisms for the Hot Corrosion of Nickel-Base Alloys, Metall. Trans., Vol 4 (No. 1), 1973, p 261–278 5. R.F. Reising, Effects of Chromium, Molybdenum and Tungsten on Sodium Sulfate Induced High Temperature Corrosion of Nickel, Corrosion, (No. 5), 1975, p 159– 163 6. J. Stringer, Hot Corrosion of High Temperature Alloys, Annu. Rev. Mater. Sci., (No. 7), 1977, p 477–509 7. M. Kawakami, K. Goto, and R.A. Rapp, Accelerated Oxidation (Hot Corrosion) of Alloys by Molten Salt, Iron Steel Inst. Jpn., (No. 9), 1980, p 646–658 8. G.W. Goward, Low-Temperature Hot Corrosion in Gas Turbines, Trans. ASME, (No. 2), 1986, p 421–424 9. V.I. Nikitin and I.P. Komossarova, Method of Protective Coating Testing for Lifetime, Proc. CKTI, (No. 176), 1980, p 67–73 10. V.I. Nikitin, Calculation of Metal Heat Resistance, Moscow, Metallurgy, 1976, 208 p 11. V.I. Nikitin, Y.A. Tamarin, and N.V. Zabrodina, Corrosive and Mechanical Properties of Protective Coatings, Collected Volume: Aqueous-Chemical Conditions and Corrosion of Power Equipment, No. 201, Leningrad, CKTI, 1986 12. N.V. Zabrodina, Y.A. Tamarin, and V.I. Nikitin, High-Temperature Corrosion of Aluminide Alloys for Overlay Coatings, Col-

13.

14.

15.

16.

17.

18.

19.

lected Volume: Heat-Resistant Coatings, St. Petersburg, Science, 1992, p 107–110 Y.A. Tamarin, E.B. Kachanov, and N.V. Zabrodina, Corrosion a Haute Temperature des Alliages Base Aluminure Destines aux Revetements Protecteurs, High Temperature Corrosion Advanced Materials and Coatings, R. Streiff, J. Stringer, R.C. Krutenat, and M. Caillet, Ed., Elsevier, London, 1989, p 361–365 A.I. Desipry, N.V. Zabrodina, and S.B. Shutova, Heat-Resistant Aluminide Coatings Containing Platinum and Palladium, Surface Layer, Accuracy and Performances of Machine Parts, Moscow, MDNTP, 1989, p 55–57 Y.A. Tamarin, E.V. Lopatin, and N.V. Klyueva, Research into Structural Characteristics of Diffusion Aluminide Coatings Alloyed with Palladium, Surface Layer, Accuracy and Performances of Machine Parts, Moscow, MAI, 1995 S. Alperine, P. Steinmetz, P. Josso, and A. Constantini, High Temperature-Resistant Palladium-Modified Aluminide Coatings for Nickel-Base Superalloys, High Temperature Corrosion Advanced Materials and Coatings, R. Streiff, J. Stringer, R.C. Krutenat, and M. Caillet, Ed., Elsevier, London, 1989, p 367–372 B.A. Movchan, I.S. Malachenko, and K.Y. Yakovchuk, Two- and Three-Layer Coatings Produced by Deposition in Vacuum for Gas Turbine Blades Protection, Surf. Coat. Technol., Vol 67, 1994, p 55–63 B.A. Movchan, Y.D. Skliarov, A.I. Rybnikov, and I.S. Malachenko, Corrosion Lifetime of Two-Layer Overlay Co-Cr-Al-Y/ ZrO2 Coating in Gas Turbine Fuel Ash, Adv. Spec. Electrometall., (No 3), 1988, p 52–56 A.S. Osyka, A.I. Rybnikov, S.A. Leontiev, N.V. Nikitin, and I.S. Malachenko, Experience with Metal/Ceramic Coating in Stationary Gas Turbine, Surface and Coatings Technology, Elsevier, 1995, p 187–192



CHAPTER 7

The Effect of Protective Coatings on the Mechanical Properties of Superalloys COATINGS PROTECT the surface of turbine blades from damage caused by high-temperature corrosion and thus preserve the structural shape of blades and their mechanical properties for the required time. There are many papers in the technical literature that deal with experimental studies of heat resistance, fatigue strength, and thermal fatigue of coated superalloys. However, the connection between the physical and mechanical properties of protective coatings and their effect on the mechanical properties of turbine-blade superalloys has not been studied sufficiently well. For thin diffusion coatings, this can be explained by the complexity of the task, both in reproducing the correct composition and in testing the coatings separately from the alloy. However, for these coatings, information on their physical and mechanical properties is necessary for their proper application and for accurate calculations of coated-blades service lives. The effect of coatings on mechanical properties of superalloys is a combination of several factors:

The efficiency of revealing the effect of coatings on superalloy properties depends on the testing methods employed. Cooled blades are essentially thin-walled constructions, and the effect of the coatings increases as the mismatch between the coating thickness and the bladewall thickness rises. Thus, to correctly determine the properties of blades protected with coatings, the specimens preferable for mechanical testing should have the ratio of the coating thickness and the specimen cross section close to those of the blades in use. Mechanical and physical properties of NiAl and Ni3Al aluminides, which form diffusion coatings, are dealt with in numerous publications (Ref 1, 2). Such aluminide properties as ultimate strength, yield strength, and elongation values depend to a considerable extent on their structure set by the process of test-specimen fabrication and their heat treatment. Table 7.1 refers to the test results presented in one of the earliest publications that dealt with

●

Table 7.1 Properties of Ni3Al and NiAl intermetalide-base cast alloys at tensile tests at 20 C

Coating deposition alters superalloy surface stresses. ● The coating can be damaged under action of static or alternate stresses; the cracks nucleating in the coating may accelerate the superalloy damage process. ● The mechanical properties of superalloys may be affected by the coating deposition thermal cycle. ● The coating changes the surface layer microstructure of a superalloy, and this effect is boosted as high-temperature exposure time increases.

Aluminum content, wt% 13.3 14.0 25.0 28.0 30.0 31.5 Source: Ref 3

Ultimate strength (rB), MPa

Yield strength (r0,2), MPa

Elongation (d), %

205 280 163 162 103 102

86 110 … … … …

1.1 1.0 0 0 0 0


research into aluminide features (Ref 3). Mechanical properties of NiAl compound versus its test temperatures are shown in Fig. 7.1 (Ref 3, 4). Despite different elongation values for the NiAl samples reported by different authors, they agreed that, in fact, elongation-temperature relation yielded quite the same patterns. At a certain temperature dependent on a NiAl compound composition, it experiences ductile-brittle transition that is typical of all the materials with body-centered cubic lattice. Research into the high-temperature strength of NiAl and Ni3Al compounds has revealed that their high-temperature strength is not so high as that of superalloys. It can be improved by alloying and by forming directional or single-crystal structures (Ref 5). The previously mentioned

patterns of aluminide mechanical properties variations can be taken into consideration as a certain approximation in the research into the behavior of diffusion aluminide coatings on superalloys. Reproducing aluminide compositions identical to numerous diffusion coating modifications is not a simple task. It is as difficult as their testing, because a sample thickness is comparable with a coating thickness. That is why the ductile-brittle transition temperature (DBTT) of a diffusion coating is recommended as its service life criterion under static and cyclic loading (Ref 6). The fact that aluminides experience ductile-brittle transition is crucial for many diffusion coatings characteristics and their effect on superalloy properties. As an example, Fig. 7.2 demonstrates test results for diffusion coatings 50 lm thick (Ref 7, 8). The increase in aluminum content of the coatings results in the increase in their ductilebrittle transition temperature. For coatings including 16 to 17% Al, the DBTT is below 0 C, while for coatings including 36% Al, the DBTT is 800 C. Ductile-brittle transition temperatures of the diffusion coatings on superalloys are higher than those of the coatings on nickel.

Thermal Expansion Coefficients and Elasticity Modulus of Coating Alloys Mechanical properties of NiAl compound of stoichiometric composition at different temperatures. 1, Ultimate strength; 2, yield strength; 3 and 4, elongation. Source: Ref 3, 4

Fig. 7.1

Ductile-brittle transition temperatures for diffusion coatings 50 lm thick with different aluminum contents of their outer layer. 1, Coatings on JS-type superalloys; 2, coatings on nickel. Source: Ref 7, 8

Fig. 7.2

The mismatches between the thermal expansion coefficients (TEC) and the elasticity modulus of coatings and superalloys determine the level of stresses that arise on the blade surfaces after coating deposition, heat treatment, and under their service conditions. Thermal stresses on the surface have an effect on all the mechanical properties of coated superalloys. For the Ni-Cr-Al and Ni-Co-Cr-Al systems, TEC studies were carried out on samples made by extruding cast billets. The Co-Ni-Cr-Al system samples are studied as-cast. To study TEC, the 0.5 to 1.0 mm thick condensates deposited by the electron beam (EB) method were also used. Before dilatometric analysis, all the samples were subjected to vacuum annealing at 1100 C for 2 h followed by slow cooling down. The structure and phase compositions of the extruded samples are close to those of overlay coatings with the same chemical composition. Although the cast alloys feature larger structural constituent sizes, their phase composition is also

The Effect of Protective Coatings on the Mechanical Properties of Superalloys / 121

similar to that of the coatings of the same chemical composition. Studies of TECs were carried out using a quartz dilatometer of 1500 N (made by SincuRico) equipped with an induction transducer as a sensor. Sample elongation and temperature recording were carried out with the interval of 2 C. In combination with high sensitivity of the induction transducer, it allowed detection of phase changes and transformations that cause relative variations of the sample length as small as 0.001%. The total error at measuring mean TEC was (0.25 to 0.35) • 10ⳮ6, degⳮ1, assuming confidence coefficient (P) level of P ⳱ 0.95. However, in the case of measuring elongation of flat samples made of condensates, the error was as high as 5 to 10%. Dynamic modulus of elasticity was determined using the Elasomat instrument and cylindrical cast samples that were 120 mm long and 8 mm in diameter. Flat samples made of condensate with the dimensions of 100 by 10 by 2 mm were also tested. Relative errors of measuring elasticity modulus for cast samples and flat samples were 1% and 10%, respectively. Ni-Cr-Al Alloys. The TECs for the Ni-Cr-Al system are given in Table 7.2. (Regression equations derived from statistics-based processing of the experimental data on TECs are given in the Appendix.) Thermal expansion coefficient values for the alloy group under study are determined in each alloy by the volume ratio of its basic phase components, such as c-solid solution, c⬘-phase

Table 7.2

(Ni3Al), b-phase (NiAl), and ␣-Cr. Each of these phases has TEC monotonically rising as the temperature increases, whereupon its absolute value is less for the phases with greater aluminum contents. By the mean TEC values in the temperature range of 20 to 1000 C, the phases of the system under consideration are ranged as follows: c (19.8 • 10ⳮ6, Kⳮ1), Ni3Al (16.7 • 10ⳮ6, Kⳮ1), NiAl (16.3 • 10ⳮ6, Kⳮ1), and ␣Cr (9.6 • 10ⳮ6, Kⳮ1). The increase in the aluminum content of the alloy and, hence, a greater number of aluminumrich phases (NiAl) cause reduction of the TEC mean values in the range of 20 to 1000 C. More chromium in the alloy produces an effect similar to that of aluminum, because it leads to an ␣-Cr volume increase. The said general features of the effect of aluminum and chromium on the TECs of multiphase alloys of the Ni-Cr-Al system are effective up to about 600 C; in a higher temperature range, phase transformations lead to sudden TEC variations. On consideration of the TEC effect on the coating-alloy system properties, it is not the absolute value of this characteristic that is really important but its relative value while comparing it with superalloy TECs. The Ni-Cr-Al-system alloys and superalloys are much the same in their thermal coefficients. When overlay coatings interact with superalloys while being high-temperature tested, cobalt diffuses from the superalloys into the Ni-Cr-Alsystem coatings, and after about 100 h at temperatures higher than 1000 C, its content reaches the average cobalt content of a super-

Thermal expansion coefficients, ␣, of Ni-Cr-Al alloys ␣ • 106, Kⴑ1

Alloy(a) Ni14Cr3AlY Ni18Cr5AlY Ni20Cr6AlY Ni16Cr9AlY Ni21Cr9AlY Ni22Cr11AlY(b) Ni7Cr12AlY Ni19Cr12AlY Ni8Cr15AlY Ni18Cr16AlY Ni8Cr18AlY Ni5Cr25Al Ni31Al Ni19Cr10AlY(c) CMSX-4(d)

20–100 C

100–200 C

200–300 C

300–400 C

400–500 C

500–600 C

600–700 C

700–800 C

800–900 C

900–1000 C

13.7 13.3 … 12.4 12.0 11.7 13.1 12.0 11.8 12.2 11.6 13.8 13.4 11.8 12.1

14.4 13.9 12.7 12.9 12.7 12.6 13.7 12.7 12.6 12.8 12.6 14.4 13.9 12.7 12.7

15.1 15.0 13.4 13.6 13.4 13.2 14.4 13.4 13.5 13.3 13.4 15.4 14.3 13.4 13.3

15.9 15.3 14.4 14.4 14.4 14.6 15.4 14.0 14.0 13.8 14.4 16.5 13.0 14.4 13.9

16.2 16.0 14.7 15.1 14.7 14.8 15.7 14.8 14.3 14.6 13.7 16.3 13.3 14.7 14.4

17.9 17.2 15.4 16.3 15.2 15.5 16.9 14.3 14.6 13.2 18.4 9.8 13.9 15.4 15.0

19.9 19.1 17.1 17.7 16.3 17.2 18.6 16.1 17.6 15.1 16.1 19.6 14.4 17.1 16.4

20.3 21.2 15.1 15.8 14.9 16.9 18.6 15.0 15.3 11.3 12.0 19.6 14.5 15.1 17.4

19.4 23.0 19.0 13.2 10.6 18.7 18.6 16.2 15.6 19.3 19.7 18.6 14.6 9.0 19.1

20.4 28.6 25.2 26.5 21.6 31.0 19.7 31.1 17.0 39.1 … 19.5 14.8 25.2 21.5

(a) Alloy chemical compositions are presented in Table 4.10. (b) In the range of 1000 to 1100 C, ␣ ⳱ 28.0 • 10ⳮ6, Kⳮ1; in the range of 1100 to 1200 C, ␣ ⳱ 22.6 • 10ⳮ6, Kⳮ1. (c) Condensate: nickel base, 18.5% Cr, 10.4% Al, 0.01% Y. (d) In the range of 1000 to 1100 C, ␣ ⳱ 25.6 • 10ⳮ6, Kⳮ1; in the range of 1100 to 1200 C, ␣ ⳱ 31.4 • 10ⳮ6, Kⳮ1


alloy. Therefore, when analyzing coating TEC, especially if the coatings are intended for longterm service, it is worth using the TEC values of the Ni-Co-Cr-Al system. The results of elasticity modulus measurements for the Ni-Cr-Al alloys are given in Table 7.3. Information on statistics-based processing of experimental data is presented in the Appendix. The value of the alloy elasticity modulus is a result of the combined elasticity modulus of the phases that form it. The NiAl (Ni8Cr18AlY)base alloy features the lowest value of elasticity modulus, wherein, with a temperature raising to 800 C, the modulus value does not actually change. The maximum elasticity modulus is demonstrated by the alloys whose phase composition corresponds to c-solid solution (Ni15Cr3AlY) and c⬘-phase (Ni7Cr12AlY) with the low chromium content. The value of elasticity modulus for all the alloys (except for those with the NiAl structure) monotonically decreases as the temperature rises. The absolute value of elasticity modulus for the said group of alloys is lower than that for nickel-base superalloys with an equiaxial structure. Ni-Co-Cr-Al Alloys. To determine TECs of the Ni-Co-Cr-Al alloy major phases, these phases were extracted electrochemically and their TECs were determined by high-temperature x-ray diffraction methods. As for the Ni-CrAl system, their mean TEC in the temperature range of 20 to 1000 C decreases as the aluminum content of the phases increases: c (17.2 • 10ⳮ6, Kⳮ1), Ni3Al (15.2 • 10ⳮ6, Kⳮ1), and NiAl (13.8 • 10ⳮ6, Kⳮ1). The experimental TEC data for the Ni-Co-CrAl-system alloys are given in Table 7.4 (Ref 9). The statistics-based processing of the experi-

Table 7.3

mental TEC data for the Ni-Co-Cr-Al alloys was conducted in different temperature ranges. In the temperature range of 100 to 400 C, the influence of cobalt, chromium, and aluminum on thermal effects of TEC variations was considered. In the range of 900 to 1000 C, the effects of TEC variations during phase transformations were discussed; in the range of 100 to 1000 C, the effect of alloying on the mean TEC values was determined. The regression equations are given in the Appendix. A cobalt-content increase from 10 to 30% causes an increase of the maximum TEC level for the Ni-Co-Cr-Al alloys in the temperature range of 100 to 400 C (Fig. 7.3a). Alloying the alloys with 8 to 14% Co reduces TEC. In this case, with the chromium content raised from 15 to 30%, TEC values do not actually vary. Aluminum, chromium, and cobalt exert a crucial effect on alloy TEC in the temperature range of 900 to 1000 C (Fig. 7.3b). A cobalt content increase from 10 to 30% results in the reduction of the maximum TEC level. The increase in aluminum and chromium contents of the alloys causes TEC reduction too, which is mainly associated with an increase in NiAl and ␣-Cr volume fractions. If the aluminum, chromium, and cobalt contents of the alloys are at their maximum levels, a region with comparatively low TEC values of ␣ 20 • 10ⳮ6, Kⳮ1, appears. The TEC behavior in the temperature range of 900 to 1000 C is well correlated with variations of the NiAl-Ni3Al ratio, depending on alloying. The more NiAl is in the alloy, the lower the level of TEC. The results of the research into Ni-Co-Cr-Al alloy elasticity modulus are given in Table 7.5. Due to statistics-based processing of test results, a regression equation is derived that links the

Elasticity modulus, E, of Ni-Cr-Al alloys E • 10ⴑ2, MPa

Alloy

20 C

200 C

300 C

400 C

500 C

600 C

700 C

800 C

Ni14Cr3AlY Ni18Cr5AlY Ni16Cr9AlY Ni21Cr9AlY Ni22Cr11AlY Ni22Cr11AlY(a) Ni7Cr12AlY Ni19Cr12AlY Ni8Cr15AlY Ni18Cr16AlY Ni8Cr18AlY

1784 1743 1665 1571 1696 1670 1806 1596 1424 1624 1079

1714 1681 1599 1517 1672 … 1749 1615 1370 1604 1065

1657 1614 1565 1477 1644 … 1696 1602 1346 1599 1092

1601 1564 1510 1438 1612 … 1656 1575 1329 1579 1091

1536 1500 1469 1393 1580 … 1609 1545 1315 1552 1085

1469 1431 1429 1352 1530 … 1545 1515 1283 1523 1072

1431 1361 1379 1305 1481 … 1488 1486 1249 1495 1062

1354 1304 1348 1270 1441 1440 1446 1462 … 1469 1047

(a) Condensate from Ni22Cr11AlY alloy


elasticity modulus value with alloying element contents of the Ni-Co-Cr-Al alloys in the temperature range of 20 to 800 C. Calculation data available from the equations for alloy elasticity modulus at the temperatures of 20 and 800 C are given in Fig. 7.4(a, b). Aluminum reduces the value of elasticity modulus mainly due to a higher volume of NiAl in the alloys. The cobalt effect is much stronger, especially if the alumiTable 7.4

num content is high. The influence of both elements is explained by a rise in atom interaction energy when the nickel-base c-solid solution is alloyed with these elements. Elastic modulus values determined for the vapor-deposited samples are close to those determined for cast alloys (Table 7.5). These similarities are caused by measurement errors in thin, flat samples and by differences in the composi-

Thermal expansion coefficients, ␣, of Ni-Co-Cr-Al alloys ␣ • 106, Kⴑ1

Alloy(a) Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni30Co30Cr14Al Ni10Co22Cr11Al Ni30Co22Cr11Al Ni20Co15Cr11Al Ni20Co30Cr11Al Ni20Co22Cr8Al Ni20Co22Cr14Al Ni20Co22Cr11Al Ni20Co22Cr11AlY(b) Ni20Co22Cr11AlY(c) Ni6Co20Cr12AlY Ni6Co10Cr13AlTaReHf(d)

100–200 C

200–300 C

300–400 C

400–500 C

500–600 C

600–700 C

700–800 C

800–900 C

900–1000 C

13.2 13.6 13.0 13.3 13.3 13.4 13.0 13.3 13.2 12.1 13.7 13.4 14.1 13.9 13.8 13.5 13.0 13.2 13.2

13.9 14.4 13.8 13.6 14.2 14.1 13.5 13.6 13.6 14.2 14.3 14.3 14.5 14.1 14.5 14.0 15.3 13.6 14.2

15.0 15.3 14.6 14.6 14.6 15.5 14.1 15.0 15.2 15.7 15.5 15.2 16.3 15.4 15.3 15.1 15.3 15.2 14.7

15.5 16.0 15.0 15.4 16.7 17.1 14.8 15.4 15.8 16.7 16.6 16.4 16.7 16.1 16.8 15.7 15.3 15.8 14.2

16.7 17.8 15.7 16.6 16.7 17.5 16.2 15.7 16.0 18.3 17.7 16.4 17.4 17.2 16.8 16.9 16.7 16.0 13.6

19.9 20.7 18.1 19.0 19.1 19.3 17.1 17.3 18.0 20.8 20.9 20.1 21.8 18.3 21.3 20.0 16.7 18.0 17.5

21.5 22.8 21.8 19.7 21.9 21.0 18.8 18.1 20.5 21.3 23.7 20.4 25.2 21.6 25.0 22.8 16.7 20.5 19.2

25.3 25.7 25.2 19.3 23.7 21.0 20.6 17.8 25.3 21.3 25.4 21.3 27.7 23.7 24.7 22.9 20.0 25.3 22.1

29.5 28.3 29.9 20.1 28.4 21.5 21.6 19.8 31.1 23.5 28.5 25.3 27.7 25.9 25.1 22.8 20.0 31.1 27.8

(a) Alloys chemical compositions are presented in Table 4.11. (b) In the range of 1000 to 1100 C, ␣ ⳱ 25.7 • 10ⳮ6, Kⳮ1; in the range of 1100 to 1200 C, ␣ ⳱ 23.8 • 10ⳮ6, Kⳮ1. (c) Condensate from Ni20Co22Cr11AlY alloy. (d) Chemical composition: nickel base, 6.0% Co, 9.9% Cr, 12.9% Al, 3.9% Ta, 0.55% Re, 1.4% Hf, and 0.4% Si

Fig. 7.3

Thermal expansion coefficients of Ni-Co-Cr-Al alloys (Kⳮ1). (a) 100 to 400 C. (b) 900 to 1000 C


tions of the condensate samples and the alloys used for their deposition. Co-Ni-Cr-Al Alloys. The results of the research into Co-Ni-Cr-Al-system alloy TECs are given in Table 7.6. Up to 500 C, the temperature elevation results in a monotonical TEC rise. At the temperatures above 500 C, an abrupt TEC change is observed due to magnetic transformation and dissolution of b- and r-phases. The statistics-based processing of the experimental measurement results allows for drawing the following conclusions. In the temperature range up to 500 C, the increase in aluminum and chromium contents reduces TEC from 16 • 10ⳮ6 to 13 • 10ⳮ6, Kⳮ1. Alloying with nickel produces the same effect. If the effect of alloying elements within the whole temperature range from 20 to 1000 C is considered, chromium has the strongest effect on TEC. It reduces TEC. Nickel slightly reduces TEC; the effect of aluminum is not so strong. When the temperature rises to 800 C, the elasticity modulus of Co-Ni-Cr-Al alloys monotonically decreases to 1262 • 102 to 1492 • 102, MPa (Table 7.7). The regression equation linking elasticity modulus with alloying element content and temperature is given in the Appendix. The effect of the alloying elements is similar at both low and high temperatures.

Mechanical Properties of Coating Alloys The coatings deposited on turbine blade surfaces are affected by the same static and alterTable 7.5

nate stresses that have an impact on the blade materials. The level of coating mechanical properties has an effect on the basic characteristics of blades, such as their high-temperature strength, endurance, and thermal fatigue. Determining the mechanical properties is a hard task. Even fabrication of cast samples of the alloys with high aluminum content is hampered by their low ductility. It is also worth taking into account that at high-temperature testing and using the blade coatings, their compositions and properties vary continuously. Strength and ductility of the alloys used for coatings were investigated using the standard procedures for cast and vapor-deposited samples. The cast samples 5 mm in diameter and the flat samples 0.5 to 1.0 mm thick made by vapor deposition were annealed before testing at 1050 C for 4 h. Test results for Ni-Cr-Al-system alloys are presented in Tables 7.8 and 7.9 and in Fig. 7.5. The Ni20Cr6AlY alloy with (cⳭc⬘)-phase composition features moderate-strength values. Ductility of this alloy at temperatures up to 1000 C does not exceed 5%. The Ni8Cr15AlY alloy based on Ni3Al has higher strength. At temperatures above 800 C, its strength exceeds the strength of all the alloys studied. Ductility of the alloys with (cⳭc⬘)-phase composition depends on the aluminum content and other alloying elements. The Ni3Al-base systems are used for fabrication of superalloys and can be considered most promising for designing the coatings highly resistant to thermal fatigue. The Ni22Cr11AlY alloy is most widely used as a coating on blades operating under conditions of high temperatures and thermal stresses.

Elasticity modulus, E, of Ni-Co-Cr-Al alloys E • 10ⴑ2, MPa

Alloy

20 C

200 C

300 C

400 C

500 C

600 C

700 C

800 C

Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni10Co22Cr11Al Ni30Co22Cr11Al Ni20Co15Cr11Al Ni20Co22Cr8Al Ni20Co22Cr11Al Ni20Co22Cr11AlY(a)

1560 1699 1664 1858 1307 1550 1364 1486 1737 1566 1602 1633 1549

1499 1637 1602 1799 1308 1523 1341 1461 1688 1530 1548 1595 1514

1461 1586 1573 1747 1301 1501 1324 1440 1656 1500 1506 1563 1476

1461 1533 1538 1705 1286 1476 1306 1404 1625 1472 1472 1533 1444

1394 1482 1491 1662 1274 1446 1278 1373 1583 1435 1431 1496 1410

1358 1439 1449 1661 1255 1414 1259 1337 1541 1400 1379 1464 1365

1300 1357 1404 1556 1233 1373 1238 1299 1496 1359 1318 1422 1311

1237 1267 1322 1473 1173 1316 1193 1249 1426 1302 1263 1373 1248

(a) Condensate from Ni20Co22Cr11AlY alloy


Fig. 7.4

Table 7.6

Elasticity modulus of Ni-Co-Cr-Al alloys (MPa). (a) 20 C. (b) 800 C

Thermal expansion coefficients, ␣, of Co-Ni-Cr-Al alloys ␣ • 106, Kⴑ1

Alloy(a) Co10Ni15Cr6AlY Co30Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni15Cr12AlY Co30Ni15Cr12AlY Co10Ni30Cr12AlY Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY Co22Cr9AlY

20–100 C

100–200 C

200–300 C

300–400 C

400–500 C

500–600 C

600–700 C

700–800 C

800–900 C

900–1000 C

14.3 12.8 13.7 13.3 13.9 13.8 12.3 13.8 13.4 14.2 12.8 13.8 13.0 13.1 …

15.2 13.6 14.2 14.2 14.9 14.7 12.9 14.3 14.0 14.9 13.5 14.6 13.6 13.8 13.3

16.1 14.6 14.9 15.1 15.7 15.7 13.5 14.8 14.7 15.6 14.0 15.3 14.5 14.4 13.7

17.1 15.5 15.5 16.1 16.9 16.7 14.2 15.3 15.7 16.5 14.8 16.1 15.3 15.2 14.1

17.8 16.5 16.2 16.7 17.5 17.4 14.8 16.6 16.2 17.1 15.8 16.8 16.5 16.1 14.5

21.4 17.3 19.2 19.3 19.9 18.8 15.3 17.6 17.5 19.4 17.0 19.3 17.4 18.0 14.8

35.5 18.0 20.6 21.7 27.6 20.3 15.8 15.6 22.5 26.6 18.8 22.4 19.0 19.9 14.8

23.1 19.0 19.2 25.2 27.7 21.4 16.5 14.8 22.8 21.0 15.6 19.3 17.3 18.6 14.7

20.2 20.3 18.7 18.1 21.3 20.3 19.2 16.8 20.1 18.9 … 16.8 18.5 17.0 15.2

20.5 22.0 18.2 15.6 15.4 20.3 21.7 21.8 … 18.5 … 18.1 23.3 20.6 16.1

(a) Alloy chemical compositions are presented in Table 4.13

Table 7.7

Elasticity modulus, E, of Co-Ni-Cr-Al alloys E • 10ⴑ2, MPa

Alloy

20 C

100 C

200 C

300 C

400 C

500 C

600 C

700 C

800 C

Co10Ni15Cr6AlY Co30Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni15Cr12AlY Co30Ni15Cr12AlY Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY

2071 1893 1960 1820 1956 1826 2013 1820 1792 2037 1888 1778 1847

1956 1856 1923 1787 1911 1787 1982 1795 1759 2006 1843 1760 1799

1874 1803 1861 1726 1839 1763 1937 1749 1706 1970 1781 1728 1749

1809 1740 1792 1683 1779 1714 1888 1704 1657 1919 1717 1702 1700

1738 1655 1702 1603 1703 1674 1823 1649 1586 1866 1648 1662 1646

1662 1601 1618 1549 1641 1621 1764 1575 1540 1805 1576 1643 1600

1573 1549 1535 1469 1568 1554 1690 1516 1666 1739 1476 1578 1534

1492 1386 1447 1366 1458 1463 1625 1434 1363 1679 1372 1515 1426

1262 1266 1353 1284 1334 1308 1492 1345 1295 1492 1289 1438 1344


at 1100 C. Such a change in the properties of this alloy is due to the development of phase transformations. Intensive softening of the Ni20Cr12AlY alloy starts at temperatures above 700 C, when c⬘-phase is dissolving in the alloy. Tables 7.10 and 7.11 and Fig. 7.5 present the test results on the mechanical properties of the NiCo-Cr-Al alloys. At temperatures below 600 C,

Its properties are typical of all the materials of the Ni(Co)-Cr-Al system used for depositing overlay coatings. At temperatures under 800 C, the alloy features high strength, which abruptly diminishes as the temperature rises to 1100 C. At 500 C, elongation of the Ni22Cr11AlY alloy does not exceed 3%. However, at higher temperatures, elongation increases abruptly to 69% Table 7.8

Ultimate strength, rB, of Ni-Cr-Al alloys rB, MPa

Alloy Ni22Cr11AlY Ni22Cr6AlY Ni20Cr6AlY(a) Ni8Cr15AlY(a) Ni5Cr25Al

500 C

600 C

700 C

800 C

900 C

975 C

1000 C

1100 C

728 316 550 … …

… … 530 650 300

672 406 390 500 250

396 341 190 … …

216 306 … 400 200

… … … 358 150

68 236 65 … 120

34 … … 157 50

(a) Samples are vapor deposited from alloy of the said composition

Table 7.9

Elongation, d, of Ni-Cr-Al alloys d, %

Alloy Ni22Cr11AlY Ni22Cr6AlY Ni20Cr6AlY(a) Ni8Cr15AlY(a) Ni5Cr25Al

500 C

600 C

700 C

800 C

900 C

975 C

1000 C

1100 C

1200 C

… 1 16 … …

… … … 0 0

24 1.2 16 0 10

31 2.4 … … …

74 2.6 49 10 20

… … … 12 25

105 3.2 85 … 30

… … … 31 …

… … … … 70

(a) Samples are vapor deposited from alloy of said composition

Fig. 7.5

(a) Ultimate strength and (b) elongation of alloys. 1, Ni20Co20Cr12AlY; 2, Ni20Cr12AlY; 3, Ni10Co20Cr12AlY; 4, Co20Cr12AlY; 5, Ni20Cr7AlY; 6, NiAl (33% Al)


(bⳭc) phase composition, which causes an abrupt drop of its ultimate strength and an increase of its ductility as the temperature rises. Cobalt-base alloys also feature monotonically decreasing strength and increasing ductility as the temperature rises from 500 to 1100 C (Tables 7.12 and 7.13 and Fig. 7.5). The strength of these alloys depends slightly on their alloying. It is related to the fact that the said system lacks c⬘-strengthening phase. The Appendix contains the regression equations linking strength and ductility of the Co-Ni-Cr-Al-Y-system alloys with their alloying element contents. The comparison of strength and ductility of the cast and vapor-deposited samples of the

the Ni20Co20Cr12AlY and Ni10Co20Cr12AlY alloys have high ultimate strength and relatively low ductility (d 5%). As the temperature rises, the alloy strength diminishes monotonically, and at 1100 C, it is 30 to 40 MPa. Ductility of the alloys under study rises monotonically, and at temperatures above 1000 C, some alloys (e.g., Ni10Co22Cr11AlY) are in a superductile state; their elongation exceeds 100%. Different alloy behaviors, which manifest themselves in temperature-dependent variations of their strength and ductility parameters, are caused by their phase composition and alloying with cobalt. In the temperature range higher than 600 C, the Ni20Co20Cr12AlY alloy alloyed with 20% Co shows no trace of c⬘-strengthening phase. It has Table 7.10

Yield strength (r0,2) and ultimate strength (rB) of Ni-Co-Cr-Al alloys r0,2, MPa

Alloy Ni10Co15Cr8AlY Ni30Co15Cr8AlY Ni10Co30Cr8AlY Ni10Co22Cr11AlY Ni10Co20Cr12AlY(a) Ni20Co22Cr8AlY Ni20Co22Cr11AlY Ni20Co20Cr12AlY(a)

rB, MPa

800 C

900 C

1000 C

500 C

700 C

800 C

900 C

1000 C

1100 C

… … 343 … … 344 292 …

335 246 194 … … 230 113 …

178 103 73 … … 72 69 …

… … … 668 … … 840 520

… … … 602 … … 514 510

698 572 434 429 260 470 319 320

415 268 234 233 190 249 148 …

219 112 88 62 55 82 84 50

70 41 56 … … 48 45 …

(a) Samples are vapor deposited from alloy of the said composition.

Table 7.11

Elongation (d) of Ni-Co-Cr-Al alloys d, %

Alloy

500 C

700 C

800 C

900 C

1000 C

1100 C

… … … 0.5 … … 4 3

… … … 3 … … 28 …

4 5 27 14 … 20 39 25

7 7 32 23 3 41 65 70

14 15 68 84 130 42 68 130

34 37 66 … … 15 … …

Ni10Co15Cr8AlY Ni30Co15Cr8AlY Ni10Co30Cr8AlY Ni10Co22Cr11AlY Ni10Co20Cr12AlY (a) Ni20Co22Cr8AlY Ni20Co22Cr11AlY Ni20Co20Cr12AlY (a) Condensate from alloy of said composition

Table 7.12

Yield strength (r0,2) and ultimate strength (rB) of Co-Ni-Cr-Al alloys r0,2, MPa

Alloy Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY

rB, MPa

500 C

700 C

900 C

500 C

700 C

900 C

1100 C

496 … 392 564 494 … 493

392 428 315 479 334 202 376

161 126 122 165 158 … 178

567 … 690 866 787 … 810

546 547 461 567 496 474 531

205 165 142 173 190 249 188

60 57 42 54 53 64 60


Ni20Co20Cr12AlY alloy (Fig. 7.6a,b) demonstrates that the principal trend of properties change versus temperature is the same. At the same time, the mechanical properties of the vapor-deposited sample are slightly worse than those of the cast sample (50 and 84 MPa at 1000 C, respectively), but its ductility is higher. The vapor-deposited and cast samples of other alloys have the same relationship between their properties. Differences in the properties are explained by structural differences, such as more dispersed phases in the vapor-deposited samples, different chemical compositions of vapordeposited samples and evaporated alloys, and sample shapes and sizes. To determine the moment when the alloys become ductile, the alloys of the Ni-Cr-Al and NiCo-Cr-Al systems were bend tested. Testing was carried out by applying certain steady loading to the central part of the sample and then measuring its strain. The test samples were fabricated by the extrusion method and had a dispersed structure similar to that of vapor-deposited samples and coatings. The Ni20Co20Cr12AlY alloy experiences deformation even at room temperature. At 500 C, the sample is quite ductile when bent (Fig. 7.7). The cobalt-free alloy of the same chromium and aluminum contents demonstrates noticeable strains only at 600 C. The Ni10Co15Cr8AlY and Ni30Co15Cr8AlY alloys differing only in cobalt contents have different ductility levels. The 30% Co alloy becomes ductile even at room temperature, while the 10% Co alloy becomes ductile at temperatures above 500 C. The increase in aluminum content of the alloy up to 14% results in overriding the cobalt effect. The high aluminum content and, respectively, high NiAl content (of 70 to 80%) raise the alloy plasticization temperature to 800 C. The previously mentioned test results show that at 500 to 1000 C, the strength levels of the cast alloys used for coatings, their vapor-depos-

Table 7.13

Elongation, d, of Co-Ni-Cr-Al alloys d, %

Alloy Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY

500 C

700 C

900 C

1100 C

1 … 11 12 9 … 10

13 15 19 29 25 … 28

49 27 72 66 53 37 50

74 37 61 72 56 57 63

ited samples, and the respective coatings are considerably lower than those of modern superalloys. That is why the protective coating cannot bear any static and alternate stresses generated in turbine blades. This should be taken into account when calculating stresses and strains in coated turbine blades.

Thermal Stresses in Superalloy Coatings A study of the coating stressed state is one of the basic elements used for development of coatings and their use. Thermal stress generation during heating and cooling of coated parts caused by a mismatch between the TECs of a coating and a superalloy is typical for both diffusion and overlay coatings. Thermal stress in a coating (rc) may be calculated from a ratio of coating and superalloy TECs: rc ⳱ Ec •

(␣c ⳮ ␣sa) • (T ⳮ T0) 1 ⳮ lc

where Ec is coating modulus of elasticity; ␣c and ␣sa are TECs of a coating and a superalloy in the temperature range of T0 to T, respectively; T0 is a temperature of zero thermal stress level in a coating-superalloy system; T is a current temperature; and lc is Poisson’s ratio of a coating. Thermal stresses generated in the coating-superalloy system are crucial to the destruction of the coating on turbine blade surfaces during heating and cooling under transient conditions typical of aircraft gas-turbine engines. Experimental Studies of Thermal Stresses. To study stressed coatings, two methods were used: ●

Technique for measurement of strain caused by coating chemical removal from a sample surface ● X-ray diffraction technique for assessing coating stresses by measuring crystal lattice strain (Ref 10, 11) As mentioned in many publications, compressive stresses occur in diffusion aluminide coatings at room temperature. Their generation is caused by two factors: ●

When nickel superalloys undergo aluminizing, NiAl3, Ni2Al3, NiAl, and Ni3Al form, their specific volumes (0.25, 0.21, 0.17, and


0.14 cm3 /g, respectively) being larger than the specific volume of the superalloy (0.11 to 0.12 cm3 /g). ● Thermal expansion coefficients of aluminides are less than those of superalloys. The compressive stress level in aluminide coatings depends on the aluminizing technique in use. The results of the studies pursued show that on aluminizing a nickel superalloy of JStype from the slurry containing 100% Al at the temperature of 700 C for 4 h, the residual com-

Fig. 7.6

(a) Ultimate strength and (b) elongation of samples made from alloy condensates. 1, Ni20Co20Cr12AlY; 2, Ni10Co20Cr12AlY; 3, Ni20Cr7AlY; 4, Ni5Cr8AlY; 5, Ni20Cr12AlY

Ductility of alloys during bend testing. 1, Ni10Co15Cr8Al; 2, Ni30Co15Cr8Al; 3, Ni10Co30Cr8Al; 4, Ni10Co15Cr14Al; 5, Ni30Co15Cr14Al; 6, Ni22Cr11AlY; 7, Ni20Co22Cr11AlY

Fig. 7.7

pressive stress of r ⳱ ⳮ200 MPa generates in the coating surface layer at room temperature. In the coating applied by aluminizing at 950 and 1200 C for 4 h, the compressive stresses are ⳮ160 and ⳮ140 MPa, respectively. The stress level in diffusion coatings decreases with the temperature increase. The data on thermal stresses in the diffusion coating measured by the x-ray method at elevated temperatures are presented in Fig. 7.8. The coating was applied by aluminizing the JS6U superalloy in a

Thermal stresses in (1) diffusion coating on JS6U superalloy versus temperature. Thermal expansion coefficients of (2) NiAl and (3) superalloy. Source: A.I. Samoilov and I.A. Ignatova

Fig. 7.8


mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 950 C for 4 h. The extreme points on the stress-temperature curve are caused by different TECs of the superalloy and the coatingbase aluminide. At high-temperature testing, the stress level of the diffusion coating depends on its test time and temperature. However, on cooling from the test temperature, the compressive stresses always rise in diffusion coatings. When analyzing a stressed state of an overlay coating, it is necessary to take into account the following initial conditions. When the EB method is used, the coating is deposited on a substrate heated to 850 to 950 C, and evaporation heat generated during its condensation also heats its thin surface layers. Despite the fact that no special preheating equipment is used at electric arc (EA) deposition, the substrate is heated by plasma flow up to 500 to 600 C. Under these conditions, no stresses are generated at the interface of the coating and the substrate on which the coating is deposited from vapor or plasma. At the heat treatment temperatures of 1000 to 1050 C, the coatings are in a ductile or, in some cases, superductile state (d 100%); their yield stress is lower than 50 MPa. In addition, at 1000 C, recrystallization occurs in the coat-

ings of the Ni-Cr-Al and Ni-Co-Cr-Al systems. On cooling from 1000 C, a mismatch between the TECs of the coating and the protected alloy causes thermal stress generation. As the temperature decreases, thermal stresses increase and reach their maximum at room temperature. The results of x-ray diffraction analysis of stresses at different stages of the coating deposition process and their heat treatment and machining are shown in Fig. 7.9. The coatings 80 to 100 lm thick made of Ni20Cr12AlY, and Ni20Co20Cr12AlY alloys were deposited by the EB method. All the measurements were carried out at room temperatures; the JS6U superalloy was used as a substrate material. The tensile stress level of Ⳮ(70 to 90) MPa is typical of the coatings as-deposited by the EB method on the substrate heated to 850 to 950 C. On annealing at 1030 C for 2 h, tensile stresses in the coatings increase. The tensile stress level in the coating alloyed with cobalt is higher. On annealing at 1030 C of the Ni20Cr12AlY coating applied by the EA method, its tensile stress level is Ⳮ90 MPa. This indicates that the level of tensile stresses in the coatings depends little on a deposition technique and is mainly determined by a mismatch between the TECs of the coating and the superalloy.

Thermal stresses at 20 C in overlay coatings on JS6U superalloy after different technological treatment processes. 1, after deposition (D); 2, after D Ⳮ annealing; 1030 C, 2 h (A); 3, after (DⳭA) Ⳮ peening (P); 4, after (DⳭAⳭP) Ⳮ annealing: 700 C, 2 h; 5, after (DⳭAⳭP) Ⳮ annealing: 800 C, 2 h; 6, after (DⳭAⳭP) Ⳮ annealing: 900 C, 2 h; 7, after (DⳭAⳭP) Ⳮ annealing: 1000 C, 2 h; 8, after (DⳭAⳭP) Ⳮ annealing: 1030 C, 2 h

Fig. 7.9


The next step in the technological process for the coatings deposited by the EB method is a compacting treatment by microball peening, which generates high compressive stresses, whereupon their absolute value exceeds that for the Ni20Cr12AlY coating. This is due to a higher yield stress of this coating at room temperature. Annealing the specimens peened with microballs at a temperature above 700 C completely eliminates coating strain hardening. On annealing of the specimen peened with microballs, all stresses are relaxed and tensile stresses regenerate during cooling. The threshold recrystallization temperature for the Ni20Cr12AlY coating is 1000 C. The processes of stress relaxation in the coating alloyed with cobalt run in another manner, compared with the Ni20Cr12AlY coating. This is because coating strain hardening and its subsequent heat treatment initiate phase transformations and recrystallization. This results in a steadier state, compared to as-deposited and asannealed for the first-time coating. The increase in thermal stresses of the Ni20Co20Cr12AlY coating is caused by a greater mismatch between its TEC and that of the JS6U superalloy, compared with the Ni20Cr12AlY coating and the JS6U superalloy. The stressed Ni20Cr12AlY and Ni20Co20Cr12AlY coatings were x-rayed in the temperature range from room temperature to 900 C. The specimens tested were subjected to a full treatment cycle, including annealing at 1030 C for 2 h, microball peening, and final annealing at 1030 C for 2 h. The results of the investigation are shown in Fig. 7.10. Stresses

Fig. 7.10

Thermal stresses in (1) Ni20Cr12AlY coating and (2) Ni20Co20Cr12AlY coating on JS6U superalloy

were gradually relaxing in both coatings. For the Ni20Cr12AlY coating, the coating-alloy system shows no stresses at 800 C. Further temperature increases to 900 C result in the generation of low compressive stresses (20 MPa). When cooling the alloy from the said temperature to room temperature, stresses recover to their previous level. In the Ni20Co20Cr12AlY coating, the temperature increase results first in the abrupt drop of tensile stresses from 240 MPa at 20 C to 0 MPa at 650 C, then in the generation of low compressive stresses (ⳮ50 MPa); as the temperature approaches 900 C, their absolute value decreases to ⳮ20 MPa. The compressive stress generation in the Ni20Cr12AlY and Ni20Co20Cr12AlY coatings is explained once again by a mismatch between the TECs of the coating and the alloy. The results of the research into thermal stresses using the technique of strain measuring during chemical removal of coatings are shown in Fig. 7.11. The stresses were measured in samples cut out of blades made of the JS6U superalloy and subjected to annealing at 1030 C for 2 h, compacting by microball peening, and further annealing at 1030 C for 2 h. Tensile stresses at a distance of 20 lm from the surface of the Ni20Cr12AlY coating are 100 to 180 MPa. With due account of possible errors introduced by both methods, these results are in good agreement with the results of x-ray diffraction. As the distance from the surface increases, the stresses gradually decrease, and at a distance equal to the coating thickness, the stresses approximate to a zero level. The study of the stressed coating on the blades subjected to testing in an engine for 100 h demonstrates that both the sign and the level of the stresses do not actually change. The experimental methods allow assessment of the stressed state of the coated specimen surfaces. However, these methods are rather timeconsuming and may introduce considerable errors when the changes in chemical and phase compositions take place during high-temperature testing. The method based on etching off coatings allows studying the stressed state only at their room temperature. The use of calculation methods allows the analysis of the coating physical and mechanical properties effect on their thermal stress levels, as well as the assessment of design features of any type and composition of multilayer coatings deposited on substrates of any type. The calcu-


Fig. 7.11

Distribution of thermal stresses at 20 C in Ni20Cr12AlY coating on JS6U superalloy blades

Calculated thermal stresses at 20 C in (1) Ni20Co20Cr12AlY coating and (2) Ni20Cr12AlY coating on JS6U superalloy specimens. hc, distance from a “coating alloy” interface into the coating; hs, distance into the alloy

Fig. 7.12

Fig. 7.13

Forces (P) acting in Ni22Cr12AlY coating on JS6U superalloy versus coating thickness

lation methods are a step toward controlling stresses in coatings and the respective characteristics of their thermal fatigue. Coating-Thermal-Stresses Modeling. To calculate the thermal stresses occurring in the coating-superalloy system, a strained body stressed-state numerical modeling method was used (Ref 12). The specimen taken for model calculation was 2 mm thick and 4 mm long; a coating 0.1 mm thick was applied to one of its sides. The length of 4 mm allows elimination of any edge effect. On the basis of the previously mentioned xray diffraction results, within tolerable limits, the T0 temperature was taken to be 920 C. At the T0 temperature, the stresses at the coating/ superalloy interface approximate to a zero level. Thermal expansion coefficients and modulus of elasticity determined from the investigation of the coating alloys were used for calculations. The calculation results for the stresses occurring in the Ni20Cr12AlY coating/JS6U superalloy and the Ni20Co20Cr12AlY coating/JS6U superalloy systems at their cooling from 920 to 20 C are shown in Fig. 7.12. The calculated stresses in the surface layer of the Ni20Co20Cr12AlY coating are Ⳮ392 MPa, while in the surface layer of the Ni20Cr12AlY coating they are Ⳮ128 MPa. The coated specimen surface stress profile shows that the tensile stresses peak is in the coating zone near the coating/alloy interface. In the vicinity of the interface, the stress sign reverses abruptly, and compressive stresses are observed in the alloy. Beneath the Ni20Co20Cr12AlY coating, compressive stresses reach ⳮ8 MPa; beneath the Ni20Cr12AlY coating they are ⳮ2 MPa. The compressive stresses gradually decrease as the


distance from the coating/alloy interface increases. The diagram showing the value of forces in the coating-alloy system versus coating thickness is presented in Fig. 7.13. The force, P, in the coating was calculated from the following formula: n

Pn ⳱

兺 • ri • Fi i⳱1

Level of thermal stress ratio, r/r0, in (1, 3) coating and in (2) substrate versus (1, 2) coating thermal expansion coefficient variation, D␣, and (3) superalloy elasticity modulus, DE

Fig. 7.14

Calculated thermal stresses in (1) Ni20Co20Cr12AlY and (2) Ni20Cr12AlY coatings on JS6F superalloy specimens. Orientations of (a) [001] and (b) [111]

Fig. 7.15

where ri is stress in the sections into which the coating is divided; Fi is the area affected by stress ri; and n is the number of sections into which the coating is divided. The separation of the stresses and forces acting in the coating has a physical meaning. The stresses in the coating have an effect on the coating material and may cause its failure. The forces in the coating are applied to the coating/ alloy interface and may cause spalling of the coating along this interface. The force in the coating increases monotonically as the coating becomes thicker. Thus, coating thickness is limited by the strength of the coating/coated alloy interface (i.e., by its adhesive strength). The unsatisfactory adhesive strength level may cause spontaneous spallation of a “thick” coating. Spalling from the substrate surface may occur without any stress applied. Superalloy TECs vary slightly, depending on their alloying. Coating TECs can vary within a rather wide range. Coating TECs also vary during high-temperature testing due to the change of the coating chemical and phase compositions. The calculated thermal stresses in the coating and the substrate versus differences of TEC (D␣) and elasticity modulus (DE) for the coating and superalloy are shown in Fig. 7.14. The Ni20Cr12AlY coating/JS6U superalloy system was taken as the basis for calculation. Coating and substrate stresses vary linearly with D␣. As a TEC mismatch between the coating and alloy increases by 20%, the level of thermal stresses in the coating increases 3.7 times. The D␣ has a weaker effect on the stresses in the substrate. The effect of DE on the stresses in the coating is not strong. The calculation results for thermal stresses in single-crystal specimens made of JS6F superalloy with grain orientations of [001] and [111] are given in Fig. 7.15. The Ni20Cr12AlY and Ni20Co20Cr12AlY coatings are deposited on the specimens. In the case of the Ni20Cr12AlY coating, the stress level in the substrate with the orientation of [001] is Ⳮ29 MPa, while in the case of the Ni20Co20Cr12AlY coating it is Ⳮ275 MPa. The stress level in the coatings on the specimens with the orientation of [111] is lower. The main factor, which decides the value of thermal stresses, is a mismatch between the coating and alloy TECs. If this mismatch is not great, as for the Ni20Cr12AlY coating, the variation of elasticity modulus is crucial to the value and sign of thermal stresses. If the mismatch is high,


the use of the material with a high elasticity modulus has a minor effect on the thermal stress level, which decreased by only 5% for the Ni20Co20Cr12AlY coating. The development of multilayer coating systems is very promising and can improve coating stability, for example, in the case of applying the Ni3Al layer or thermodynamically stable carbides (NbC, HfC) to the coating/alloy interface. By using multilayer coatings, it is possible to control the level of thermal stresses in them. The experimental determination of stresses in multilayer systems is a complicated process and, as a rule, it is reduced to measuring total stresses. The results of thermal stress calculations for different modifications of two-layer coatings are shown in Fig. 7.16. Depending on the position in the coating pattern, the layers of the same chemical composition may have different stress signs.

Effect of Coatings on High-Temperature Strength of Superalloys High-temperature strength of the protective coating materials is substantially lower than that of modern superalloys. That is why the coating itself cannot bear the loads applied to the sample during its high-temperature strength testing carried out at temperatures above 700 C. The stresses arising in the coating at the moment of

Calculated thermal stresses at 20 C in two-layer coatings on JS6F superalloy specimens [001]. 1, Ni20Cr12AlY (95 lm)/Ni8Co20Cr12AlY (5 lm); 2, Ni8Co20Cr12AlY (50 lm)/Ni20Cr12AlY (50 lm). hc, distance from a “coating alloy” interface into the coating; hs , distance into the alloy

Fig. 7.16

the sample loading, as a rule, exceed the coating yield strength, so they are relieved and redistributed over the cross section of the superalloy. The stresses in the tested sample should be calculated taking into account the sample cross-sectional area, while the coating thickness is negligible. The nature of the protective coating effect on high-temperature strength of nickel superalloys is closely related with the peculiarities of their failure in an oxidizing atmosphere at high temperatures. For a majority of the superalloys, oxidation at testing promotes the development of surface cracks and contributes to their damaging processes. Application of coatings that protect the surface of such superalloys from oxidation has a favorable effect on high-temperature strength, especially in the case of long-term testing. At the same time, the coating may have a negative effect on high-temperature strength of the superalloy to be protected. This is due to the formation of an interaction zone that reduces the sample cross-sectional area, especially in the case of precipitation of laminated, closely packed phases (such as carbides, r-phase). The scope of testing carried out to evaluate the coating effect on high-temperature strength is great enough. However, the authors of most papers restricted themselves to comparison of the time-to-failure of the coated and uncoated samples. As a rule, in this case it is stated that coating deposition does not reduce sample timeto-failure. Creep-Rupture Strength of Superalloys Tested in Media Provoking Hot Corrosion. Testing for creep and stress-rupture properties in a medium simulating fuel combustion products with elevated sulfur content substantially accelerates the failure processes occurring in the specimen surface layers and their destruction. The testing procedure was as follows (Ref 13, 14). The gas turbine fuel (GZT) synthetic ash was applied from an alcohol suspension to the surface of the standard specimens intended for high-temperature testing. The specimen was 6 mm in diameter, with the working part 30 mm long. The ash layer was 120 g/m2, with the layer application repeated every 500 h of testing. Air supplied at a rate of 20 L/h was forced through the furnace muffle where the specimen was held for testing. The results of the JS6U superalloy testing in air and with a layer of ash on it at 800 and 900 C are shown in Fig. 7.17. The superalloy was considerably damaged during testing where the


Fig. 7.17

Stress-rupture curves of JS6U superalloy in air (1, 䡬) and in GZT ash (2, 䉭). Ni20Co20Cr12AlY coating (䢇) and diffusion coating (Al5Si1.5Y) (䉱) in GZT ash

GZT ash was used. It manifested itself in an intensive time-dependent drop of the superalloy strength. In this case, the strength versus test time is still the same, both under hot corrosion conditions and in the air. It is an exponential equation establishing a relationship of time-tofailure to stress (Ref 13): s ⳱ a • rⳮn

where s is time-to-failure, r is stress, and a and n are constants depending on material properties, corrosive medium, and temperature. At longer test time and higher test temperatures, the effect of the medium on the superalloy is greater, as is shown in Fig. 7.17 (Ref 15). After 100 h of testing at 800 C and 40 to 50 h at 900 C, creep-rupture strength of the JS6U

Table 7.14 Time-to-failure of JS6K superalloy in air and in GZT ash at the same stress level. (Ref 13) Test Stress Time to failure temperature, (r), at testing in air C MPa (sA), h 800 850 900 800 850 900

410 297 179 293 190 112

1000

10,000

Time to failure at testing in GZT ash (sGZT), h 180 165 100 670 280 190

superalloy with the GZT ash starts to decrease. If the initial life of the JS6U superalloy in air is 1000 h, the GZT ash on it will reduce it by 74% at 800 C and by 91% at 900 C. The difference in high-temperature creep-rupture strength levels of the specimens tested in air and in the corrosive media becomes more pronounced as their testing time increases. The same effect can be observed for the JS6K superalloy (Table 7.14). At long-term testing, time-to-failure decreases by a few orders of magnitude. The higher the test temperature is and the smaller the diameter of the specimen tested, the greater the medium effect will be. In addition to shortening superalloy specimen lives under hot corrosion conditions, their initial, secondary, and tertiary creep lives become shorter, too; total ductility (specimen elongation) at failure and ductility at the initial and secondary creep phases are reduced. The JS6U superalloy behavior at testing is typical of all nickel superalloys. Two test-coated JS6U and JS6K superalloys under corrosive conditions, two types of coatings were used: a diffusion coating made of the Al5Si1.5Y alloy deposited by the EA method and an overlay coating made of the Ni20Co20Cr12AlY alloy deposited by the EB method. The coating thickness values were 50 and 100 lm, respectively. The test results on creep-rupture strength of the coated alloys under hot corrosion conditions


are shown in Fig. 7.17. The Ni20Co20Cr12AlY coating on the JS6U and JS6K superalloys guarantees their protection against hot corrosion as well as ensures their long-term strength characteristics similar to those typical of testing in air. The creep-rupture strength level of the coated JS6U superalloy at 2000 h testing at 800 C is 400 MPa, while that for the uncoated specimen is 240 MPa. At 900 C, the coating guarantees the alloy protection for more than 1000 h. In this case, the alloy high-temperature strength level is 220 MPa. At the same time, the uncoated alloy has creep-rupture strength of less than 100 MPa.

Creep rates (V, %/h) of JS6U superalloy at 900 C at (a) initial, (b) second, and (c) tertiary creep stages. 1, Air, no coating; 2, GZT ash, Ni20Co20Cr12AlY coating; 3, GZT ash, diffusion coating (Al5Si1.5Y); 4, GZT ash, no coating

Fig. 7.18

Diffusion coatings demonstrate worse protective properties. However, they reliably protect the alloys against hot corrosion at 800 C for at least 1000 h. The JS6U properties under these test conditions are the same as at its testing in air. At 900 C, the diffusion coating becomes faulty after 200 h, and a loss of creep-rupture strength is observed for the specimens with this coating. However, it is higher than that of the uncoated JS6U and JS6K superalloys. The creep rate for the coated JS6U and JS6K superalloys when tested in the GZT ash is the same as at their testing in air (Fig. 7.18). In addition, the coated specimens do not lose their ductility (Fig. 7.19), as compared with the uncoated specimens tested under the same conditions. The study of the nature of coating failure during testing has not revealed any peculiarities or distinctions from the failure of these coatings without any load applied. The size of the Ni20Co20Cr12AlY coating/JS6U superalloy interaction zone after 3000 h testing at 900 C does not exceed 20 lm. The typical feature of the Ni20Co20Cr12AlY coating detected during testing is its high ductility. At testing, cracking occurs beneath the coating that is strained above the crack but is not faulty. Creep-Rupture Strength of Superalloys Tested in Air. High-temperature testing of superalloys with directional solidification (DS) and single-crystal (SC) structures corroborates the general features revealed of the protectivecoating effect on their high-temperature strength. The results of testing for creep-rupture strength of the directionally solidified JS26-DS superalloy with the Ni20Co20Cr12AlY coating are shown in Fig.7.20. At testing at 1100 C for 1000 h, creep-rupture strength values for the coated and uncoated specimens are 55 and 45 MPa, respectively. The mismatch between creep-rupture strength of the coated and uncoated alloys is much less than that obtained at testing in the GZT ash medium and than could have been expected from the uncoated alloy oxidation depth. No evidence of the expected effect is likely to result from the fact that under high-strain conditions typical of the specimens with directional structures, the surface crack-growth rate exceeds the specimen oxidation rate. Despite coating ductility and protection of the alloy against oxidation, some cracks develop beneath the coating, too (Fig. 7.21). The cracks gradually propagate and cause damage to specimens.


The development of interaction processes between the coating and the alloy leads to a decrease of the specimen cross-sectional area. Dissolution of the c⬘-strengthening phase in the interaction zone beneath the coating reduces the alloy strength. As a result, the working crosssectional area of the JS-type superalloy specimen 5 mm in diameter is reduced by 5 to 7% after 500 h of testing. It is much less than in the case of oxidation effect. Yet, for the blades less than 2 mm thick and bearing high loads, the ef-

Relative elongation of JS6U superalloy at 900 C versus stress at (a) initial, (b) second, and (c) tertiary creep stages. 1, Air, no coating; 2, GZT ash, Ni20Co20Cr12AlY coating; 3, GZT ash, diffusion coating (Al5Si1.5Y); 4, GZT ash, no coating

Fig. 7.19

fect of interaction between the coating and the superalloy should be taken into account. The structural investigations do not reveal differences in the interaction zone between coatings and the superalloys tested under no-load conditions and with high stresses and strains applied. It is explained by the fact that the energy contribution to the diffusion processes due to high temperatures substantially exceeds the energy contribution of the stresses. In all the nickel superalloy specimens tested at long-term static loading, the initial failure zone is located on the specimen surface, which points to similar failure patterns based on crack development and propagation causing destruction at the final creep stage. The coating does not change the failure mechanism; it only neutralizes the effect of oxidation on it. Mechanism of Coating Effect on Superalloy Creep-Rupture. Summarizing the experimental data on the coating effect on high-temperature strength of superalloys, the following conclusion can be drawn. The plot of stress level as a function of time-to-failure that is a straight line has a point corresponding to a certain test time of sc (Fig. 7.22). Starting from this moment, the uncoated specimens start losing their creep-rupture strength due to oxidation effect, surface softening, and reduction of their crosssectional areas. Starting from this moment, creep-rupture strength of the specimens with coatings that protect their surface against oxidation, softening, and respective reduction of their cross-sectional areas is higher in comparison with that of the uncoated specimens. Time of sc is a measure of an uncoated specimen test life. Testing for a longer period should be carried out only on protected specimens. The sc point position depends on heat resistance of the alloy tested, test temperature, environment aggressiveness, and the diameter of the specimen tested. Time of sc for specimens 5 mm in diameter and made of heat-resistant superalloys (such as JS6K, JS6U, JS32, etc.) tested in air at 900 to 1000 C is more than 1000 h. That is why most investigators do not reveal a positive effect of coatings on creep and stress-rupture properties of these alloys, because they reduced their test time to 100 to 500 h. A more aggressive medium, elevated temperatures, smaller specimen diameters, and reduced heat resistance shift the sc point to the left. When testing the JS6U superalloy in the GZT synthetic ash medium at 800


Fig. 7.20

Stress-rupture curves of JS26-DS alloy in air at 1100 C. 1, D, No coating; 2, 䢇, Ni20Co20Cr12AlY coating

and 900 C, the sc time values are 100 and 500 h, respectively.

Effect of Coatings on Fatigue Strength of Superalloys One of the factors that may cause turbineblade failure is blade-material fatigue. Fatigue

cracks usually develop in the blade surface layer in the zone of the greatest alternate stress effect. Blade fatigue strength depends on the surface condition and can vary substantially when a coating with physical and mechanical properties different from those of the superalloy is applied. At testing for fatigue, it is necessary to take into account the fact that the coating effect manifests itself clearly during bend-testing when the maximum strain area is located on the surface. The following factors have an effect on fatigue strength of coated specimens: ● ● ●

● ● ●

Surface microstructure of JS26-DS superalloy sample with Ni20Co20Cr12AlY coating after testing for stress-rupture strength. Applied stress (r) ⳱ 150 MPa; time to rupture (s) ⳱ 22 h. 300⳯

Fig. 7.21

Coating structure (defects, texture, grain size and orientation, and phase composition) Coating thickness and ratio of coating thickness and specimen cross-sectional area Structure of the coating-alloy interaction zone (presence of needlelike precipitate of closely packed phases) Sign and level of thermal stresses in the coating Coating-material fatigue strength Coating surface roughness

Fatigue Strength of Superalloys with Diffusion Coatings. Both thermal stresses generated in diffusion coatings and strength and ductility differ substantially from the respective properties of overlay coatings. As a result, the


effect of these two groups of coatings on superalloy fatigue strength is different. The results of extensive studies of fatigue strength of superalloys with diffusion aluminide coatings revealed the following main relationships. In the temperature range of 20 to 250 C, fatigue strength of the specimens with diffusion coatings is higher than that of unprotected specimens (Ref 16, 17). The increase in fatigue strength depends on the coating chemical composition, the alloy properties, and the thickness of the specimens tested. This effect of diffusion coatings is explained by substantial compressive stresses (ⳮ150 to ⳮ300 MPa) in them. These stresses compensate for the low strength and zero ductility of coatings in the temperature range in question. Testing at high stresses when tensile stresses arise in the surface layer may cause cracking even at the first stages of testing. Such cracking reduces total service lives of the specimens with diffusion aluminide coatings compared to the uncoated alloy in the range of 20 to 250 C. At 900 to 950 C, no deterioration of fatigue strength of the specimens and blade models made of the JS6K and JS6U superalloys and protected with diffusion coatings has been detected compared with unprotected specimens (Ref 18, 19). Oxidizing protected and unprotected specimens before testing under the air atmosphere at 950 C for 1000 h reveals the positive effect of the applied coatings. Depending on the test time, preoxidation of the uncoated samples results in the reduction of the JS6K superalloy sample fatigue lives by 10 to 17%. Fatigue lives of the samples with the diffusion coatings also de-

crease, although this reduction is 5 to 12% and also depends on the test time. Thus, fatigue lives of the coated samples are higher than those of uncoated samples by 6 to 9%. Similar results were obtained on testing the blades. Reduction of sample fatigue lives at longterm testing in the high-temperature range under oxidizing atmospheres results from two processes: changes in superalloy structure and sample surface oxidation. For the uncoated sample, its fatigue life drop is the consequence of more intensive oxidation of its surface and softening of its surface layer. The same effect of diffusion coatings has been revealed at testing in an aggressive medium of fuel combustion products (Ref 20). Test results for blade models with edge radii of 0.5 mm and made of the Ni10CrWMoCo alloy are presented in Fig. 7.23. The tests were carried out under conditions of asymmetrical bending for the samples supported as a cantilever (cycle stress range rm ⳱ 200 MPa) in the T-1 aircraft fuel combustion products. The test temperature was 900 C, its cycle vibration frequency was 50 Hz, and test time provided for 106 to 107 cycles. The diffusion coating was applied to the samples from the powder mixture of 98% Al-Fe alloy (50% Al) and 2% NH4Cl at 950 C. The coating thickness was 50 lm; its aluminum content was 34 to 36%. In corrosive media, diffusion coatings on blade models guarantee a fatigue-strength level 25 to 30% higher than that of the uncoated specimens. The JS6K and JS6U superalloys showed the same results. Favorable outcome of fatiguestrength testing of blades with diffusion coatings

Fatigue strength of blade models made of JS6U superalloy. Testing in combustion products of T-1 aircraft fuel at 900 C. 1, diffusion-coated samples; 2, uncoated samples

Fig. 7.23 Fig. 7.22

Diagram illustrating the effect of coatings on creep-rupture strength of superalloys


was also observed after their service in engines (Ref 21). Fatigue Strength of Superalloys with Overlay Coatings. Fatigue-strength testing of superalloys with different modifications of overlay coatings in the temperature range of 20 to 1100 C (Ref 22, 23) has revealed the following relations. At temperatures lower than 500 C, tensile stresses arise in the surface layers of the specimens with overlay coatings. Their values depend on the coating chemical composition. Ultimate and yield strength values of overlay coatings are rather high in this temperature range. They are higher than the stresses applied to the specimens during testing. Under these conditions, no increase in fatigue strength, which is typical of diffusion coatings, is observed. When testing round, standard specimens 8 mm in diameter with the Ni20Cr12AlY coating 60 lm thick under pure bending conditions at 20 C, no decrease in fatigue strength is detected (Table 7.15). As a rule, a single crack that causes failure (Fig. 7.24) appears on the specimen surface. The site of failure is located on the coating surface. The same results were obtained on the specimens made of the JS6F, JS26-DS, and JS32-DS superalloys. At 20 C, an insignificant rise in fatigue strength is noted on round specimens made of the VJL12U superalloy with the working part 3 mm in diameter and the Ni20Cr12AlY overlay coating 100 lm thick (Fig. 7.25). It may result from the fact that on testing the specimens 3 mm in diameter, their stresses were calculated by ignoring coating thickness contribution to their cross-sectional areas. Fatigue-strength testing of finished blades at 20 C is a mandatory check operation in bladelot manufacturing at the plants. Heat treatment of the blades and machining of their surfaces have an important effect on the reference values of blade fatigue strength (Table 7.16). Fatigue life of the blade with a diffusion coating at the

Table 7.15 Fatigue strength, rⴑ1, of JS6U superalloy specimens 7 mm in diameter at 20 and 950 C rⴑ1, MPa (N ⴔ 2 • 107 cycles) 20 C

Coating Uncoated Ni20Cr12AlY Ni20Co20Cr12AlY

190 … 190 (d ⳱ 60 lm)

950 C 280 280 (d ⳱ 100 lm) 300 (d ⳱ 120 lm)

test stress of 180 MPa is more than N ⳱ 2 • 107 cycles. When the Ni20Cr12AlY overlay coating is applied instead of a diffusion coating, reduction of fatigue life to N ⳱ 6 • 106 cycles is observed. Peening with microballs and generation of high-level compressive stresses did not extend blade lives, although the site of failure shifted to their inner-cavity surface. The initial blade-fatigue life was obtained again after reducing the temporary annealing temperature to 1000 C. At 20 C, the defects of an overlay coating, such as intercrystalline porosity, microdrops, and so on that cause stress concentration, are the sites of fatigue-crack nucleation. These blade defects in the zones of high alternate stresses are prohibitive. At temperatures above 700 C, overlay coatings are free from stresses or show low compressive stresses. During fatigue-strength testing, the stress level on the specimen surface considerably exceeds coating yield strength and ultimate strength. The effect of combined strains, depending on the superalloy mechanical properties and test conditions, causes an accelerated accumulation of ductile strain followed by a respective accumulation of defects. The process is more intensive in the coating than in the bulk superalloy. It results in the formation of fatigue cracks and, as a rule, the cracks are not longer than the coating thickness. The results of fatigue strength testing of the JS6U superalloy with the Ni20Co20Cr12AlY and Ni20Cr12AlY coating at 950 C are given in Table 7.15. Fatigue strength tested on the basis of 2 • 107 cycles for the coated samples was the same as fatigue strength for the uncoated samples. The lack of a wholesome effect of the overlay coating is due to the fact that the rather short test time of 120 h does not allow the effect of oxidation that reduces fatigue strength of uncoated samples to be noticed. Similar results were obtained for other superalloys. On testing the coated samples 3 mm in diameter at 950 C, a slight increase in fatigue strength was observed as compared with uncoated samples (Fig. 7.25). The salient feature of high-temperature testing of specimens with diffusion and overlay coatings is the presence of numerous cracks on their surfaces (Fig. 7.26). These cracks are stress concentrators and cause formation of numerous sites of fatigue failure beneath the coating. Further crack propagation depends on the failure mechanism typical of the alloy tested at high


temperatures and on the specimen shape. Numerous surface sites merge and form a circular crack, thereby reducing the specimen cross-sectional area. However, this reduction of the crosssectional area is insignificant for specimens 7 mm in diameter and does not affect their fatigue strength. The main site causing failure of the round specimens is located at a distance of 0.2 to 0.8 mm from the surface (Fig. 7.26c). Studies of fatigue strength of superalloys with EB thermal barrier coatings (TBC) (the coating design is as follows: inner layer/intermediate layer/outer layer of Ni20Cr6AlY/ Ni20Cr12AlY/ZrO2-8%Y2O3, 35/45/60 lm thick, respectively) were carried out for two types of specimens. Round SC specimens 7 mm in diameter and made of the JS32-DS superalloy were tested under clear-bending conditions at 1100 C. Flat specimens with thickness varying from 2 to 5 mm and made of the JS6U superalloy were bend-tested at 975 C. Bending vibrations were exhibited in a magnetostrictor at a frequency of 3200 Hz. The study of the nature of alloy round sample failure has confirmed the general features stated previously. The application of a TBC to a sample surface results (as has been observed for the Ni20Cr12AlY coating) in nucleating numerous sites of failure, followed by cracks propagating into the coating and further into the alloy. However, the main site of nucleation of the crack causing a sample failure is located beneath the coating surface at a depth of 1 to 1.5 mm, and it does not result from the coating effect. At 1100 C, fatigue strength (on the basis of N ⳱ 2 • 107 cycles) was rⳮ1 ⳱ 150 MPa. The pres-

Fig. 7.24

ence of a ceramic layer changed neither the nature of sample failure nor fatigue strength values. The situation changes on testing sheet (flat) samples (Fig. 7.27). In this case, the surface site of failure forms in the TBC in the maximum stress zone in the corner of the flat sample. The corner is a stress concentrator; it is the site from where the fatigue crack causing failure propagates. In this case, fatigue strength after N ⳱ 1 • 108 cycles diminishes by 30 MPa in comparison with unprotected samples, and the TBC has the same effect as the Ni20Cr12AlY coating lacking a ceramic layer. Mechanism of Coating Effect on Superalloy Fatigue Strength. Fatigue-strength tests showed that the effect of overlay coatings on this superalloy property was different in the highand low-temperature ranges. At room temperature, the level of coating strength and respective fatigue strength approaches that of superalloys. That is why, despite tensile thermal stresses in the coating, sample and blade fatigue strength values do not actually change when overlay coatings are deposited. On testing small-sized (ⱕ3 mm) samples, deposition of a coating 100 lm thick, which has high strength characteristics, can result in upgrading the level of fatigue strength. On bend-testing in the low-temperature range, coating defects act as stress concentrators. If the coating is not free from defects, the sample and blade fatigue strength values may decrease. At temperatures above 800 C, overlay coating strength and respective fatigue strength are substantially lower than superalloy fatigue

(a) Appearance and (b) fracture of JS6U superalloy specimen with Ni20Cr12AlY coating. Specimen tested at 20 C. r ⳱ 200 MPa; N ⳱ 1.8 • 107 cycles


Fatigue strength of VJL12U superalloy (3 mm bars). 1 and 2, 20 C; 3, 700 C; 4 and 5, 900 C; 1 and 4, uncoated; 2, 3, and 5, Ni20Cr12AlY coating

Fig. 7.25

strength. The coating quickly accumulates defects, and numerous fatigue cracks, which are effective stress concentrators, develop in it. In addition, the cracks open the access of oxygen to the alloy surface. Further sample damage depends on the temperature, alloy heat resistance, sample shape, and medium aggressiveness. The amount of oxygen passing through the cracks to the alloy surface is much less than if there is no coating at all. The efficiency of the coating cracks as stress concentrators, taking into account the high ductility of the coating, does not exceed the efficiency of stress concentrators formed due to oxidation. In addition, when testing round samples under clear-bend conditions, a site of failure is formed, as a rule, beneath the surface and inside the sample. That is why on testing in the air at

900 to 1000 C for N ⳱ 2 • 107 cycles (120 h), the values of fatigue strength for coated and uncoated samples are the same. In the case of testing in aggressive media, the coated samples have higher fatigue strength in comparison with uncoated samples. The fatigue-strength testing procedure has much in common with testing for creep-rupture strength. A certain number of cycles, Nc, corresponds to the test time of sc (Fig. 7.22). On a lapse of these cycles, surface damage caused by oxidation and hot corrosion results in deterioration of uncoated sample fatigue strength. The values of fatigue strength for coated samples are higher than those for uncoated ones after the said number of cycles. At temperatures of 900 to 1000 C, heat resistance of the JS6U, JS6K, and VJL12U superalloys is rather high, and, in fact, their surfaces remain free from oxidation products after the said N ⳱ 2 • 107 cycles (120 h). That is why no effect of the coating on fatigue strength has been noticed in the tests carried out. However, when blades are used in turbines, the degree of their surface damage substantially exceeds the degree of sample surface damage at their testing for fatigue strength. That is why fatigue strength of the uncoated blades used in the engines is always lower than that of the coated blades.

Effect of Coatings on Thermal Fatigue of Superalloys The blade designs featuring the most-effective cooling systems, the increase in temperature gradients through the blade-wall thickness, and the nonuniform temperature fields over the blade surfaces have resulted in the growth of thermal fatigue cracking events in the coatings. That is why the study of coating effect on superalloy resistance to failure during their heating and

Table 7.16

Fatigue strength of JS6U superalloy blades at 20 C

Stresses on leading edge

Number of test cycles before cracking

180 MPa

2 • 107

180 MPa

6 • 106

180 MPa

(8 ⳰ 15) • 106

180 MPa

2 • 107

Location of cracks Inner cavity at leading edge where stresses are accumulated Outer surface of loading edge where it develops into shroud Inner cavity at leading edge where stresses are accumulated Cracks are not detected

Type and characteristics of coating and technological process Diffusion coating, 40 lm thick, annealing: 1000 C, 4 h Ni20Cr12AlY coating, 100 lm thick, annealing: 1040 C, 5 h Ni20Cr12AlY coating 100 lm thick, annealing: 1040 C, 5 h; peening Ni20Cr12AlY coating 100 lm thick, annealing: 1000 C, 5 h


Fig. 7.26

(a) Appearance, (b) microstructure, and (c) fracture of JS6U superalloy specimen with Ni20Cr12AlY coating. Specimen tested at 950 C. r ⳱ 300 MPa, N ⳱ 2.4 • 107 cycles


cooling cycles is an important point in the investigation of the properties of the coatings designed to protect turbine blades. Numerous publications deal with this problem. A complicated dependence of thermal fatigue on test rig parameters and on the structural design of the blades and test samples, as well as the lack of a single-valued criterion for thermal fatigue estimation, to some extent, make interpretation of numerous test results more complicated. Resistance of blades and samples to thermal fatigue depends on mechanical and thermal-physical parameters of the materials, such as strength, ductility, elasticity modulus, thermal expansion coefficient, and thermal conductivity. The test conditions, which have a great effect on thermal fatigue characteristics, are as follows: rates of heating and cooling, cycle maximum and minimum temperatures, holding time at the maximum temperature, and static loads applied. During thermofatigue testing of the blades and samples, their surfaces interact with gaseous atmosphere. As a result of this interaction, the top layers of the metal are depleted of alloying elements. This reduces a fraction of strengthening phases in them and degrades their hightemperature strength. The changes in the chemical composition of the surface zone and the respective variations of thermal expansion coefficients result in an additional increasing of thermal stresses. They can reach a level considerably exceeding the surface zone strength. Thermal fatigue cracking in the surface zone followed by crack propagation into the base material resulted in test sample failure. The presence of heat-resistant coatings on sample or blade surfaces, which protect the base

Fig. 7.27

Fatigue strength of JS6U superalloy (1) without any coating and (2) with EB-TBC

material against oxidation and softening, is expected to have a favorable effect on resistance to thermal fatigue. In this case, physical-mechanical and thermophysical characteristics of the coating have a substantial effect on the test results. As a rule, rigidity of coatings (Ec • h) is much less than that of a blade wall or a sample (E • H): Ec • h 0.1 E • H

where Ec and E are the coating and superalloy elasticity modulus, respectively, and h and H are the thickness of a coating and a blade wall or a sample. Under these conditions, even at slow cooling or heating of the sample with the coating deposited at the temperature of T0, thermal stresses arise in the coating. Under the conditions of long-term cycling and accumulation of the strain induced by thermal stresses, thermal fatigue cracking may occur even if no additional stresses are generated by impressed forces. Thermal stresses may vary in the coating if a temperature gradient in the coating/blade-wall zone arises. It is always present in cooled blades and substantially rises at turbine transient power settings. If the sample or blade construction is stressed or strained because of their nonuniform temperature pattern along its full height and surface, coating thermal stresses are added to structural thermal stresses. As a rule, thermal fatigue tests are carried out on gas-dynamic test rigs at alternating heatingcooling cycles. Test results depend on specimen shape, stiffness and rigidity, maximum and minimum cycle temperatures, rates of heating and cooling, and gas parameters and composition. Under these test conditions, thermal stress and strain values in the specimens may be controlled by specimen rigidity (e.g., solid or hollow specimens, etc.), rates of heating and cooling, and maximum and minimum cycle temperatures. Thermal fatigue is characterized by the number of cycles to appearance of the crack of the specified size on the specimen surface. This characteristic, to a considerable extent, is a qualitative one. It may be used for revealing the advantages of a certain blade construction or a certain superalloy or protective coating. However, this characteristic cannot be used for blade service-life calculations. In addition, as a rule, gasdynamic test rigs do not simulate heat flows typ-


ical of the engines. That is why the results of the rig tests cannot reveal the features of coating behavior on the cooled blades at very high rates of their heating and cooling that occur under transition working conditions in modern engines. Proceeding from extensive testing results for specimens with different aluminide coatings, it is noteworthy that aluminum content of the coating outer zone and the coating thickness are the main factors that have an effect on the test results. The reduced aluminum content of the coating involving the respective reduction of the NiAl compound ductile-brittle transition temperature improved thermal fatigue resistance of the specimens and blades with diffusion aluminide coatings. The reduction of coating thickness had the same effect.

Thermal fatigue resistance of JS6K superalloy blades versus maximum cycle temperature. 1, Uncoated; 2, diffusion coating of aluminum, Al2O3, and NH4Cl mixture; 3, diffusion coating of aluminum, chromium, Al2O3, and NH4Cl mixture. Source: Ref 24

Fig. 7.28

Fig. 7.29

Schematic diagram of test rig for thermal fatigue tests

On setting up the program for testing specimens with diffusion coatings, it is necessary to take into account some specific coating features, such as low ductility of the coatings with the aluminum content of more than 32% and gradual aluminum content reduction during testing, followed by the increase in coating ductility. Crack nucleation may take place in a diffusion coating with high-aluminum content even after the first few heating-cooling cycles at testing for thermal fatigue under conditions of high thermal stresses and respectively high strains. On running the same tests for the coated specimens with low-aluminum content of the coating, their cracking occurs much later than in uncoated specimens. Typical results of testing for thermal fatigue of aluminized blades made of the JS6K superalloy on the gas-dynamic test rig are presented in Fig. 7.28 (Ref 24, 25). The maximum gas temperature available in the test rig is 1500 C, the time of heating to the preset temperature is 8 s, and the cooling time is 17 s. The maximum test temperatures were 1000, 950, 900, and 850 C. The minimum cycle temperature was 300 C. After every 100 cycles, the specimens were examined visually for crack detection. Under the conditions of the tests carried out, the aluminized coatings increased thermal fatigue resistance. Under the conditions of 850 ⇔ 300 C cycling, aluminized blades withstood 4 times as many cycles to crack nucleation as uncoated blades. The increase in the maximum cycle temperature followed by the respective increase in thermal stress and strain values resulted in less difference in thermal fatigue resistance of coated and uncoated blades. The addition of salt to the gas flow under the same test conditions gives rise to hot corrosion and reduces blade service life considerably. However, in all cases, the service lives of the blades with aluminized coatings are longer than those of uncoated blades. Experimental Studies of Coated Superalloy Thermal Fatigue. A large body of information on comparison tests for specimens and blades with different coatings is received using the gasdynamic test rig. Its layout is given in Fig. 7.29. Knife-edged solid and hollow samples with an edge radius of 1 mm (Fig. 7.30) are used for testing. Eight samples are simultaneously fastened in the special fixtures of the test rig. Then, they are moved in turn into a combustion chamber, where they are heated to 1000 or 1100 C,


and to compressed air supply systems, where they are cooled to 200 C. The combustion chamber uses aircraft fuel. The preset program may change the cycle duration. Two options of the cycle were used for testing: ● ●

sheat ⳱ scool ⳱ 30 s sheat ⳱ scool ⳱ 30 s, holding time at the cycle maximum temperature of 60 s

The second option was intended to determine coating resistance to thermal fatigue at surface

Fig. 7.30

Table 7.17 No.

Sketch of specimens for thermal fatigue tests. R1, radius equal to 1 mm

Test conditions

Temperatures

Time

1

1000 C ↔ 200 C

2

1000 C ↔ 200 C

3

1100 C ↔ 200 C

4

1200 C ↔ 200 C

5

1000 C ↔ 200 C

Time required for heating the sample to the maximum temperature is 30 s; cooling time is 30 s Time required for heating the sample to the maximum temperature is 30 s; holding time at the maximum temperature is 60 s; cooling time is 30 s Time required for heating the sample to the maximum temperature is 60 s; cooling time is 60 s Time required for heating the sample to the maximum temperature is 60 s; cooling time is 60 s Time required for heating the sample to the maximum temperature is 120 s; cooling time is 120 s

exposure to severe conditions caused by oxidation. Within the first seconds of heating hot gases from the combustion chamber, heat the sample central portion that is not large (20 to 30 mm). It gives rise to a temperature gradient through its full length and thickness. This gradient induces thermal stresses and strains, which peak on the sample edge with a radius of 1 mm that faces the flow of heating and cooling gases. Strain accumulating in this place causes thermofatigue cracking of the sample. The number of cycles to development of a crack 0.5 mm long on the edge of the sample is its life criterion. Test conditions used are given in Table 7.17. The test results are given in Table 7.18 and shown in Fig. 7.31. When testing uncoated samples of the JS6U superalloy at the cycle maximum temperature of 1000 C, accelerated oxidation of their working edges and surfaces is observed. Numerous cracks nucleate in these zones. Protective coatings of all compositions protect the superalloy surfaces against oxidation and softening and, thus, increase the number of cycles to thermal fatigue cracking. A narrower range of strains arising on a hollow sample in comparison with the solid one leads to a cyclic life of 1.2 times as long. The sample holding time at the cycle maximum temperature has a substantial effect on the test results. The development of creep at a high temperature and oxidation of coating surfaces encourages accumulation of strains and facilitates a failure of both the coating and the superalloy. The samples tested at a holding time at the maximum temperature of 60 s feature cyclic lives 1.5 to 3 times shorter in comparison with those of the samples tested without the hightemperature holding. Under long-term cycling conditions (120 s), the distinctions caused by sample physical-mechanical characteristics diminish. Cyclic life of a JS6U superalloy sample with an overlay coating is 1.2 times longer than that of a sample with a diffusion coating, while at testing under short-term cycling conditions, the respective parameter is 2.2. Thermal fatigue of the coated samples, which are thermally stressed on the test rig under consideration, is closely related with both high-temperature properties of the alloys and sample fabrication techniques. The test results for the samples made of the JS6U superalloy (equiaxial solidification), the JS6F-DS superalloy (direc-


tional solidification), and the JS30-SC (single crystal) with the diffusion aluminized coating deposited from a mixture of 98% of the Al-Fe (36% Al) alloy and 2% NH4Cl are presented in Table 7.18. The aluminum content of the coating outer zone is 22 to 24%; the coating is 30 to 50 lm thick. According to their resistance to thermal fatigue, the alloys are ranged as follows: JS6U (1), JS6F-DS (1.35), and JS30-SC (1.45). Better ductility and strength of coatings as well as a smaller mismatch between TECs of a coating and a superalloy increase sample service lives and reduce the distinction of thermal fatigue for superalloys with different structures. The same alloys tested with the Ni20Cr12AlY/ Ni10Co20Cr12AlY coatings 100 lm thick deposited by EB technique have the same cyclic lives: JS6U (1), JS6F-DS (1), and JS30-SC (1). This is explained by the fact that long-term cycling gives rise to the accumulation of damages not in a superalloy but in a coating, in the zone of maximum oxidation and thermal strain effect. On testing the samples made of the JS6U superalloy at long-term cycling (120 s) for their resistance to thermal fatigue, the coatings are ranged as follows: aluminized coating (1),

(Al5Si1.5Y) diffusion coating (1.10), Ni20Cr12AlY coating (1.20), and Ni20Co20Cr12AlY coating (1.30). The same testing under short-term cycling conditions (60 s) gives the following ranges: aluminized coating (1), Ni20Cr12AlY coating (2.25), and Ni20Cr12AlY/Ni10Co20Cr12AlY coating (2.40). Longer cyclic lives of the samples with the Ni20Cr12AlY/Ni10Co20Cr12AlY coating are due to their better surface roughness characteristics and fewer number of defects in the coating structure when the two-stage surface preparation technique is used, which is described in Chapter 3, “Phase Composition of Coatings on Superalloys.” The surface study of the samples tested on a gas-dynamic test rig has shown that the nature of their failures is similar in all cases; that is, numerous thermal fatigue cracks develop on the working edge at a distance of 0.6 to 1.0 mm from one another, and the crack edges are open and oxidized. Depending on coating ductility, the cracks either propagate perpendicularly to the surface or their abundant branching is observed in the coating.

Table 7.18 Number of cycles (N) to cracking for hollow samples at their testing for thermal fatigue on 8UTS test rig Superalloy

Coating

Deposition method

Thickness, lm

Test condition

NC

JS6U JS6U(a) JS6U JS6U(a) JS6U JS6U JS6U JS6U JS6U JS6U JS6U JS6U JS30-SC JS30-SC JS30-SC JS6F-DS JS6F-DS JS6F-DS VJL12U VJL12U VJL12U VJL12U JS26-DS(a) JS30-SC(a) JS30-SC(a) JS30-SC(a) JS32-DS(a) JS32-DS(a)

Uncoated Uncoated Aluminized Aluminized Aluminized Al5SiY Ni20Cr12AlY Ni20Cr12AlY Ni20Cr12AlY Ni20Cr12AlY/Ni10Co20Cr12AlY Ni20Co20Cr12AlY Ni20Co20Cr12AlY Uncoated Aluminized Ni20Cr12AlY/Ni10Co20Cr12AlY Uncoated Aluminized Ni20Cr12AlY/Ni10Co20Cr12AlY Al5SiY Ni20Cr12AlY Ni20Cr12AlY Ni20Co20Cr12AlY Ni20Cr12AlY/Al Uncoated Aluminized TBC Uncoated TBC

… … Aluminizing Aluminizing Aluminizing EA EB EB EB EB EB EB … Aluminizing EB … Aluminizing EB EA EB EB EB EB/aluminizing … Aluminizing EB … EB

… … 40 40 40 40 100 50 100 20/80 100 100 … 40 40/60 … 40 40/70 40 50 100 100 100 … 50 180 … 180

1 2 1 2 2 2 1 2 2 1 2 5 1 1 1 1 1 1 2 2 2 5 4 3 3 3 3 3

1190 580 1610 830 1010 1080 3680 1160 1210 4720 1310 1330 2200 2420 4030 1900 2260 4000 700 740 860 870 620 1190 1300 1530 1320 1510

(a) Solid samples


The maximum range of thermal stresses and strains in a coating at lab testing can be reproduced on rapid cooling of heated samples in cold water. In this case, the range of strains in a coating induced by a temperature gradient between the surface and the cross-sectional area of a sample wall increases. Such tests reveal the extreme manifestation of thermal stresses that is a thermal shock. They allow the comparison of coatings by their ability to realize deformation within the microzone of a coating. The typical feature of thermal shock testing is a formation of a network of small cracks in the coating. Their length does not exceed coating thickness. Cracks nucleate on cooling. At that period, the strains exceeding coating ductility arise in surface layers. The results of thermal shock tests carried out under 1050 C ↔ cold water conditions are given in Table 7.19. At such testing, an aluminized coating is fragmentized into a network of minor cracks after 8 to 10 cycles. The Ni20Cr12AlY coating is much more stable. It shows a network of cracks after 40 to 50 cycles. The addition of a highly ductile

sublayer results in deceleration of crack growth and in changing the mechanism of their propagation from brittle to ductile. The advantages of the Ni20Cr12AlY coating tested under these conditions are due to a better ratio of TECs of an alloy and a coating and a respectively lower level of thermal stresses. Thermal fatigue tests of coated samples have revealed that the coatings may both raise and reduce superalloy thermal fatigue. When carrying out short-term tests in air, which did not result in unprotected sample surface damage by oxidation, their lives could be longer than those of the samples with protective coatings. The detrimental effect of coatings is caused by a wider range of strains on a coated alloy surface due to additional thermal stresses in the coating and lower mechanical properties of the coating in comparison with those of the superalloy. Similar to testing for creep-rupture strength, the situation changes at long-term testing or at testing in aggressive media, which prompt surface damage of unprotected samples.

Number of cycles to cracking caused by thermal fatigue of samples versus types of superalloys and coatings: JS6U superalloy (1, 2, 3, 4); JS30-SC superalloy (5, 6, 7); JS6F-DS superalloy (8, 9, 10). 1, 5, 8, Uncoated superalloy; 2, 6, 9, superalloy with aluminized coating; 3, superalloy with Ni20Cr12AlY coating; 4, 7, 10, superalloy with Ni20Cr12AlY/Ni10Co20Cr12AlY coating

Fig. 7.31

Table 7.19 Alloy JS6F-DS JS6U-DS JS6F-SC JS6F-DS JS6F-DS JS6F-DS

Thermal shock testing of different-type coatings. Test conditions: 1050 C ↔ water Coating

Thickness, lm

Deposition method

Number of cycles before cracking of coating

Coating effect factor, K ⴔ N/NAl

Aluminized Aluminized Aluminized Ni20Cr12AlY Ni20Co20Cr12AlY Ni20Cr12AlY/Ni10Co20Cr12AlY

40 40 40 100 100 m 40/60

Aluminizing Aluminizing Aluminizing EB EB EB

8 10 10 50 ⳰ 60 30 ⳰ 40 50 ⳰ 70

… … … 5⳰6 3⳰4 5⳰7


Distinctions in service lives of the samples with different coating compositions are caused by a number of factors, such as the level of thermal stresses arising in the coating, its ductility, and the strength and protective properties. The efficiency of any coating type manifests itself, to a greater extent, at high thermal stresses in a coating, which are generated by high rates of heating and cooling. Coating Effect on Superalloy Low-Cycle Fatigue. Low-cycle fatigue tests are an efficient technique for evaluation of coating resistance to cracking caused by high strains typical of thermal loading. Flat samples 1 mm thick of the VJL12U superalloy with coatings of different compositions were tested for low-cycle fatigue by bending at 20 and 800 C. At 20 C, all types of coatings tested have ductility less than 1% and the maximum thermal stress level. At 800 C, all coatings are ductile and thermal stresses are not high in them. An aluminized coating deposited from a mixture of 98% Al-Fe alloy (36% Al) and 2% NH4Cl and the Ni20Cr12AlY and Ni20Cr12AlY/Ni10Co20Cr12AlY overlay

coatings 100 mm thick deposited by EB technique have been tested. The test results are presented in Fig. 7.32 and given in Table 7.20. At 20 C, all types of protective coatings reduce life of the VJL12U superalloy. This feature is most conspicuous in the samples with the Ni20Cr12AlY coating and caused by defects (microdots) in them, which are stress concentrators. A two-layer coating affects their lives to a lesser extent. Unlike overlay coatings, diffusion coatings are under compressive stresses at 20 C. However, despite it, they also shorten lives of the samples made of the VJL12U superalloy 5 to 7 times. During testing (100 h) at 800 C, no failure of the coating surface caused by oxidation occurs. Under these conditions, the coatings do not exhibit their protective properties, and accumulation of strains in them due to low yield strength enhances cracking and sample failure. Overlay coatings having high strength and ductility margin at 800 C shorten sample lives to a lesser extent in comparison with diffusion coatings (Fig. 7.32). The lives of coated and uncoated samples differ less and less and have a tendency of becoming equal if cyclic testing lasts long enough. The relationships derived for the VJL12U superalloy can be extended to other superalloys. It is worth determining coating material lowcycle fatigue separately from that of a superalloy and comparing these results with coated sample test results. A batch of samples was fabricated from the Ni20Cr12AlY alloy condensate produced by EB evaporation technique. The samples were 0.6 to 0.8 mm thick. Low-cycle fatigue testing was carried out by tension at 950 C under two sets of test conditions: ●

On testing under the first set of test conditions, certain loads were cyclically applied to the samples (4 cycles/min). Their levels neither varied nor depended on sample strains. ● On testing under the second set of test conditions, samples were cyclically strained. The strain range value, De, did not vary at testing.

Low-cycle fatigue of VJL12U superalloy at (a) 20 C and at (b) 800 C. 1, Uncoated; 2, aluminized coating; 3, Ni20Cr12AlY coating

Fig. 7.32

The test results are presented in Fig. 7.33(a) and (b) in comparison with the test results for the samples made of the JS26-DS superalloy, both uncoated and coated by the EA method with the Ni20Cr12AlY alloy 100 lm thick. As a result of testing under the first and second sets of test conditions, lives of coated samples are longer than those of uncoated ones. This is due to the fact that coating ductility is very


high at 950 C. A coating material sample tested under the first set of test conditions is shown in Fig. 7.34. Its ductility exceeds 100%. In both cases, the effect of the coating application is due to the JS26-DS superalloy surface protection against oxidation. Under the first set of test conditions, when high mechanical properties are most conspicuous, the coated superalloy life is much longer than the coating material life. The conditions for loading a coating on a superalloy sample are fully dictated by the mechanical properties of a superalloy. Under these conditions, the coating accumulates failures much slower than a coating material tested separately does. Under the second set of test conditions, in the certain range of strain range, a coating material has a longer life. This is also due to high ductility of the coating material at 950 C. Failures in the superalloy that features lower ductility are

accumulated much faster. Perhaps, coating failure occurs after cracking of the superalloy.

Coating Effect on Thermomechanical Fatigue of Superalloys Basic Principles and Methods. The most detailed information on coating properties under conditions similar to blade service conditions can be obtained from thermomechanical fatigue (TMF) tests. For nickel-base superalloys used as turbine blade materials, the main TMF test types are load-adjusted thermomechanical fatigue test (LA-TMF test) and strain-control thermomechanical fatigue test (SC-TMF test) (Ref 25–27). At LA-TMF testing, a constant stress range of Dr ⳱ const (gentle loading) is preset, while at SC-TMF testing, a constant mechanical strain

Table 7.20 Experimental K and C in equation of De • N k ⴔ C characterizing low-cycle fatigue of VJL12U superalloy with coatings of different types Test temperature 20 C

800 C

Coating

Thickness, lm

Deposition method

K

C

Uncoated Aluminum Ni20Cr12AlY Ni20Cr12AlY/Ni10Co20Cr12AlY Uncoated Aluminum Ni20Cr12AlY

… 40 100 100 … 40 100

… Aluminizing EB EB … Aluminizing EB

0.119 0.215 0.127 0.129 0.144 0.126 0.131

3.11 4.99 2.15 2.58 2.97 2.34 2.56

Results of low-cycle fatigue testing of JS26-DS superalloy and vapor-deposited samples under (a) the first and (b) the second set of test conditions. 1, Uncoated superalloy; 2, superalloy with Ni20Cr12AlY coating; 3, samples made of condensate alloy Ni20Cr12AlY

Fig. 7.33


range of De ⳱ const (severe loading) is usually held. At elastic deformation of a sample, both types of testing are, in fact, similar and thus yield similar outcome. Each of the previously mentioned TMF tests allows only approximate simulation of bladematerial deformation processes under service conditions, because in a real blade, complicated cyclic deformation takes place. This deformation is a combination of gentle loading caused by centrifugal and gas loads and of severe loading caused by thermal stresses. It is considered that severe cyclic loading that occurs at SC-TMF test is most similar to the real deformation conditions of a blade airfoil material. However, this statement is true under the conditions of elasticplastic cyclic deformation of a blade material only when thermal stresses in a blade airfoil are noticeably higher than the mechanical stresses

caused by centrifugal and gas loads. In addition, the irreversible creep strains take place in the blade airfoil under service conditions. This results in additional complication of a real blade material deformation pattern. Both of the previously mentioned test methods are used for experimental investigations of TMF of overlay coatings deposited on turbine blades. In the case of experimental research into TMF of the coatings, the following factors typical of turbine blade coatings under their real deformation and failure conditions should be taken into consideration (Ref 28, 29): ●

●

●

●

●

●

Appearance of vapor-deposited Ni20Cr12AlY sample after testing under the first set of test conditions. 1, Initial sample; 2, r ⳱ 55 MPa, d ⳱ 320%; 3, r ⳱ 25 MPa, d ⳱ 240%, N ⳱ 2320 cycles; 4, r ⳱ 15 MPa, d ⳱ 500%, N ⳱ 5180 cycles

Fig. 7.34

Basically, thermal fatigue cracking starts in the coating and then penetrates into the substrate. This is very typical for the SC blades cast with the primary crystallographic orientation 001 because of the low elasticity modulus of material and, therefore, low thermal strain ranges for thermal cyclic loading. It is necessary to take into account the twoaxial stress state in the coating during analysis of the TMF test results of coated bars. Cracking of the coating on the cylindrical bar during TMF testing can be both diametrical and longitudinal. The stress-strain state of the coating cannot be directly measured by existing control methods during TMF testing of coated bars. For analysis of nonelastic straining of the coating under these conditions, it is necessary to use suitable constitutive models. Strains in coatings are basically dependent on the difference between TECs of the substrate and the coating. These strains in the coatings are usually non-linear (plastic) even when linear (elastic) straining of the substrate occurs. Coating strain range values are, as a rule, higher than corresponding strain range values of the surface layer of the blade (or bar) during thermomechanical cyclic loading, especially for the SC and DS materials. The number of cycles to the first crack initiation and their distribution on the outer surface of the SC blade airfoil cast with primary crystallographic orientation 001 depend on the secondary orientation of the crystallographic lattice in the airfoil cross section. This is greatly dependent on the circumferential thermal stresses, rs, in the blade, resulting in longitudinal coating cracking, and the value of circumferential elasticity modulus, Es, for the SC superalloy. So, when TMF-cycle testing of SC-coated bar, it is necessary to know


the positioning of the axis of the secondary orientations of 010 and 100 for each bar, because distribution of the first crack on the coated surface can depend on this axis in the bar. The following experimental method of TMF investigation of the protective coatings for turbine blades is suggested. Because elastic-plastic straining of the coating is possible even under elastic straining of the substrate, it is suggested that TMF testing of the coated bars be carried out under conditions of elastic straining of the base material only. This method is expedient for SC turbine blades with primary crystallographic orientation 001, because there are no plastic strains in the outer layers of such blades at operating conditions. This is the result of the low elasticity modulus and high yield strength of blade material for the 001 orientation. The exceptions to the rule are small stress-concentration zones near perforation holes at the cooled blade walls where plastic strains of the blade metal are possible. In accordance with that mentioned previously, the LA-TMF test is recommended for TMF testing of coated bars. The results of this method are similar to SC-TMF testing under conditions of elastic straining of the bar material but are simple in practice. The main problem for SC-TMF testing is the necessity for temperature uniformity of the metal in the extensometer-controlled zone at stationary and nonstationary thermal conditions. This is not necessary for the LATMF test, because during the test, the loading of

Fig. 7.35

the bar is controlled. Moreover, nonuniformity of temperature distribution along the central work zone with a temperature maximum is preferable, because it helps to fix the location of maximum failure in the cross section of the bar. Using standard whole cylindrical bar with constant square cross-sectional work zone is recommended for the LA-TMF test. The stress state of the coating applied to the outer surface of the bar is biaxial, although the stress state of the bar itself during this test is uniaxial. It is recommended to use a control system with two independent channels: first, to control the load, and second, to control the temperature of the bar during LA-TMF testing. This control system makes possible model changes in stress and maximum temperature of the work zone of the bar versus time with any phase shift among these independently controlled parameters. The possibility of stress asymmetry, Rr ⳱ rmin / rmax, varying for all types of LA-TMF tests must also be limited. (rmin and rmax are correspondingly minimum and maximum values of stress during cycle.) The stress-strain state of the bar coatings is calculated for the TMF test conditions by using the developed constitutive model based on the deformational theory of plasticity. The calculational hysteresis loop for the overlay coating Ni20Co20Cr12AlY, obtained on the basis of this constitutive model for the out-ofphase LA-TMF test of the coated bar cast of MAR-M-002 superalloy, is shown in Fig. 7.35(a). The calculations are performed for out-

Hysteresis loops for (1) MAR-M-002 bar and (2) Ni20Co20Cr12AlY coating material of (a) whole cylindrical coated bar and (b) when out-of-phase TMF testing


of-phase LA-TMF test conditions shown in Fig. 7.35(b) for a solid cylindrical specimen made of the MAR-M-002 superalloy with the Ni20Co20Cr12AlY coating. For coating design, the mechanical properties of cast specimens of the same chemical composition as the coating material were used in a first approximation. It is worthwhile to perform the calculations in question before testing, because it allows the proper choice of test conditions for coated specimens capable of realizing the required range of elastic-plastic strains in the coating. It is also worthwhile to perform calculations based on real measurements of the sample temperature and the load applied after testing, for estimation of the actual strain range in the coating. Calculations of strains in the coatings after testing are also required for plotting thermal fatigue curves specialized for coatings. For this purpose, sample thermal fatigue curves plotted from LA-TMF test data are used. Thermomechanical Test Results for CMSX-4 Superalloy with Coatings. Three modifications of the coating were deposited on the test specimens (Ref 29, 30). The first modification of the coating 80 lm thick was deposited by EA method. After its deposition, the coated specimens were annealed at 1080 C for 2 h. Then, they were shot peened with microballs and underwent another annealing under vacuum at 1080 C for 2 h. The coating features a dense, fine-grained structure typical of coatings deposited by EA method. The coating characteristics are presented in Table 7.21. The second modification of the coating designated as Ni7Co12Cr17Al under study was deposited in two stages. At the first stage, a layer of the Ni8Co12Cr7Al coating was deposited by the previously mentioned technological process. At the second stage, a layer of Al-5%Si-1.5%Y alloy was deposited on the surface of Ni8Co12Cr7Al layer by EA method at the rate of 40 g/m2. A diffusion layer 50 lm thick was Table 7.21

formed on the surface on annealing at 1080 C for 2 h. The total coating thickness was 80 lm. The average chemical compositions in the inner and outer zones of the coating are given in Table 7.21. The outer zone of the Ni7Co12Cr17Al coating consists of b-phase. Its structure is typical of the diffusion coatings deposited by EA method (Fig. 4.21d). A multiphase zone is formed under b-phase layer on diffusion annealing. The Co32Ni21Cr8Al coating 100 to 120 lm thick was deposited by the low-pressure plasma spray (LPPS) method using the same technological process as for turbine blades. Thermomechanical Fatigue Tests. Proceeding from the principles discussed previously, the following test conditions were used: temperature and load antiphase alteration, and asymmetric cycle of load alteration without holding the specimen at the maximum cycle temperature (Fig. 7.36).

Fig. 7.36

Stress and maximum temperature vs. time distribution during out-of-phase LA-TMF testing of

coated bar

Characteristics of coatings tested Chemical composition, wt%

Coating Ni8Co12Cr7Al Ni7Co12Cr17Al Co32Ni21Cr8AlY

Deposition method

Thickness, lm

Coating zone

Ni

Co

Cr

Al

Ta

Re

Hf

EA EA EA LPPS

80 80 … 100–120

… Outer Inner …

Base Base Base 32

7.6 6.5 8.1 Base

12.4 7.5 16.9 21

6.5 17.0 6.8 8

4.5 2.0 4.5 …

0.3 0.2 1.1 …

0.4 … 0.5 …

Note: EA, electric arc; LPPS, low-pressure plasma spray


Fig. 7.37

Specimen for TMF tests (in mm)

Fig. 7.38

Temperature pattern in specimen working area

Table 7.22 No.

Specimen loading conditions

Temperature, C

1 2 3 4

Fig. 7.39

100 1100 100 1100 100 1100 100 1100

Stress, MPa rtensile ⳱ rcompressive ⳱ rtensile ⳱ rcompressive ⳱ rtensile ⳱ rcompressive ⳱ rtensile ⳱ rcompressive ⳱

Stress range, MPa Ⳮ500 ⳮ150 Ⳮ200 ⳮ150 Ⳮ200 ⳮ100 Ⳮ200 ⳮ50

Dr ⳱ 650 Dr ⳱ 350 Dr ⳱ 300 Dr ⳱ 250

Thermal strain range distribution in the specimen maximum temperature area

Whole cylindrical specimens of the CMSX-4 superalloy with diameters of 6.5 mm were used to carry out TMF tests (Fig. 7.37). Specimens were heated by conducting electric current with current density of I ⳱ 20 A/mm2. The temperature distribution pattern in the specimen working zone is given in Fig. 7.38. Earlier research has revealed that the previously mentioned heating conditions do not cause changes in superalloy mechanical properties (Ref 31). At the same time, specimens were loaded using a hydraulic loading device. The main parameters of specimen loading conditions are presented in Table 7.22. The coating service life was set by the moment when cracking on the surface was detected. Control over cracking was carried out by an optical method and by nondestructive liquid-penetrant testing. The resolving power of the technique in use was as follows: crack length ⱖ0.5 mm and depths of penetration of 0.05 mm. To specify specimen-loading conditions during the testing cycle, strain was measured in the zone of the maximum temperature. Marks made of the wire with a diameter of 0.05 mm were spot welded to the specimen surface. The distance between the marks was measured at the minimum and maximum temperatures of the cycle using an optical measuring device (Fig. 7.39). Then, measurements were carried out at heating under mechanical loading, and the total strain was calculated in the central zone of the specimen. The total strain range versus the applied stress range in the center of the specimen is shown in Fig. 7.40. For each coating modification, fatigue tests under thermal cycling were carried out at three levels of stress amplitudes in the specimen. Based on the experimental data, a thermomechanical fatigue curve was plotted on the stress

Fig. 7.40

Strain range, De, of specimen versus stress range, Dr, in the maximum temperature area


range (Dr, MPa)-service life (N, cycles) coordinates. The terms used for test data description are as follows: coating service life is the number of cycles to crack(s) detection on the coated specimen surface, and CMSX-4 superalloy service life is the number of cycles to crack(s) detection on the uncoated specimen surface. Service lives of uncoated specimens of the CMSX-4 superalloy tested by the previously mentioned method are of much interest, because these data may be used for comparison with coating service lives and the analysis of the effect of coatings on the alloy properties. When cracking was detected on the specimen surface, the test was not terminated. It was run to failure of the specimen. Test results are given in Table 7.23 and Fig. 7.41. Fractographic study of the specimens has shown that they are typical of the failure at TMF testing at narrow stress range. The nucleation site of the crack is situated on the specimen surface (Fig. 7.42). The cracks grow into the specimen and cause its failure. During testing, most of the cycles are spent on crack nucleation on the specimen surface. After crack nucleation, the service life of the cracked specimen is up to 8 Table 7.23 test results Number of cycles to cracking 2850 6190 … 15,780 20,420 48,670 24,120 41,270

Fig. 7.41

Uncoated CMSX-4 alloy specimen

Number of cycles to specimen failure

Stress range (Dr), MPa

3630 6510 4840 16,950 21,820 … 25,480 44,860

650 650 650 450 350 300 300 250

Cyclic life of CMSX-4 superalloy. 1, Cracks on surface specimen; 2, failure of specimen

to 20% of the total service life of the specimen under study. Measuring specimen dimensions after their testing has revealed no elongation or changes of their diameters. This is evidence of the compliance of the tests with the chosen loading conditions in the range of superalloy elastic deformation. Test results for the specimens with the Co32Ni21Cr8AlY coating are shown in Fig. 7.43. The number of cycles to crack appearance on the specimen surface with the Co32Ni21Cr8AlY coating is much less than for an uncoated specimen. Crack appearance and spacing in the surface network are similar to those typical of thermal fatigue cracks, which develop in the coatings on blade surfaces (Fig. 7.44). The cracks were detected along the full length of the specimen. There were cracks in the areas where the maximum temperatures did not go over 1000 C. Crack nucleation is a consequence of the accumulation of plastic strain in the coating as well as the formation of pores and minor surface ruptures, which join together gradually and form a thermomechanical fatigue crack in the coating (Fig. 7.45a,b). Crack propagation rate in the CMSX-4 superalloy is much less than that in the coating. While cracks 80 to 100 lm long were detected in the coating after 1000 cycles, a further 3100 cycles resulted in crack penetration of not more than 100 to 120 lm into the alloy. Test results for specimens with the Ni8Co12Cr7Al coating are shown in Fig. 7.43. The mechanism of crack formation and their appearance and surface network are identical with those typical of specimens with the Co32Ni21Cr8AlY coating. The Ni8Co12Cr7Al coating service life is 4 times as long as that of the Co32Ni21Cr8AlY coating at the stress range of Dr ⳱ 300 MPa and 2 times as long at the stress range of Dr ⳱ 250 MPa. Test results for specimens with the Ni7Co8Cr17Al coating are shown in Fig. 7.43. The mechanism of crack formation and their appearance and surface network are identical with those typical of specimens with the Co32Ni21Cr8AlY and Ni8Co12Cr7Al coatings. The Ni7Co12Cr17Al coating service life is 13 times as long as that of the Co32Ni21Cr8AlY coating at the stress range of Dr ⳱ 300 MPa and 8 times as long at the stress range of Dr ⳱ 250 MPa. Mechanism of Coating Effect on Thermomechanical Fatigue of Superalloys. A coating with physical and mechanical properties identical to those of the protected superalloy can be


considered as an “ideal” coating, with high thermomechanical fatigue resistance. This coating can hardly be developed, because a thin layer of polycrystalline coating cannot have mechanical properties typical of a modern SC superalloy. However, the development of a coating with characteristics that are as similar to those of the uncoated superalloy as possible is an advantageous line in producing a coating with high thermal fatigue resistance. The use of LA-TMF test methods with anti-

Fig. 7.42

phase change in temperatures and stresses allows the simulation of stress and strain conditions in the coatings on turbine blade surfaces. Under the selected test conditions, the applied loads contribute to the increase in compressive stresses in coatings at high temperatures and to the increase in their tensile stresses at low temperatures. The test technique in use allows the distinctions in coating resistance to thermal fatigue crack nucleation and propagation to be revealed.

Fracture of specimen CMSX-4 superalloy, 3658 cycles, Dr ⳱ 650 MPa. (a) Site of failure, 200⳯; (b) crack development area, 500⳯


Table 7.24 Comparison of average lives of uncoated CMSX-4 specimens and CMSX-4 specimens with different coatings Number of cycles to cracking (%) Stress range (Dr), MPa 650 350 300 250

CMSX-4 2850 20420 24120 41270

Co32Ni21Cr8AlY

Ni8Co12Cr7Al

Ni7Co12Cr17Al

368 (13%) … 911 (4%) 2330 (6%)

… 2840 (14%) 4178 (17%) 6592 (16%)

… 10550 (52%) 12484 (52%) 18587 (45%)

(100%) (100%) (100%) (100%)

Specimen fractography after thermal fatigue testing has revealed that, in all cases, the crack nucleation sites are located on the specimen surfaces and the fracture type is similar to that caused by applying alternating loads. Cracks are developed in the coatings after accumulation of considerable strain. The tests have revealed a noticeable distinction between the time to thermomechanical fatigue crack detection on the uncoated specimens and on the specimens with different coatings. All test data are given in Table 7.24 for com-

Cyclic life of uncoated CMSX-4 superalloy (4) and CMSX-4 with coatings. 1, Co32Ni21Cr8AlY coating; 2, Ni8Co12Cr7Al coating; 3, Ni7Co12Cr17Al coating

Fig. 7.43

Crack location on specimen surface with Co32Ni21Cr8AlY coating (liquid penetrant testing), 900 cycles

Fig. 7.44

parison. The Co32Ni21Cr8AlY coating service life is only 6 to 8% of uncoated CMSX-4 alloy service life. Thus, if there are zones in turbine blades with the stress range similar to that used at testing, numerous thermal fatigue cracks will develop in the Co32Ni21Cr8AlY coatings on their surfaces. The development of these cracks on the blades may be unexpected, because, as a rule, coating service life is not taken into account in blade service-life calculations. Changes in coating physical-chemical properties can have a significant effect on their service lives. The Ni8Co12Cr7Al coating service life is 2 to 4 times as long as that of Co32Ni21Cr8AlY coating. The Ni7Co12Cr17Al coating has also demonstrated high service life. The Ni7Co12Cr17Al coating service life is 8 to 10 times as long as that of the Co32Ni21Cr8AlY coating, and it is 45 to 52% of the CMSX-4 superalloy service life. This service life is primarily due to close agreement of linear TECs of the coating and the CMSX-4 superalloy. However, the effect of technological factors in forming the outer diffusion layer from a melted aluminum-base liquid phase cannot be excluded. The presence of the outer diffusion layer has a desirable effect on the coating service life, too. As is known, unlike the applied coatings, diffusion coatings accumulate residual compressive stresses in the temperature range below 900 C. The chosen testing technique that uses the specimens with diameters of 7 mm does not allow the estimation of the effect of crack formation in the coating on CMSX-4 specimen service lives. The comparison of the number of cycles to failure for coated and uncoated specimens has revealed that the number of specimens tested to failure and the test data are insufficient for unbiased estimation of this effect. The agreement of the test data from the coating service lives on the specimens with those on the blades depends on blade service conditions in the engine. In short-term cyclic tests of the engines under high thermal loading conditions,


Fig. 7.45

Cracks in Co32Ni21Cr8AlY coating at (a) the initial stage of crack development and (b) the stage of its propagation into the superalloy after 550 cycles. 500⳯

there is a high probability that the data from the coating service lives on the blades will show close agreement with the respective results of laboratory testing of the specimens. In the case of long-term tests in the engines, these data will depend to a considerable extent on such coating characteristics as oxidation resistance and resistance to hot corrosion.

REFERENCES

1. M.I. Wood, The Mechanical Properties of Coatings and Coated Systems, High Temperature Corrosion Advanced Materials and Coatings, R. Streiff, J. Stringer, R. Krutenat, and M. Caillet, Ed., Elsevier Science Publishers, 1989, p 633–643


2. P.T. Kolomytzev, Gas Corrosion and Strength of Nickel Alloys, Moscow, Metallurgy, 1984, 216 p 3. E.M. Grale, Research into NiAl and Ni3Al, Mechanical Properties of Intermetallic Compounds, J.H. Westbrook, Ed., John Wiley & Sons, New York, 1960, p 266–299 4. A.G. Rozner and R.J. Wasilelevski, Tensile Properties of NiAl and Ni3Al, J. Inst. Met., Vol 94 (No. 5), 1966, p 169–175 5. V.P. Buntushkin and O.A. Bazyleva, Casting Ni3Al Intermetallic Compound-Base Alloy for Single-Crystal Engine Turbine Rotor Blades, Aviatsionnaya Prom-st., N 3– 4, 1997, p 61–65 6. G.W. Goward, Current Research on the Surface Protection of Superalloys for Gas Turbine Engines, J. Met., Vol 22 (No. 10), 1970, p 31–39 7. P.T. Kolomytzev and N.V. Abraimov, Research into Temperature Effect on Coatings Ductility, Phys. Chem. Mech. Mater., No. 2, 1982, p 104–107 8. Y.S. Eliseev, N.V. Abraimov, and V.V. Krymov, Chemical-Heat Treatment and Protective Coatings in Aircraft EngineBuilding, Moscow, Vysshaya Shkola, 1999, 525 p 9. V.I. Nikitin, Y.A. Tamarin, and N.V. Zabrodina, Investigations of Structural and Physical-Chemical Characteristics of NiCoCrAl System Alloys, Protective Coating Application and Use, Leningrad, Nauka, 1987 10. A.I. Krivko, A.I. Epishin, and A.I. Samoilov, X-Ray Method of Research into Stressed-Strained State of Composite Materials Based on Calculations and Experiments, Zavod. Lab., Vol 1 (No. 5), 1985, p 34–38 11. A.I. Samoilov, I.A. Ignatova, A.M. Vorobiev, and V.S. Kozlova, Measurements of Thermal Stresses in Ni-(Co)-Cr-Al-Y HeatResistant Coatings, Zavod. Lab., Vol 50 (No. 11), 1984, p 36–39 12. E.G. Kuzovkov, Numerical Simulation of Stressed State for Strained Bodies by Elementary Cells Method, Kiev, Institute of Strength Problems, 1981, 37 p 13. V.I. Nikitin, Corrosion and Protection of Gas Turbine Blades, Leningrad, Machinostroenie, 1987, 272 p 14. V.I. Nikitin and I.P. Komossarova, Method of Protective Coating Testing for Lifetime, Proc. CKTI, (No. 176), 1980, p 67–73

15. V.I. Nikitin, Y.A. Tamarin, and N.V. Zabrodina, Corrosive and Mechanical Properties of Protective Coatings, AqueousChemical Conditions and Corrosion of Power Equipment, No. 201, Leningrad, CKTI, 1986, p 72–81 16. J.F. Paskiet, D.H. Boon, and C.P. Sullivan, Effect of Aluminide Coating on High-Cycle Fatigue Behavior of Nickel-Base HighTemperature Alloy, J. Inst. Met., Vol 100 (No. 2), 1971, p 58–62 17. M.M. Gell, G.P. Leverant, and C.H. Wells, The Fatigue Strength of Nickel-Base Superalloys, ASTM STR, No. 64, 1970, p 113– 153 18. I.I. Ischenko, V.I. Omelchenko, and B.N. Sinaisky, Research into Effect of Heat-Resistant Coatings on Superalloy Fatigue Strength, Strength Probl., No. 10, 1971, p 76–79 19. I.I. Ischenko, V.I. Omelchenko, and B.N. Sinaisky, Research into Effect of Heat Treatment and Protective Coatings on JS6K Superalloy Fatigue Strength, Strength Probl., No. 8, 1976, p 10–15 20. L.M. Akimov, Superalloys Fatigue Lives, Moscow, Metallurgy, 1977, 220 p 21. P.T. Kolomytzev and P.G. Lebedev, Cyclic and Long-Term Rupture Strength of Alloys with Protective Coating, Strength Probl., No. 9, 1972, p 92–95 22. L.B. Getsov, A.I. Rybnikov, and I.S. Malashenko, Fatigue Resistance of Coated Superalloys, Strength Probl., No. 5, 1990, p 51–56 23. T.N. Rhys-Jones and T.P. Cunningham, The Influence of Surface Coatings on the Fatigue Behaviour of Aero Engine Materials Surface and Coatings Technology, Vol 42, 1990, p 13–19 24. L.P. Lozinsky, B.I. Musienko, and A.A. Ivanenko, Effect of Diffision Coatings on Turbine Blade Damage at Thermal Cycling, Vol 8, Kiev, Naukova Dumka, 1974, p 134– 136 25. G.R. Halford, T.G. Meyer, R.S. Nelson, and D.M. Nissley, “Fatigue Life Prediction Modeling for Turbine Hot Section Materials,” NASA Technical Memorandum 100291, 33rd Intl. Gas Turbine and Aeroengine Congress and Exposition, (Amsterdam, The Netherlands), ASME, 5–9 June 1988 26. E.G. Ellison, Thermomechanical Strain Cycling and Testing at Higher Temperature,


Measurement of High Temperature Mechanical Properties of Materials, M.S. Loveday and M.F. Day, Ed., HMSO, London, 1982, p 204–218 27. D.A. Miller and R.H. Priest, Materials Response to Thermal Mechanical Strain Cycling, High Temperature Fatigue: Properties and Prediction, R.P. Skelton, Ed., Elsevier Applied Science, 1987, p 113–175 28. Y.A. Nozhnitsky, R.A. Doulnev, and V.G. Sundyrin, Damage Mechanisms for Thermomechanical Fatigue of Aircraft Engines Materials, 81st Meeting of the AGARD SMP, (Banff, Canada), 2–4 Oct 1995, and published in CP-559

29. Y.A. Tamarin, V.G. Soundyrin, and N.G. Bychkov, Thermo-Mechanical Fatigue Tests of Coatings for Turbine Blades, High Temperature Surface Engineering, J. Nicholls and D. Rickerby, Ed., IOM Communications Ltd., 2000, p 157–169 30. Y.A. Tamarin and H.C. Low, Thermal Fatigue of Protective Coatings on CMSX-4 Superalloy, High Temperature Corrosion 5, R. Streiff, I.J. Wright, R. Krutenat, M. Caillet, and A. Cailerie, Ed., Trans Tech Publication, 2001, p 647–654 31. N.G. Bichkov, Test Rig for Thermo-Mechanical Fatigue Tests, Zavod. Lab., Vol 57 (No. 12), 1991, p 57–58.



CHAPTER 8

Electron Beam Thermal Barrier Coatings DEVELOPMENT of thermal barrier coatings (TBCs) applied to cooled blades is one of the trends for improving gas turbines. Unlike aluminide protective coatings, ceramic coatings not only protect blade surfaces from high-temperature oxidation and corrosion but also prevent base material softening at high temperatures. Thermal barrier coating application allows the reduction of the blade temperature and the significant increase in its service life. Under both steady-state and transient conditions, the application of TBCs can diminish temperature gradients over the blade surfaces as well as reduce thermal stresses in them. Between 1980 and 2000, a great number of scientific papers and patents were published. They covered the issues concerning development, deposition techniques, and use of TBCs. A typical design of a TBC is presented in Fig. 8.1. The ceramic layer deposited directly on the superalloy surface does not show the required service life. Penetration of oxygen through the ceramic layer to the superalloy surface results in its quick oxidation and in spallation of the ceramic layer. That is why, as a rule, a TBC consists of at least two layers. An inner aluminide heat-resistant bond coat may be formed by different techniques. It may be either a diffusion or an overlay coating, depending on the requirements of its physical-mechanical properties and protection targets. The requirements of bond coat properties and protective coatings properties are much the same, yet the bond coat should meet some special requirements. First of all, it must be highly heat resistant; the oxides formed on its surface should have high adhesion to both the bond coat and the outer ceramic layer. When choosing a bond coat composition, one should pay special attention to its yttrium content as

well as to the contents of the other elements, which guarantee high oxide adhesion to the surface and reactive element effect (Ref 1). It is of special importance for bond coats deposited by the electron beam (EB) technique, because their yttrium contents depend on the yttrium content of the liquid bath and vary within wide limits (Ref 2). In this case, the required yttrium content of 0.2 to 0.3% is guaranteed by different technological procedures, such as direct yttrium addition to the liquid bath. Under these conditions, it is noteworthy that high yttrium contents of the liquid bath cause slag formation on its surface, thus resulting in occurrence of microdrops. These microdrops on the bond coat surface may provoke defects in the ceramic layer.

Fig. 8.1

Thermal barrier coating design


It should be taken into consideration that TBCs are usually applied to the blades of hightemperature turbines. The blades of such turbines feature directionally solidified or singlecrystal structures, thin walls, and high cooling efficiency. Under service conditions, high thermal stresses and strains arise in these blades, especially in their surface layers. That is why thermomechanical fatigue characteristics are as important in choosing a bond coat composition as its heat resistance. During thermal cycling, the bond coat should not experience considerable plastic strain. For example, the effect of a “rippled” blade surface (Fig. 2.10) always entails spallation of the ceramic layer. The outer zirconium oxide/yttrium oxide (ZrO2-Y2O3) system base ceramic layer can be applied by two techniques: air plasma spraying of powders (TBC-APS) or vapor condensation at EB evaporation of ceramic pellets (TBC-EB). For this system, ceramic coating service life depends on Y2O3 content. The ZrO2-(6 to 9%) Y2O3 compositions are usually applied, because they have demonstrated maximum service lives in the tests carried out (Ref 3–5). However, one should bear in mind the fact that the coating service life depends not only on its chemical composition but also on its structure and adhesive strength at the ceramic layer/bond coat interface, which depends on deposition technique. For coatings deposited by different techniques, the optimal chemical compositions may be other than that stated previously. The ceramic layer deposition technique determines such characteristics as ceramic layer structure and adhesive strength, its corresponding service life, thermal stresses in the ceramic layer, and its surface roughness. The main difficulty in designing TBCs for turbine blades lies in the combination of the ceramics on the blade surface and the superalloy that they are made of. At heating-up/cooling-down cycling, considerable difference between the ceramics and superalloy expansion coefficients (D␣ 5.0 ⳯ 10ⳮ6 1/C) causes the generation of high thermal stresses in ceramics, which in turn results in ceramic layer spalling from the surface. To reduce thermal stresses, various technological procedures are used. In the ceramic layer deposited by the APS technique, special heat treatment is used to form a network of microcracks that break the ceramics into isolated fragments (Ref 6). In the ceramic layer deposited by the EB technique, some specific columnar structure is formed that is readily

fragmentizing when tensile stresses arise in it (Ref 7). The point crucial to success in the development of TBCs lies in obtaining the required adhesive strength of the ceramic layer/heat-resistant bond coat, providing for holding of the ceramics on the blade surface during all the blade service life. As a rule, in aircraft engine manufacturing, the technique of plasma deposition is used for nozzle vanes; in aircraft engine turbine blades, the EB technique is considered to be preferable. This is due to the fact that the following properties can be rendered to the ceramic layer: ●

The specific columnar structure, with the crystallites oriented perpendicular to the surface, forms in the ceramic vapor-deposited layer. In the case of tensile stresses, the ceramic layer is readily fragmentizing, thus reducing ceramic tearing stress during thermal cycling. ● In the temperature range of 850 to 950 C, which is below the blade heating temperature at ceramic layer deposition, compressive stresses arise in it. Their generation is due to the different values of the ceramic and superalloy thermal expansion coefficients. These stresses do not relax on subsequent process annealing and under service conditions. ● The adhesive strength of the ceramic layer is controlled by physical-chemical reactions occurring between the ceramics and the metallic bond coat. As-deposited ceramic layer adhesive strength is above 70 MPa. ● The surface roughness of the ceramic coatings does not exceed 1.5 lm after their deposition. This chapter deals with TBCs deposited by the EB technique.

Ceramics for EB Evaporation For EB evaporation, it is necessary to produce ceramic pellets that are conveniently loaded into a vaporizer and sufficiently strong to preserve their shape during shipment and treatment with EB. The composition of the pellets should be similar and should provide for forming a coating of the required chemical and phase compositions when used for applying ceramic layers. Both phase composition of the ceramic pellets and their chemical homogeneity depend on chemical and phase compositions of the initial ZrO2, the

Electron Beam Thermal Barrier Coatings / 163

conditions of its stabilization with Y2O3, the production process for ceramic pellets, their annealing time and temperature, and their cooling rate. The process of TBC ceramic layer deposition and the stability of its properties depend, to a large extent, on the characteristics of the material evaporated. Zirconium oxide is used as a base material to form the TBC ceramic layers. The choice of this material is based on its low thermal conductivity and high thermal expansion coefficient (TEC) of the ZrO2 cubic phase (␣ 11 ⳯ 10ⳮ6 Kⳮ1) in the temperature range of 20 to 1200 oC. The specific feature of ZrO2 is its polymorph transformations, causing significant volume changes (Fig. 8.2) (Ref 8). When heating ZrO2, its low-temperature monoclinic (M) modification transforms into the tetragonal (T) and then into the cubic modification (C): M ↔ T ↔ C. M ↔ T transformation occurs martensitically and features substantial hysteresis. When the T phase transforms into the M one, the crystal lattice volume increases by 3 to 12%. Making oxide additions prevents phase transformations in ZrO2. They form solid solutions of fluorite crystalline structure with ZrO2. The oxides stabilizing ZrO2 cubic modifications are as follows: Y2O3, MgO, CaO, and Yb2O3 (Ref

Fig. 8.2 Ref 8

Low-yttrium region of ZrO2-Y2O3 phase diagram. M, monoclinic; T, tetragonal; C, cubic phase. Source:

9). Among the previously mentioned oxides, the most stable solid solutions are formed in the ZrO2-Y2O3 system taken as a base of high-temperature-resistant ceramic coatings. The required addition contents are taken from the state diagrams, although under actual conditions, it is of great importance to consider the way of making additions, system homogeneity, and heat treatment. The quality of an overlay ceramic coating depends on the material homogeneity of the ceramic pellet being evaporated. This is caused by the evaporation of a relatively shallow liquid bath on the surface of the ceramics. Under other equal conditions, mixture homogeneity depends on the properties of the mixed components: the finer they are, the less their difference in density is, and the more their ranges of particle dimensions overlap, the more homogeneous mixture can be produced. A few techniques for ZrO2-Y2O3 solid-solution synthesis are available. Each of them has both advantages and disadvantages (Ref 10): ●

Mixing ZrO2 and Y2O3 powders, followed by mixture heat treatment at 1600 to 1800 C and sintering in electric-arc furnaces ● Mechanical mixing of hydroxides or zirconium and yttrium salts, followed by their heat treatment at 1000 to 1200 C ● Simultaneous deposition of homogeneous hydroxide mixtures or other unsolvable zirconium and yttrium compounds from salt solutions If ZrO2 and Y2O3 powders are used and the initial mixture is prepared by their mechanical mixing, the phase composition of the pellet fabricated depends on a mixing process quality and an annealing temperature. Zirconium oxide full stabilization does not occur even after adding 20% Y2O3 to it (Table 8.1). Some Y2O3 remains in the free state after annealing. As a rule, different densities of the initial components and the size and shape of their particles do not guarantee the uniform distribution of both oxides within the pellets on mixing. In the mechanically mixed pellets, one can detect the zones of different chemical compositions (Table 8.2). The pellets made of ZrO2-Y2O3 powders using the hydroxide method are found to be much more homogeneous in their chemical and phase compositions (Ref 11). Zirconium and yttrium hydroxides are deposited simultaneously from chloride solutions at pH ⳱ 8 to 9, using aqueous


Table 8.1 Phase composition of ZrO2-Y2O3 ceramic pellets made of mechanical mixture of ZrO2 and Y2O3 powders(a) Phase content(b), vol% Annealing at 1350 oC(c), 10 h

Annealing at 1700 oC(c), 10 h

Chemical composition

M

T

C

Y2O3(d)

M

T

C

Y2O3(d)

ZrO2 ZrO2-2Y2O3 ZrO2-4Y2O3 ZrO2-6Y2O3 ZrO2-8Y2O3 ZrO2-10Y2O3 ZrO2-15Y2O3 ZrO2-20Y2O3 ZrO2-25Y2O3 ZrO2-30Y2O3 ZrO2-40Y2O3

79 72 60 42 20 16 15 15 5 5 5

21 28 40 58 60 43 45 42 40 40 40

0 0 0 0 20 41 40 43 55 55 55

Not detected Not detected Detected Detected Detected Detected Detected Detected Detected Detected Detected

75 75 50 40 10 5 0 0 0 0 0

25 25 50 60 60 40 40 20 15 10 0

0 0 0 0 30 55 60 80 85 90 100

Not detected Not detected Not detected Not detected Not detected Not detected Not detected Not detected Not detected Detected Detected

(a) Size of mixed powders, 20–60 mm. (b) Cooled in furnace. (c) M, monoclinic; T, tetragonal; C, cubic. (d) Free Y2O3

solution of ammonia in accordance with the reaction: ZrOCl2 Ⳮ YCl3 Ⳮ NH4OH ⇒ Zr(OH)4↓ Ⳮ Y(OH)3↓ Ⳮ NH4Cl

After simultaneous deposition of the components, the deposit is isolated by filtration and annealed at 1000 C. Some characteristics of commercial ZrO2-Y2O3 powders produced by the hydroxide method and the phase compositions of the powders after annealing under various conditions are presented in Table 8.3. After agglomeration, the powders feature a particle size of 30 to 80 lm. Unlike the powders produced by mechanical mixing, the powders produced by simultaneous deposition, as well as the pellets made of them, have more homogeneous phase compositions. To fabricate ceramic pellets from ZrO2-Y2O3 powders for their further evaporation, different techniques can be used, for example, the semidry molding technique (Ref 12). The main feaTable 8.2 Changes in phase composition along side face of ZrO2-8%Y2O3 ceramic pellet with dimensions of 50 ⴒ 50 ⴒ 50 mm made of mechanical mixture of ZrO2 and Y2O3 powders Phase constituents(a), vol% Surface zone

M

T

C

End ↓ Center ↓ End

21 12 20 10 15

60 88 60 65 75

19 ... 20 25 10

(a) M, monoclinic; T, tetragonal; C, cubic

tures of the ceramic pellet fabrication technique are as follows: the ZrO2-Y2O3 powders produced by the hydroxide process are screened and, if necessary, mixed with certain oxide additions in the special equipment designed for this purpose. The prepared powders are moistened with a 4% solution of polyvinyl alcohol, and, after becoming granulated, they are kept in a closed vessel for 24 h. After being conventionally molded in special dies to form pellets of the specified size, they are annealed in a furnace at 1350 oC for 10 h. The furnace heating-cooling rates should be below 100 C/h. On heat treatment, noticeable pellet shrinkage may occur. It depends on the granulometric compositions of the powders in use as well as their molding and firing conditions. The results of research into the effect of dimension ratio on cylindrical pellet densities before and after firing are presented in Table 8.4. The pellets for evaporation were produced using the semidry technique. Firing zirconium-dioxide-base semiproduct pellets for EB evaporation does not bring about glassy phase formation, and the process requires a long time and high temperatures. The firing process may be divided into several stages: preheating, with a preset rate of temperature rise; firing at a preset temperature for a preset period; and cooling down. At the preheating stage, the rate of temperature rise is chosen depending on the parameters of the process under way, such as rates of moisture and plasticizer removal and product hardening. The rate of temperature rise and drop is controlled, to avoid thermal stresses causing cracking and straining. The maximum firing temperature and holding time depend mainly on oxide compositions and their particle


Table 8.3 Chemical and phase compositions (vol%) of ceramic powders fabricated by simultaneous deposition from zirconium and yttrium hydroxides Annealed(a) at 1350 C, 10 h

No annealing(a)

Annealed(a) at 1700 C, 10 h

Grade

Specific surface, cm2/g

Average grain diameter, mm

ZrO2

Y2O3

M, %

T, %

C, %

M, %

T, %

C, %

M, %

T, %

C, %

YSZ-3 YSZ-8 YSZ-12 YSZ-15

1996 2007 2265 ...

5.83 5.30 4.73 ...

Base Base Base Base

3–4 8–10 11–12 15–16

40 13 0 0

69 74 20 0

0 13 80 100

... 13 0 0

... 74 22 0

... 13 78 100

... 6 0 0

... 74 15 0

... 20 85 100


Table 8.4

Effect of molding and annealing conditions on ZrO2-8%Y2O3 pellet density Molding force, kg

Cylinder height before/ after annealing, mm

Density before/ after annealing, g/cm3

Porosity after annealing, %

YSZ-8

10,000

YSZ-8

20,000

22/18 45/38 25/21 43/36

2.12/3.59 2.07/3.29 2.53/4.18 2.45/4.05

36 42 25 28

Composition(a)

(a) Compositions are additionally doped with 5% NiO.

sizes. Firing temperature is chosen with allowance for completion of the sintering process for powder particles within acceptable time intervals. At firing, the access of air to the pellets is limited. Under conditions of air deficit, polyvinyl alcohol decomposes, forming more simple compounds and releasing elemental carbon that disperses in the pellet in the form of the finest coke particles. The carbon content of as-produced pellets may be as high as 0.02 to 0.05%. Its presence can affect the processes of ceramics evaporation. The stability of ZrO2-Y2O3 initial powder properties and the process of its manufacturing can guarantee stability of as-produced pellet dimensions. The properties of the finished products depend not only on chemical compositions

Table 8.5 ZrO2-8%Y2O3 ceramic pellet phase composition Phase composition(a), vol% Lot No.

M

T

C

1 2 3 4 5 6 7

30 30 40 40 40 20 10

30 30 30 30 30 25 25

40 40 30 30 30 55 65


of oxide powders but also on their phase compositions, grain sizes, and manufacturing processes. That is why an important component of the quality-control system is a procedure of initial powder control on delivery from the suppliers. On compacting semiproduct pellets, their visual inspection is carried out to detect outer pellet defects. Their density and geometry should also be controlled. After firing, the pellet quality is controlled by visual inspection and measurements of their shrinkage and density. Their phase and chemical compositions are also analyzed. Phase composition of the pellets depends on the chemical compositions of their initial powder components, the firing temperatures, and the cooling rates. Data on phase compositions for several commercial batches of ZrO28%Y2O3-5%NiO pellets are presented in Table 8.5. Firing conditions were 1350 C for 5 h. The presence of monoclinic phase is caused by slow cooling of the pellets in the furnace.

Main Features of TBC Deposition Technique There are basic differences between the deposition techniques for metallic and ceramic overlay coatings. Deposition of metallic coatings is not a finishing procedure in forming their structure. Subsequent mechanical, heat, or chemical treatments can cause noticeable


changes in their structures, and, in some cases, they may change chemical composition of the coating as well. These treatments can favor the healing of defects formed in the case of violation of the coating deposition technique parameters. As for ceramic coatings, their deposition technique is the only factor that sets a coating structure pattern and its adhesive strength on the bond coat surface. These characteristics cannot be changed by subsequent treatments. That is why the technique of ceramics deposition should feature consistency. Malfunction within a technological cycle results in spoilage. The only way to reclaim this spoilage is to remove the deposited ceramic layer and repeat its deposition. This is not an easy task for finished blades arriving for TBC deposition. Thermal barrier coating deposition technique includes the following main stages: bond coat deposition, preparation of a bond coat surface, ceramic layer deposition, and coating inspection. The technological operations of a TBC bond coat deposition are similar to the technological operations of heat resistant coating deposition discussed previously. The peculiarities of the ceramic layer deposition technique are mainly concerned with the preparation of the heat resistant bond coat surface for ceramics deposition as well as with ceramics evaporation and condensation conditions. The main characteristics governing a TBC service life are ceramic layer structure and adhesive strength at the ceramic layer/metallic bond coat interface. The ceramic layer structure depends on ceramics evaporation and condensation conditions, while adhesive strength at the ceramics/bond coat interface is affected by different process parameters. It depends on such factors as methods of bond coat surface preparation, blade heating method and temperature, conditions of its revolution past evaporator and heating systems, ceramics evaporation rates, and the quality of vacuum in the unit. Insufficient initial adhesive strength at the ceramic layer/ bond coat interface may be caused by various optimal parameter violations of the designed technique for ceramic layer deposition, especially at the early stage of ceramics vapor condensation. Bond Coat Surface Preparation. The quality of bond coat surface preparation is crucial for physical-chemical processes of bond coat/ ceramic layer interaction, adhesive strength at this interface, its stability at high temperatures, and ultimately, for a TBC service life. There are

two basic approaches to bond coat preparation for ceramic layer deposition: ●

Clearing oxides from the bond coat surface and finishing it to the surface roughness of 0.32 to 0.80 lm ● Preoxidizing the bond coat surface to form Al2O3 on it The first finishing variant is a conventional process that is used for the surfaces to be coated. As a rule, it includes any abrasive machining of the surface, such as vibrogrinding, vibropolishing, and so on. For bond coats deposited by EB technique, a compacting procedure of microball peening is required. Bond coat surface hardening that occurs at vibroprocesses and microball peening facilitates further activation and development of diffusion processes between the metallic bond coat and the ceramic layer. When selecting the machining technique and tolerable surface roughness of the bond coat surface, it is necessary to take into account the requirement of the maximum heat resistance. The surfaces with low roughness feature the best heat resistance. It is advantageous that the surface roughness (R) of the bond coat surface prepared for ceramics deposition should not exceed 0.80 lm. Sharp microprojections on the surface become readily oxidized and cause the spallation of the ceramic layer. As experience shows, bond coat surface roughness affects coating texture. X-ray diffraction analysis of crystallite texture has shown that, on condensing the ceramic layer on the surface with R 0.65 lm, the grainboundary angle for the crystallites oriented in [011] direction is not more than 1 to 2. The ceramic layer deposited on the surface with the surface roughness of 1.25 lm has columnar structure, with a grain-boundary angle of 10. It contains crystallites oriented in the directions of [311] and [111]. This structure is more susceptible to erosion. Special attention should be paid to the surface quality. In view of earlier publications (Ref 13), glass particles left on the surface after peening with glass microballs may form phases, causing premature spallation of the ceramic layer. Similar results are reported (Ref 14) that throw light on the presence of silicon on the surfaces processed with the abrasive material, and the presence of iron on the surfaces peened with steel microballs. To prepare a bond coat surface that is clean of oxides, the following technique can be used.


Immediately before the ceramic layer condensation onset in the EB unit, a thin metallic layer (1 to 5 lm thick), with the composition identical to that of a bond coat or close to it, is deposited for 1 to 2 min. A thin metallic layer with a juvenile surface free from oxides and dirt facilitates the formation of a strong diffusion bond between the ceramics and the metal. The second variant of bond coat surface preparation includes its oxidation before ceramic layer deposition (Ref 15). Bond coat preoxidizing should guarantee the formation of a thin ␣Al2O3 layer (1 to 2 lm thick) on its surface. It should ensure high adhesive strength of a bond coat/ceramics interface and its long service life at elevated temperatures. Preoxidizing may be achieved by different methods. It may be carried out either in special units or in an EB unit before ceramics evaporation. The crucial feature of the process is a need for preoxidizing under the conditions, which guarantees the formation of Al2O3 with the most perfect crystal structure available over the whole blade surface. To meet this requirement, the following factors should be controlled: thorough cleaning of foreign particles, such as SiO2, from the surface before the process; the specified aluminum content on the surface and high homogeneity of bond coat surface composition; and a special pattern of supplying the surface with oxygen. It has been established that one can find a variant of surface preparation procedure combining the two methods just discussed (Ref 16). A tablet made of a powder mixture, such as 13%Al7%Al2O3-80%(ZrO2-7%Y2O3), or of mixtures of more complex compositions (Ref 17) is placed on the top of the ZrO2-Y2O3 pellet to be evaporated. On EB heating, the processes of melting and continuous evaporation, first of the tablet and then of the ceramic pellet itself, take place. Because the partial vapor pressures of the previously mentioned mixture components are different, selective evaporation of the molten tablet elements occurs. Aluminum is the first to evaporate. It is followed by Al2O3 and finally by ZrO2. The corresponding composition and structure gradients through the coating thickness accompany the deposition of such a vapor flow on the substrate. It is obvious that variation of tablet composition and thickness may result in modification of the composition, structure, and thickness of the coating deposited. At the metal substrate temperature of 850 to 1000 C, which is higher than the aluminum

melting temperature, condensation proceeds according to the vapor-liquid-solid phase pattern. Melting of the thin bond coat surface layer took place at condensation onset. Then, (nickel, cobalt) aluminum intermetallide-base structure was forming. Those deposition conditions made it possible to achieve high adhesion at the coating-substrate interface. It allows less stringent requirements to the initial bond coat surface roughness. The previously mentioned features of Al-Al2O3-ZrO2 multicomponent mixture evaporation and condensation open up possibilities of bond coat and ceramic layer deposition within one technological cycle, eliminating intermediate heat treatment and mechanical finishing of the bond coat. The main evidence of the efficiency of the previously mentioned bond coat surface preparation technique is ceramic layer service lifetime on a blade surface. Comparative analysis of a large body of test results for TBCs after various bond coat surface preparation processes on the basis of this characteristic is a difficult task, because a service-life characteristic depends on many other factors. On designing a bond coat preparation technique, one should first take into account the requirements of a TBC service life. In addition, all the procedures of service preparation should be reproducible for each blade in a big lot. As a rule, any complication of a technological process makes it less stable and reproducible. The choice of the surface preparation technique also depends on bond coat composition and its deposition technique. Mechanical finishing of a bond coat deposited by diffusion techniques (such as aluminizing, platinum-aluminum application, and etc.), which make it relatively brittle, should not entail surface flaking and cracking. For bond coats deposited by low-pressure plasma spray (LPPS) or electric arc technique, mechanical finishing that guarantees the required surface roughness is a mandatory procedure. For bond coats deposited by EB technique, the mandatory procedure is surface compacting. It is also noteworthy that heat treatment used traditionally at different stages of bond coating preparation affects its surface composition. Heat treatment in vacuum at temperatures above 1000 C results in reducing chromium content on the surface, because of its evaporation, and involves the increase in aluminum content (Table 8.6) (Ref 14). This effect can also be purposely used for improving surface homogeneity before its preoxidizing.


Table 8.6

Effect of technological factors on element contents of bond coat surface Element contents of the surface, wt%

Bond coat treatment As-deposited Annealing at 1050 oC for 2 h Finishing with steel microballs Annealing at 1050 oC for 2 h Abrasive finishing

Ni

Co

Cr

Al

Y

48.2 58.3 58.4 60.9 60.4

19.4 14.2 14.7 13.2 13.1

20.5 8.5 9.1 6.7 6.7

11.2 17.2 15.5 17.2 16.8

0.08 0.11 0.09 0.11 0.09

Source: Ref 14

Pellet Evaporation and Ceramic Layer Formation. There are two variants of the technique for ceramics evaporation and condensation. In the first variant of the technological process, some oxygen is supplied into the evaporator zone that is enough for compensation of its loss in ZrO2-Y2O3 dissociation. It results in forming a white ZrO2-Y2O3 ceramic layer of stoichiometric composition. As a rule, this process is used for ceramic layer deposition on the preoxidized bond coat. In the second variant of the technique, the effects of partial reduction of the ZrO2-Y2O3 ceramics, which takes place at its evaporation, are used for developing diffusion processes of interaction between the condensing ceramic layer and the metal bond coat. It guarantees high adhesive strength of the ceramics/bond coat bonding and long service life of the ceramic layer. For this variant of technique, the bond coat surface should be free from oxides. The results of the application of this variant of the technique are discussed subsequently. Ceramics evaporation conditions, vapor density, and its condensation conditions are set by the evaporator construction used in the EB unit. The EB coaters UE described in Chapter 3, “Technological Processes for Deposition of Protective Coatings to Turbine Blades,” is equipped with flat beam guns for material evaporation. These EB guns form a beam with a cross section Table 8.7 ciation

Products and degree of ZrO2 dissoTemperature, K

Vapor phase composition, vol%

2950

3500

4000

O O2 Zr ZrO ZrO2 Degree of dissociation, ␣

36.60 0.33 0.02 36.95 26.10 0.59

38.42 0.82 0.06 40.00 20.70 0.65

39.67 1.40 0.13 42.00 16.80 0.72

Source: Ref 18

of 100 mm2 on the surface of the evaporated material. Evaporation occurs on scanning the surface of the ceramic pellet 70 mm in diameter with the EB. Its power is within the range of 17 to 25 kW (the voltage is 19 kV; the current is 0.8 to 1.3 A). In this case, the ceramic surface layer melts and forms the liquid bath, with a depth of 2 to 5 mm. Under these conditions, the evaporation rate of ZrO2-8%Y2O3 ceramics is 1 to 2 kg/h, and the rate of condensation on rotating blades is in the range of 0.6 to 2.0 lm/ min. Zirconium oxide dissociation takes place during ceramic pellet evaporation (Table 8.7). Oxygen that is present in dissociation products is partially removed from the evaporator zone under vacuum. The degree of ceramics reduction depends on a large number of technological parameters, such as a bath temperature, an evaporation rate, and vacuum in the unit. An evidence of the partial reduction and the increase in the number of anionic vacancies is a change in the ceramics surface color. After cooling, the color of the solidified bath surface and the layer 5 to 8 mm thick under it is black. It is typical of zirconia with a lack of oxygen. Reduction effects may be enhanced by the use of ceramics condensates doped with NiO that feature much less thermodynamic potential of formation (Ref 18, 19). During heating and evaporation of this system, the reduction of NiO to metallic nickel takes place. Nickel atoms facilitate metallic bonding between the ceramic layer and the metallic bond coat and give high adhesive strength to their interface. Annealing in air of the ceramic layer solidified after evaporation results in its mass gain, due to oxygen uptake. Kinetics of mass change on annealing is the same for both the bath and the condensates (Fig. 8.3). After 30 min of annealing at 700 C, ceramics gain 70% of the oxygen lost. Its full oxidation takes place in 2 h. Total loss of oxygen by the ceramics during evaporation, calculated from mass gain, was 3.0


to 3.5%. At oxidizing annealing, no phase composition change of ceramics has been detected. As the oxygen content increases to its initial level, the ceramics whiten. The phase composition of the black ceramics differs from the phase composition of the ceramic pellet used for evaporation, as mentioned previously. The initial phase composition depends on the ZrO2-8%Y2O3 powder used and the procedure of pellets manufacturing. After annealing at 1350 oC and retarded furnace cooling, the ceramic pellet contains 30% “M” phase, 30% “T” phase, and 40% “C” phase. The phase composition of the black bath after evaporation includes only cubical and tetragonal phases (Fig. 8.4). The research into chemical compositions of the ZrO2-8%Y2O3 bath and the

Fig. 8.3

Kinetics of ZrO2-8%Y2O3 black ceramics oxidation at 700 and 1000 C

Fig. 8.4

X-ray patterns of (1) ZrO2-8%Y2O3 ceramic pellet and of (2) bath surface after evaporation

ceramics condensated from it has shown that both oxides evaporate cooperatively. The difference between the Y2O3 contents of the condensate and the bath is not more than Ⳳ0.5%. The condensed ceramic layer also shows different hues. As microstructure examination has shown, a change in a color of the ceramic layer does not involve a change in its phase composition (Table 8.8). The degree of ceramics reduction and its color also depend on the substrate surface preheating temperature and its position relative to the evaporator. As published data have shown (Ref 20), the degree of ceramic layer blackening increases with the increase in the EB power and the evaporation rate. The increase in the angle of vapor condensation has the same effect (Table 8.9). Temperature conditions of blade heating at ceramic layer deposition affect its properties to a larger extent than at metallic coating deposition. While texture and porosity in the metallic coatings deposited by EB technique result in degradation of their mechanical and protective properties, the presence of porosity and columnar texture that is perpendicular to the substrate surface facilitates ceramics fragmentation on generation of thermal stresses and thus ensures its high thermal stability. On the contrary, dense ceramics provoke generation of high thermal stresses, entailing ceramic layer spallation. Ceramic layer structure depends on the temperature of the surface on which vapor condenses and on blade rotation conditions in a vapor flow. According to the scheme presented in Ref 21 columnar structure of ZrO2 layers forms on substrate heating within a temperature range of 800 to 1350 C. The blade heating temperature used on ceramic layer deposition is conventionally in the range of 850 to 950 C. However, measured substrate temperatures can be substantially different from the temperature of the thin surface layer, where vapor condensation and evaporation heat release take place. This real temperature depends on the conditions of blade surface heating, condensation rates, the parameters of blade rotation/revolution around the evaporators, the space between the blades and the evaporator, ceramic layer thermal conductivity, and optical characteristics of the ceramic layer. The structure and phase composition of the condensed ceramics experiences considerable changes, depending on the substrate heating temperature. X-ray diffraction analysis has shown that the decrease of “T”-phase content


Table 8.8

Structural features of ZrO2-8%Y2O3 ceramic layers of different colors ˚ Lattice parameters, A

Ceramics color

a (C)

a (T)

c (T)

c (T)/a (T)

Phase composition

Light gray Gray Dark gray Brown

5.115 5.115 5.119 5.115

5.136 5.141 5.045 5.159

5.163 5.163 5.169 5.156

1.005 1.004 1.024 0.9994

TⳭC TⳭC TⳭC TⳭC

Note: Cubic lattice was determined by reference to line (620); tetragonal lattice was determined by reference to lines (004) and (206).

and the increase of the “C”-phase content with the temperature increase are general trends (Fig. 8.5). At the same time, the texture gets less pronounced, and at substrate temperatures above 1050 oC, it is no longer observed. These effects are likely to result from variation of temperature gradients at the vapor/condensation surface interface. The use of electron beams for surface heating and temperature control promotes intensive heat Table 8.9 Effect of angle of vapor incidence on substrate on degree of ceramics blackening Angle of substrate with vapor flow direction, degrees 90 70–85 45–70 0–45

Condensed ceramics color Black with luster Black without luster Gradual change in color from black to white White

Note: The substrate temperature is 550 oC; the beam power is 13 kW. Source: Ref 20

X-ray patterns of EB-deposited ZrO2-8%Y2O3 ceramics on substrates heated up to different temperatures. (1) 400 C. (2) 600 C. (3) 800 C. (4) 1000 C

Fig. 8.5

release in the thin ceramic layer. That is why such a heating technique is a method of surface treatment that governs composition, structure, and properties of a thin interlayer between the ceramics and the bond coat (Ref 19). Calculations have shown that at the surface heated with EB having the power of 0.1 kW/cm2, the surface temperature increases by 300 to 500 C during the first 10 to 20 s of ceramics condensation and a ceramic layer 0.5 to 1.2 lm thick forms. An increase in the temperature at the onset of ceramics condensation results in intensification of the diffusion process between the ceramic layer and the metallic bond coat. The analysis of element contents in the ceramic layer adjacent to the bond coat (Table 8.10) has revealed that the bond coat elements diffuse into the ceramic layer to the depth of 10 lm. This results in the formation of a special interlayer between zirconia and the metallic bond coat that guarantees high strength of bonding between the layers with different physical-chemical properties. The formation of the interaction zone at the ceramicsmetal interface is the feature of the previously mentioned variant of EB technique. It ensures an adhesive strength of more than 70 MPa at the ceramic layer/bond coat interface. If the working parameters of blades heated with an EB do not change during the period of ceramic layer formation, it results in formation of thick, dense zones between the ceramics and the metallic bond coat. These layers contain an “M” phase that causes stress rise and ceramic layer spalling immediately after their deposition. The fragment of the blade with partial spallation of the ceramic layer on check annealing at 700 o C is shown in Fig. 8.6. The zone where ceramics spallation is detected has a lustrous black color. It is located on the blade airfoil pressure surface on a thin wall, where surface overheating at EB heating is possible. The research has shown that thick layers (5 lm) of dense, black ceramics formed on that surface area on condensation. It contained bond coat elements,


Table 8.10

Element contents in ceramic surface layer adjacent at ceramics/bond coat interface Distance from surface of bond coat Ni20Cr12AlY, lm

Element Zr, % Y, % Ni, % O, %

6

5

4

3

2

59.1 5.7 3.4 30.9

54.9 5.1 7.2 32.6

49.9 4.5 13.9 31.0

41.5 3.0 20.8 33.1

40.2 3.9 23.6 30.7

Source: Ref 22

Fig. 8.6

(a) Appearance of blade with spalled ceramic coating after check annealing. (b) Structure of interaction zone between ceramics and Ni20Cr12AlY bond coat. Reflected electrons, 800⳯

mainly nickel. On structure analysis, the monoclinic phase that was not detected on the other blade surface areas was found in the phase composition of the dense zone (Table 8.11). The presence of the “M” phase is probably due to high nickel content that degrades the degree of

Table 8.11 Properties of dense zone at ceramics/bond coat interface and tendency to ceramics spalling Failure after annealing at 750 oC, 2 h Not detected Detected

Phase composition

Zone thickness, lm

Outer layer

Dense zone

2 8

CⳭ15T CⳭ50T

CⳭT CⳭM

ZrO2 stabilization. At the same time, dense interlayers 3 lm thick improve adhesive strength of the ceramic layer. They contribute to its service life. To prevent the formation of thick interlayers, it is advantageous to reduce the heating electron gun power at the starting moment of ceramic condensation. The average ceramics condensation rate is a convenient characteristic of the technological process (Ref 23). The increase in a condensation rate results in a corresponding thickness increase of the dense ceramics zone that forms on the bond coat surface at the starting moment of condensation. It also makes ceramic layers less porous and much harder. An average crystallite diameter increases from 1.5 to 2.0 lm at a rate of


1 lm/min to 3 to 4 lm at a rate of 2 lm/ min (Fig. 8.7a, b). Structural changes in the ceramic layer at different condensation rates have an effect on its service life. Cyclic testing is carried out, using the samples 7 mm in diameter, under the following conditions: heating to 1100 C for 4 min, holding at this temperature for 20 min, cooling to room temperature for 6 min. The ceramic layer deposited at the condensation rate of 1.2 to 2.0 lm/min has shown the longest service life (Fig. 8.7c). Ceramic layer structure is affected by the composition of the ceramics evaporated. Surface structures of the ZrO2-8%Y2O3 and ZrO28%Y2O3-5%NiO ceramic layers are shown in Fig. 8.8. Both layers are formed under the same working conditions of evaporator guns, heating guns, and at the same specified substrate temperatures. The ZrO2-8%Y2O3-5%NiO crystal-

Effect of condensation rate on characteristics of ceramic layer. Thickness of dense zone. (a) Thickness. (b) Porosity and hardness. (c) Number of cycles to spallation. Source: Ref 23

Fig. 8.7

lites have diameters that are less than those of the nickel-free layers by a factor of 2 to 3. The use of compositions with higher Y2O3 contents has a similar effect. The crystallites of the ZrO220%Y2O3 ceramic layer formed under the same conditions have diameters 2 to 3 times as large as the diameters of the ZrO2-8%Y2O3 ceramic layer crystallites. These effects are caused by the ceramics composition effect on its melting temperature and by condensation conditions on the bond coat surface. Blade surface temperatures and condensation rates affect the thermal stress level in the ceramic layer. This depends on the temperature of the surface on which condensation occurs and on the difference between TECs in the temperature range from condensation temperature to room temperature. Stresses in blade airfoil pressure and suction surface layers depend on the actual heating conditions on these surfaces (Table 8.12). The need for maintaining the parameters of condensation and ceramic layer crystallite growth at a steady level requires a certain layout of relative positions of the blades, the vapor generator, and the EB guns for blade heating. As is shown in Ref 24, substrate rotation speeds have the same effect as temperatures. This behavior is caused by the effect of rotation on the time of growing crystal presence in the zones with different vapor density. The higher the temperature and rotation rate, the larger the diameter of an individual crystallite of the condensing ceramics (Fig. 8.9). Using different rotation speeds, structural characteristics of the ceramic layer can be governed. From the experience of ceramic layer deposition and taking into consideration intricate blade profiles and a need for simultaneous coating deposition on several blades, the best results can be achieved by combining blade revolution around the evaporator and rotation about their axes. An illustration of blade arrangement and their revolution/rotation is given in Fig. 8.10. The fixture in use revolves in the vapor flow with the speed of 12 rpm. At each fixture revolution the blades additionally revolve once around the fixture axis. The choice of blade rotation conditions depends also on the requirements to the ceramic layer thickness and its spread over the blade surface. A ceramic layer fracture pattern at the temperature of 850 to 950 C is shown in Fig. 8.11. The ceramic structure features the pronounced texture of growth perpendicular to the surface.


Fig. 8.8

Ceramic layer surfaces. (a) ZrO2-8%Y2O3. (b) ZrO2-8%Y2O3-5%NiO. 5000⳯

Table 8.12 Thermal stresses in ZrO2-8%Y2O3 ceramic layer on turbine blades. Ceramic layer is deposited on blades heated to 850 oC Suction surface Vapor deposition rate 1 lm/min 2 lm/min 4 lm/min

Pressure surface

Phase composition

Texture

Stress, MPa

Phase composition

Texture

Stress, MPa

FⳭT FⳭT F

[100] [100][111] [100]

30 130 87

FⳭT F T

[100] [100] [100]

76 113 152


Fig. 8.9

Schematic representation of substrate temperature and rotational speed effect on columnar microstructure evolution of EB ceramic layer. Tm, melting temperature. Source: Ref 24

Some individual ceramics crystals are preferably oriented in a [100] direction. Their diameters are in the range of 0.6 to 1.2 lm. They do not vary much along the full crystal lengths. Ceramic layer crystallites should have high cohesive strength and withstand an attack of a high-temperature gas flow. That is why the ceramic evaporation process feature is a requirement to its continuity. Unlike metallic bond coat deposition, in which no process interruption is

Scheme of cassette for blade fixation and revolution. (1) Cassette. (2) Blades. (3) Ceramics pellets. (4) EB gun of evaporator. (5) EB gun for blade heating

Fig. 8.10

harmful for the coating quality, any interruption of ceramic layer deposition forms an additional boundary in the ceramics. The strength of this boundary is much lower than the crystallite strength. Thus, under these conditions, the ceramic layer will never meet the requirements of its properties. In the case of any pause in ceramics evaporation, all the lot of blades being coated are rejected and sent to ceramic layer removal procedure, followed by its redeposition. When ceramic layer deposition is carried to its completion, the blades are removed from the unit and passed to heat treatment. This includes two steps: annealing in a vacuum furnace at 1000 to 1050 C for 2 h, followed by annealing in air at 750 C for 4 to 6 h. The first annealing improves the properties of TBCs formed by the technique that excludes bond coat preoxidizing. Ceramic layer oxidation in air at 750 C corrects its composition toward the stoichiometric composition. Its color changes from dark gray to white. Two-step annealing does not change ceramic layer structure and phase composition. Check operations in the TBC quality control include a visual inspection to guarantee that its surface is free from ceramics droplets; measurements of ceramic layer thickness in the specified blade zones; and a bend test of a flat check sam-


Fig. 8.11

Fractures of ZrO2-8%Y2O3 ceramic layers. (a) 850⳯. (b) 3000⳯. Substrate temperature, 850 to 950 C; rotational speed, 12 rpm

ple on the radius of 3 mm to assess its adhesive strength. On its bending to the angle of 90, ceramic layer spallation is prohibitive (Fig. 8.12). Some cracking of the ceramic layer is allowed. Then, the check samples 10 mm in diameter are subjected to oxidizing annealing at 1100 C. No

ceramic layer spallation should occur within the check annealing period.

Thermophysical Properties of Condensed Ceramics Thermal barrier coating application efficiency depends on ceramic layer thermal conductivity, which determines the cooled blade temperature drop and corresponding increase in its service life. To measure thermal conductivity of a TBC ceramic layer, a laser flash method is used (Ref 25–26). The method is based on irradiating the surface of a flat sample surface with an energy pulse, followed by recording a temperature rise on its backside (Fig. 8.13). Thermal diffusivity and heat capacity are determined experimentally using the pulse method. Then, thermal conductivity (k) is calculated from these characteristics: k ⳱ a • q • Cp

Samples after bend testing: pale sample with Ni20Cr12AlY bond coat, dark sample with ZrO28%Y2O3 /Ni20Cr12AlY (140/80 lm)

Fig. 8.12

in W/m • K, where a is thermal diffusivity (cm2 /s); q is density (g/cm3); Cp is heat capacity (J/g • K). Measuring each thermophysical characteristic is an independent task. The most developed method is that of thermal diffusivity calculation, because the main formula for thermal diffusivity includes only one experimentally measured parameter. It is a period for the temperature to


reach half of its maximum level: a ⳱ 0.1388

d2 s1/2

in cm2 /s, where d is sample thickness, and s1 / 2 is the time required for the temperature of the sample backside to reach the level equal to onehalf of the maximum temperature. Coefficient 0.1388 corresponds to an ideal case when the following conditions are met: instantaneous and uniform heat pulse, heat pulse absorption in a thin surface layer, and no heat losses. For experimental thermal diffusivity determination, one should know neither absolute temperatures nor parameters of a heat flow affecting a sample. Measuring heat capacity by the flash technique, especially for coated samples, is a much more complicated task. Analysis of thermal diffusivity and thermal conductivity of ceramic coatings are discussed elsewhere (Ref 28–30). For thermophysical studies of ceramics condensates, the TC-3000H unit manufactured by the Sinku-Riko Company was used. A ruby laser with a wave length of 6.943 lm was used as an energy source, and as a temperature pickup on the backside of the sample, either a thermocouple (Pt-PtRo) or an infrared sensor was used (Ref 31). The tested sample is essentially a flat disc 10 mm in diameter and 0.8 to 2 mm thick. When thermal diffusivity is studied in this unit, two types of experimental errors are possible. The first type of errors results from some

Fig. 8.13

lack of information on the parameter values used in the design formulas. They are due to the available accuracy of sample thickness and time of s1 / 2 measurements, exactness of the maximum temperature rise assessment, and of catching the moment of the sample irradiation start. These errors are covered in detail in Ref 32 and 33. On the basis of the results reported in the literature, one can deduce that, with the modern data collection systems used, the contribution of this type error does not exceed 0.5%. The second type of errors is due to the difference between the experimental conditions and assumptions in the mathematical model used for calculating thermal diffusivity and heat capacity. These errors are related to the finite pulse duration and its spatial inhomogeneity, to heat losses (due to irradiation, mainly), and to violation of pulse absorption conditions in the thin surface layer. These errors may be avoided by using certain corrections (Ref 34–36). For the TC-3000H unit, pulse duration and spatial inhomogeneity errors determined according to the Sinku-Riko Company recommendations are unessential (less than 1%). Heat losses in the experiment result in a quick temperature rise to its maximum and then a sharply defined smooth temperature decrease. The main cause that gives rise to measurement errors is radiation heat exchange, whose effect rises simultaneously with a temperature rise. The errors caused by radiation may account for 30%. To meet the requirements of pulse absorp-

Scheme of apparatus for measuring thermophysical properties by flash method. Q, heat flow; DT, change in temperature of sample backside due to its frontside irradiation; s1 / 2, time to temperature rise to DT/2


tion in the thin surface layer, the ceramic samples, which are partially transparent, were coated with a thin layer (10 to 12 lm) of the NiAl intermetallic compound (20% Al). This layer ensured steady surface optical parameters of the samples as well. To avoid the errors of heat capacity measuring, which are possible in the flash method, the HT-1000 calorimeter, manufactured by the Setaram Company, was used to determine heat capacity of the ceramic powders and condensates within the temperature range of 20 to 800 oC. The maximum relative errors of test measurements for thermophysical properties account for 7% for heat capacity and 4% for thermal diffusivity. To study thermophysical properties of the condensates, three systems have been chosen (Ref 22). These are ZrO2-8%Y2O3; ZrO28%Y2O3-5%NiO, and ZrO2-20%CeO2. A NiO dopant makes a ceramic layer optically opaque. The ceramics with a CeO2 dopant may be considered as promising for service in media containing sulfur compounds. The initial ZrO28%Y2O3 powders were produced by the hydroxide method. The ZrO2-20%CeO2 composition was produced by mechanical powder mixing, followed by annealing at 1350 C for 10 h. The ceramic pellets shaped as cylinders 50 mm high and 70 mm in diameter, produced by the previously mentioned method, were used for EB evaporation. Ceramics condensates were deposited on the substrates of the JS6U superalloy in the ES 1/ 3/60 (Leybold) unit. The substrate-measured temperatures were 800 to 900 C; the condensation rate was 8 to 10 lm/min; and the substrates did not rotate in the vapor flow. After deposition, the ceramic layers were separated from the substrates and annealed in air at 700 C for 2 h. The characteristics of the condensates deposited are presented in Table 8.13. Heat capacity versus temperature, plotted using the experimental data measured in the HT1000 calorimeter, is presented in Fig. 8.14. The

Table 8.13

ZrO2-8%Y2O3 and ZrO2-8%Y2O3-5%NiO powders and condensates deposited from them were used for measurements. Before measurements, the powders were annealed at 1350 C for 10 h. The main trend for all the compositions is a gradual heat capacity increase with a temperature rise. At 800 C, heat capacity is 0.6 J/g • K. As far as heat capacity is concerned, condensate behavior does not differ from that of the initial powders (Fig. 8.14b). All experimental heat capacity measurements were made in the temperature range of 20 to 800 C. To assess heat capacity values at temperatures above 800 C, extrapolation of the experimental results was done. It was based on the effect of slight gradual increase in zirconia heat capacity with the temperature rise reported in Ref 37. The results of thermal diffusivity measurements and calculation results for ceramics condensate thermal conductivity are given in Table 8.14. It also includes experimental data on heat capacity and thermal diffusivity. The results of heat capacity measurements made on the powders were used for calculations, because, in this case, one can avoid errors of measuring sample characteristics inevitable at measuring heat capacity by the flash method. Heat losses caused by radiation effects in the pulse method of thermal diffusivity measuring were calculated using the method described elsewhere (Ref 33). Correction coefficients for the ZrO2-8%Y2O35%NiO composition are given in Table 8.15. The values of correction coefficients decrease with the temperature rise. They do not go above 13%. Thermal conductivity values measured for the electron beam-physical vapor deposition (EB-PVD) condensed ZrO2-8%Y2O3 ceramics are similar to thermal conductivity of the ZrO210%Y2O3 single crystal at room temperature, which is 3 W/m • K (Ref 9). High thermal conductivity values typical of the ceramic layers deposited by vapor condensation are attributed to their specific structure. Ceramic layer crystallites are oriented perpendicularly to a condensation surface. Insignificant

Materials for investigations

Specimen ZrO2-8%Y2O3 ZrO2-8%Y2O3-5%NiO ZrO2-20%CeO2 ZrO2-8%Y2O3

Thickness, lm

Method of manufacturing

Density, g/cm3

Color

1000 815 900 1100

EB EB EB APS

5.58 5.42 5.66 5.00

White Dark gray Pale pink White

EB, electron beam; EB-PVD, electron beam-physical vapor deposition; APS, air plasma spray


Fig. 8.14 Table 8.14

Ceramics heat capacity. (a) Powders. 1, ZrO2-8%Y2O3; 2, ZrO2-15%Y2O3; 3, ZrO2-8%Y2O3-5%NiO. (b) Condensates. 1, ZrO2-8%Y2O3; 2, ZrO2-8%Y2O3-5%NiO

Thermophysical properties of ceramic specimens Temperature, oC

Specimen

Characteristics

200

400

600

800

1000

1200

ZrO2-8%Y2O3, EB

Cp, J/g • K a ⳯ 102, cm2/s a, W/m • K Cp, J/g • K a ⳯ 102, cm2/s k, W/m • K Cp, J/g • K a ⳯ 102, cm2/s k, W/m • K Cp, J/g • K a ⳯ 102, cm2/s k, W/m • K

0.40 1.18 2.63 0.56 0.70 2.12 0.52 1.05 3.09 0.40 0.42 0.84

0.54 1.07 3.22 0.52 0.61 1.72 0.58 1.05 3.45 0.54 0.40 1.08

0.55 1.05 3.22 0.56 0.60 1.82 0.58 0.95 3.12 0.55 0.39 1.07

0.60 1.04 3.48 0.61 0.65 2.15 0.58 0.95 3.12 0.60 0.40 1.20

0.62 1.03 3.56 0.66 0.63 2.25 0.6 0.88 2.99 0.62 0.41 1.27

0.70 1.04 4.06 0.67 0.62 2.25 0.6 0.9 3.06 0.70 0.44 1.54

ZrO2-8%Y2O3-5%NiO, EB ZrO2-20%CeO2, EB ZrO2-8%Y2O3, APS

Note: Corrections for heat irradiation losses excluded. Source: Ref 22

porosity (10 to 15%) brought about by substrate rotation in the vapor flow is oriented in the same direction. This porosity cannot ensure heat resistance similar to that demonstrated by the ceramic layers formed by APS. In addition, the increase in the oriented porosity fraction is detrimental for a ceramic layer service life, because of easier oxygen penetration to the bond coat surface through the ceramic layer. Research into thermal diffusivity of solid nonmetal materials has shown that this characteristic depends on the presence of point defects in the crystalline lattice of the material and on its strucTable 8.15 Correction factors for irradiation losses, K ⴔ a/ao Temperature (T), oC 700 840 930 1170

K 0.949 0.924 0.920 0.872

tural characteristics (Ref 38). As mentioned previously, structural characteristics of condensed ceramic layers depend on many technological factors of their formation. Peculiarities of technological processes and corresponding structural difference between ceramics layers deposited under different conditions have more significant effect on their thermal diffusivity than the errors due to salient features of measuring techniques. This is likely to be an explanation of the spread in thermal diffusivity and thermal conductivity data for TBC-EB reported by different authors. In this case, structural difference between the ZrO2-8%Y2O3 and ZrO2-8%Y2O3-5%NiO condensates was characterized by the nontextured condensate with a NiO dopant and predominant “T” phase in its phase composition. In the ZrO28%Y2O3-5%NiO ceramics crystallites, some precipitation was detected; it may be NiO precipitate. In addition, NiO has an effect on the ceramic layer structure, thus reducing crystallite size. It is likely that NiO effect on thermal con-


ductivity is due to both additional vacancies in the ZrO2 lattice and structural peculiarities of the ZrO2-8%Y2O3-5%NiO ceramic layer caused by its deposition technique in use. Thermophysical properties of a TBC, including a ceramic layer with a NiO content of 5%, were studied independently in a few laboratories. According to their data, thermal conductivity of the 80 to 200 lm thick ZrO2-8%Y2O3-5%NiO ceramic layers was 0.8 to 1.2 W/m • K in the temperature range of room temperature to 1200 C. To reduce thermal diffusivity of EB-deposited ceramic layers, some artificially introduced point defects may be used (Ref 22, 38). For this purpose, point defects of a crystal lattice can be used. They include foreign atoms in the lattice nodes instead of inherent ones (substitutional atoms), vacancies, interstitial atoms, or combinations of these defects. Phonon scattering by the boundaries may also be considered as an effective way of thermal conductivity reduction. For example, if a polycrystalline material with the size of 1 mm and grain size of 1 lm is taken instead of a single crystal with the size of 1 mm, its thermal conductivity at low temperatures is 1/1000 of single-crystal thermal conductivity (Ref 38). Thus, the technique of ceramic layer deposition may have an impact on its thermal conductivity. On doping ZrO2 with cations exhibiting a valence other than zirconium, additional vacancies form to keep a crystalline cell electrically neutral. They increase heat resistance of the crystal lattice (Fig. 8.15). As experiments have shown, NiO dopants make it possible to reduce thermal

Fig. 8.15

conductivity of the electron beam-physical vapor deposition ZrO2-8%Y2O3 ceramic layers. Doping the ZrO2 crystal lattice with ions having larger or smaller ion radii (such as ytterbium or erbium) produces the same effect. The results of thermal conductivity studies for different ceramic coatings formed by the EBPVD technique are presented in Ref 39. On the basis of the studies (Ref 40, 41), a two-zone model of a ceramic layer is suggested. Thermal conductivity of a dense, inner ceramic zone that forms at the starting moment of condensation is much lower than thermal conductivity of an outer zone (Fig. 8.16). This effect is attributed to the presence of numerous boundaries in the dense zone. Therefore, for reducing thermal conductivity of EB-TBC ceramic layers, it is advantageous to form thin layers 0.2 to 2.0 lm thick in the crystallites. Their boundaries ensure effective phonon scattering. Multilayer structure may be formed by plasma discharge to vary the density of the ceramics during deposition. According to the research, the efficiency of thermal conductivity reduction by means of multilayer structure may be as high as 37 to 45%. The investigation also included the analysis of doping the ZrO2-8%Y2O3 ceramics with erbium, neodymium, gadolinium, ytterbium, and NiO. These elements reduce ceramic layer transparency and facilitate formation of point defects in the ZrO2 crystal lattice. The efficiency of thermal conductivity reduction with these dopants may be as high as 52%. At room temperature, thermal conductivity values for ceramic layers 250 lm thick, with neodymium and ytterbium

(a) Pattern for ZrO2 lattice cell and (b) ZrO2 lattice cell with NiⳭ2 cation


dopant contents of 4 mol% each, are 0.86 and 0.89 W/m • K, respectively. The results of thermal conductivity measurements made on the ceramic layers deposited by the technique used in the research led to the conclusion that the efficiency of the NiO dopant is not as high as it was earlier supposed. For efficient temperature control of surfaces with deposited ceramic layers, their emissivity factor is also required. The integral emissivity factor of the condensed ceramic layers was determined in the unit, using a black body model. All measurements were made in a quiescent air atmosphere under isothermal conditions (Ref 22). Measurement error did not go above 5%. Test results are presented in Table 8.16. Some insignificant difference in emissivity factors for the Ni20Cr12AlY coating variants with different surface preparation is due to the fact that the coating contains up to 60% NiAl. For intermetallides, the effect of surface roughness on the emission is insignificant. The ZrO2Y2O3 ceramic layer deposition changes optical characteristics of the surface. For ceramic layers 10 lm thick or thinner, the increase in a test temperature involves slight decrease of emissivity factor. This results from partial transparency of the ceramic layer in the infrared spectrum range and from the substrate effect. As ceramics layer thickness increases to 100 lm, emissivity factor also increases. On calculating TBC thermal conductivity, knowledge of thermophysical characteristics of metal bond coat materials is required. These are presented in Table 8.17. Measurements were made on condensates formed by EB evaporation of the alloys in question. Thermal expansion coefficients were determined using the technique described in Chapter

7, “The Effect of Protective Coatings on the Mechanical Properties of Superalloys.” Measurements were made on the samples produced by compacting corresponding powders. After compaction, the samples were fired at 1700 C for 10 h. Test results are presented in Table 8.18. Thermal expansion coefficient measurements for the ZrO2-8%Y2O3 ceramics in the temperature range of 400 to 800 C cover the T ↔ M phase transformation. The results of TEC measurements made on ceramics condensates are shown in Fig. 8.17. Oxidation of the initial black ceramics during the first heating results in its volume increase. On repeated measurement, TEC values get stable and do not vary much.

TBC Ceramic Layer Durability The moment of ceramic layer spallation from a cooled blade surface may be considered as a TBC failure, because after spallation, its main duty of protecting blade walls against high-temperature attack is no longer fulfilled. Thermal barrier coating ceramic layer cracking and formation of different-sized fragments are evidence of its response to blade surface stresses and strains but not evidence of the lack of TBC protective properties. Main Causes of TBC Ceramic Layer Damage. A TBC ceramic layer failure on its deposition by an optimized and reproducible technique may be caused by the following different factors: ●

● ● ● ●

Scheme of an EB-TBC and corresponding variations of thermal conductivity of a two-layer model. Source: Ref 39

Fig. 8.16

Ceramic layer/metallic bond coat boundary oxidation, formation, and growth of thermally grown oxide (TGO) on this boundary Changes in ceramic layer phase composition; recrystallization and sintering Hot corrosion of the ceramic layer Erosion of the ceramic layer Deformation of a bond coat and a blade under the action of thermal stresses and centrifugal force

Each of these processes can occur separately and in combination with other processes. According to a large body of test data, the weak point of TBCs is a boundary between the ceramic layer and the heat-resistant bond coat (Ref 42, 43). During high-temperature testing or under service conditions, oxygen from the oxidizing environment (air, combustion products) penetrates to the ceramic-metal boundary. Its penetration to a bond coat surface occurs in two


Table 8.16

Coating emissivity Temperature, oC

Coating

Preparation of bond coat surface

Ni20Cr12AlY

Polishing Abrasive blasting Polishing

ZrO2-11%Y2O3

Thickness, lm

20

200

400

600

800

100 100 2 5 10 2 5 10 100

0.23 0.27 0.54 0.77 0.79 0.79 0.86 0.81 0.80

0.25 0.30 0.50 0.62 0.65 0.66 0.70 0.71 0.80

0.27 0.32 0.45 0.48 0.50 0.52 0.60 0.57 0.79

0.29 0.37 0.43 0.46 0.47 0.49 0.60 0.52 0.77

0.35 0.46 0.46 0.47 ... 0.52 0.60 0.56 0.76

Abrasive blasting

Source: Ref 22

Table 8.17

Thermophysical properties of JS32 superalloy and condensed alloys for coatings Density (q), g/cm3

Temperature (T ), oC

Heat capacity (Cp), J/g • deg

Thermal diffusivity (a ⴒ 102), cm2/s

Thermal conductivity (k), W/m • K

JS32

8.60

Ni20Co20Cr12AlY

7.05

Ni20Cr12AlY

7.60

Co20Cr12AlY

7.10

Ni8Co20Cr12AlY

7.40

Ni20Cr6AlY

7.30

20 400 600 800 1000 1200 20 600 800 1000 1200 20 400 600 800 1000 1200 20 600 800 1000 1200 20 600 800 1000 1200 20 400 600 800 1000 1200

0.40 0.44 0.49 0.55 0.64 0.78 0.50 0.57 0.65 0.71 0.93 0.38 0.40 0.42 0.45 0.53 0.65 0.50 0.62 0.67 0.73 0.90 0.44 0.48 0.54 0.66 0.76 0.39 0.43 0.46 0.50 0.58 0.73

2.05 3.00 3.55 4.00 4.00 4.00 2.90 5.40 5.50 5.55 5.80 5.40 5.80 6.00 5.90 5.60 4.90 3.20 4.58 4.93 5.10 5.60 3.98 4.79 5.38 5.20 5.22 3.30 4.70 5.40 6.10 6.50 7.30

7.05 11.35 14.96 18.92 22.02 16.83 10.22 21.70 25.20 27.78 38.03 15.60 17.63 19.15 20.18 22.56 24.21 11.36 20.16 23.45 26.43 35.78 12.96 17.01 21.50 25.40 29.36 9.40 14.75 18.13 22.27 27.52 38.90

Trademark

Courtesy of Dr. S.V. Zherzdev

Table 8.18

Thermal expansion coefficients for ceramics sintered at 1700 oC for 10 h Coefficient of thermal expansion (␣ ⴒ 106), Kⴑ1

Ceramics ZrO2-8%Y2O3 ZrO2-15%Y2O3 ZrO2-7%Y2O3-5%NiO ZrO2-7%Y2O3-15%NiO Courtesy of Dr. S.V. Zherzdev

20–400

400–600

600–800

800–1000

1000–1200

11.90 8.50 11.00 11.20

20.30 9.00 11.10 11.30

2.30 10.30 11.10 11.30

8.00 10.30 11.40 11.30

7.00 10.50 10.50 10.50


ways: by gas transfer through open porosity of the ceramic layer and by oxygen ion diffusion through anion vacancies in the ZrO2 lattice. Depending on the temperature and ceramic layer state, the contribution of these mechanisms may be different. It is noteworthy that oxygen demonstrates very high diffusive mobility in the ZrO2 lattice (a few orders of magnitude higher than the Al2O3 lattice) (Ref 44). It may set limits on the use of ZrO2 in TBCs at temperatures above 1100 C. Oxygen penetration through the ceramic layer results in bond coat oxidation. Oxides form on its surface. Their composition and structure depend on the amount of oxygen reaching the surface and on the bond coat composition. As a rule, the oxides forming on the bond coat surface

Thermal expansion coefficient of ZrO2-8%Y2O3 ceramics condensates. (1) Black (no annealing). (2) After annealing in air at 1000 C for 2 h. 䡩, along crystallites; ●, across crystallites

Fig. 8.17

Service lives of nickel superalloy samples 7 mm in diameter with TBC. 1, ZrO2-8%Y2O3 / Ni23Co20Cr11/5Al0.26Y; 2, TBC-EB (references); 3, TBC-APS (references). Source: Ref 50

Fig. 8.18

have Al2O3 structure, but some spinel may also form there. Oxide formation and growth on the bond coat surface was independent of the surface preparation technique in use. It did not matter whether it was preoxidized or if ceramics condensed on the surface free from oxides. Formation and growth of oxides on the ceramics/ bond coat interface result in additional stress generation, degradation of ceramic layer adhesion, and ultimately, in its spallation. Ceramics spallation takes place even if no stresses are generated by outside forces in a sample after hightemperature testing. Ceramic layer spallation is caused by tearing stresses arising on the convex surfaces of samples and blades (Fig. 9.16). These stresses are induced by compressive stresses, which arise in the ceramic layer on sample cooling from the temperature equal to or below its condensation temperature. Compressive stress rise is caused by the fact that the TEC of ceramics is much less than the TEC of the superalloy. The stresses of ceramic layer tearing reach their maximum at cooling to room temperature. Bond coat surface oxidation and tearing stress of ceramic layer are the main factors causing TBC damage on lab testing and under service conditions. The higher the temperature, the quicker oxidation and ceramic layer spallation occur. To assess a TBC service life, it is advantageous to use a criterion of TGO layer thickness that involves ceramic layer spallation (Ref 45) and a criterion of adhesive strength of the ceramic layer/bond coat interface (Ref 46). The latter criterion assumes that the adhesive strength of the ceramic layer/bond coat interface should be higher than the stress of the ceramic layer tearing from the blade surface over the specified period of blade service (Chapter 9, “Some Principles of Strength Designing for Turbine-Blade Protective Coatings”). Ceramic layer service life is affected by inhomogeneity of ceramics phase composition, ceramics crystallite recrystallization processes, and sintering at temperatures above 1000 C. On evaporating and vapor depositing ZrO2-8%Y2O3 ceramics, the ceramic layer is fixed in its hightemperature metastable state. Under certain conditions in the temperature range of 500 to 1200 o C, the high-temperature “C” and “T” phases dissociate in the ceramic layer, and the “M” phase may appear. The T ↔ M phase transformation occurs martensitically, with the volume expanding by 3 to 12%. This results in higher compressive stress in the ceramic layer and in


Table 8.19

Characteristics and service lifetime of samples tested

Alloy

Bond coat thickness, lm

Ceramics thickness, lm

Ceramics condensation rate, lm/min

Ceramics density, g/cm3

Ceramics hardness, MPa

Service lifetime, h

IN625 Hast X IN625 Hast X IN625 IN625 Hast X Hast X

100 130 110 105 140 150 120 120

125 130 165 170 190 95 150 185

2.9 2.9 3.56 4.25 3.28 2.38 3.75 4.40

4.72 4.69 4.73 ... 4.32 ... ... ...

5085 5260 4870 6530 4484 5300 ... 5772

2120 831 1047 832 497 945 1023 1722

Source: Ref 50

the corresponding increase of the stress, causing its tearing from the bond coat. Ceramics sintering at temperatures above 1000 C results in gradual ceramic layer densification, in shrinkage effects, and also in gen-

Fig. 8.19

Thermally grown oxide (␣-Al2O3) thickness versus test time at 1135 C. Source: Ref 50

Fig. 8.20

Effect of TBC ceramic layer thickness on its service life (sample diameter of 7 mm). Source: Ref 51

eration of additional tearing stresses of the ceramic layer on cooling. Sintering effects depend on ceramic layer composition (Ref 47). The addition of NiO to the ceramics composition enhances sintering effects. The development of the ceramics recrystallization processes in TBC-EB ceramic layers at high temperatures also results in reducing the strength of some crystallites and in worsening their resistance to damage by erosion on gas flow attack. The ceramic layer of a TBC does not actually show any ductility. That is why any plastic deformation causes cracks in the ceramic layer, facilitating oxygen access to the bond coat surface. Turbine blades with TBCs including ductile

Service lives of samples 10 mm in diameter with ZrO2-8%Y2O3 /Ni20Co20Cr11/6Al0.25Y (thickness of 130/120 lm). 1 to 9, different sample lots

Fig. 8.21


Table 8.20 Coating TBC-4 RT31/RT33

Coating characteristics Bond coat composition

Ceramic layer

Ceramics thickness, lm

Ni14%Cr12%AlY(a) Co38%Ni18%Cr8.9%AlY

ZrO2-7%Y2O3 ZrO2-7%Y2O3

120 100

(a) Yttrium content ⱕ0.1%. Source: Ref 52

Ni20Co20Cr12AlY bond coat have shown bond coat deformation at temperatures above 1100 o C. In these areas, the ceramic layer was easily damaged by gas flow. Ceramic layers of ZrO2-8%Y2O3 deposited by the EB technique feature resistance to erosion that is 3 to 7 times higher than that of ceramic layers deposited by the APS technique (Ref 48). This type of TBC damage is mainly met with on turbine blades of stationary units. The ZrO28%Y2O3 ceramics may be damaged by hot corrosion on exposure to fuel combustion products containing sulfur compounds and some other aggressive agents. Mechanisms of these processes are discussed elsewhere (Ref 49). Lab Tests of TBC Service Life. The main body of publications on the research into TBC durability deals with different high-temperature tests of samples and blades under conditions similar to those of testing for isothermal and cyclic heat resistance. The peculiarity of TBC testing is that one should detect the moment of ceramic layer spallation from their surfaces instead of measuring sample mass gain/loss. Testing cylindrical samples with different diameters is the most advantageous method for TBC durability studies, optimization of their compositions, design, and deposition technique. This is due to

Service lives of TBCs on cylindrical specimens of Hastelloy X samples 6 and 10 mm in diameter. Nf, cycles to spallation of 20% sample surface area. Source: Ref 52

Fig. 8.22

the fact that on cooling the samples from the test temperature, compressive stresses occur in the ceramic layer and induce stresses of its tearing from the cylindrical sample with the radius of R. The magnitude of tearing stress, rncr, is directly proportional to the ceramic coating thickness, hcr, and inversely proportional to the test sample radius, R. rncr

hcr • rscr R

where rscr is ceramic coating compressive stress. As an approximation, the stress in the ceramic layer may be calculated from the following equation: rscr ⳱

Ecr • (␣T ⳮ ␣Tcr) • (T ⳮ To) 1 ⳮ lcr

where T is temperature, To is conventional initial temperature (temperature of ceramic layer deposition), Ecr is the ceramic layer modulus of elasticity, and ␣Tcr and ␣T are, correspondingly, TECs of the ceramics and substrate materials in the temperature range of T to To, and lcr is Poisson’s ratio for the ceramic layer. Ceramic layer spalling from a cylindrical sample surface occurs at the moment when its tearing stress is equal to or higher than the adhesive strength at a ceramic layer/bond coat interface; that is, rncr ⱖ racr. TBC service life depends on ceramic layer composition and thickness. After the choice and optimization of these characteristics for certain turbine blades, it depends on ceramic layer deposition technique. At high-temperature cycling testing, ceramic layer service life shortens with the increase in maximum cycle temperature and rates of heating and cooling of the samples tested. By now, a large body of TBC tests for service life evaluation have been carried out. Unfortunately, comparative analysis of these test results and their use for practical purposes is a difficult task, because of the lack of information on their deposition techniques and salient features of their test conditions.


Examples of testing samples and blades with TBCs deposited using EB units and techniques designed by Paton Electric Welding Institute (Ukraine) are given subsequently. Thermal barrier coatings (including their bond coats) were deposited in EB units of the UE type. Their bond coat compositions were chosen by taking into account the blade service conditions in an engine or a stationary unit. After bond coat deposition, the samples underwent microball peening and recrystallization annealing in vacuum at 1000 to 1050 C for 2 to 4 h. Before deposition of the ceramic layer, the bond coat surface underwent hydroabrasive processing (with fused corundum with a grain size of 30 to 60 lm). Then, the samples were rinsed in an ultrasonic bath. Bond coat surface roughness before ceramics deposition was Ra ⳱ 0.6 to 0.8 lm. Bond coat surface preoxidation was not used. Table 8.21

The samples were arranged in special fixtures and loaded into the unit for ceramic layer deposition, where they were EB heated to a temperature of 900 to 930 C. On reaching this temperature, evaporation of a previously prepared ceramic bath started. The rotation speed of the fixture with the samples in the vapor flow was 12 rpm. After ceramic layer deposition, the samples were unloaded from the unit and annealed under vacuum at 1080 C for 2 h. The ZrO2-8%Y2O3 powders from the same supplier were used for ceramic layer deposition, and the pellets for evaporation were made by the same technique. The results of high-temperature TBC cycling testing are presented in Ref 50. In that case, the metallic bond coat was deposited sequentially from two alloys in a one-step process. First, an inner layer 30 to 40 lm thick was deposited

Service lives of samples 10 mm in diameter Service life, cycles

Coating

High

Mean

Low

Thermally grown oxide thickness, lm

Growth parameter (Kp), g2/cm4 ⴒ s

TBC-4 RT31/RT33

1080 710

460 260

96 19

5.5 7.5

3.9 ⳯ 10ⳮ13 8.4 ⳯ 10ⳮ13

Source: Ref 52

Fig. 8.23

Structure of surface injured by spallation. (a) Ceramic layer side. Dark surface, Al2O3; pale surface, ZrO2-8%Y2O3. (b) Bond coat side. Pale surface, bond coat; dark surface, Al2O3. 910⳯


Structure of TBC bond coat on CMSX-4 superalloy. 1000⳯. Element contents in points 1 to 5 are given in Table 8.22

Fig. 8.24

Corrosion lives of EB-TBC (1, 2, 3) and Co22Cr12AlY coating (4, 5, 6) in ash of gas-turbine fuel at different temperatures. 1 and 4, 750 C; 2 and 5, 800 C; 3 and 6, 850 C. Source: Ref 51

Fig. 8.25

from the Ni20Co20Cr5AlY alloy, and then, an outer layer 60 to 80 lm thick was deposited from the Ni20Co20Cr12AlY alloy. Yttrium content of the bond coat was controlled, and it was within the range of 0.2 to 0.26%. The layer with low aluminum content was intended to improve bond coat ductility. Deposition of the ceramic layer involved the formation of a dense zone 2 to 3 lm thick and enriched with nickel, chromium, and cobalt due to ceramic layer and bond coat interaction. The samples were tested in air in a 1710BL unit under the cycling conditions of 50 ↔ 1135 C, a sample cooling rate of 8 to 10 min, and a holding time of 1 h at the maximum temperature. The information on the samples tested and their test results are presented in Table 8.19. Service life characteristic was taken as a period (number of cycles) to the onset of ceramic layer spalling. Considerable spread in the experimental results is attributed to the problems of technological parameter controlling on TBC deposition. The spread range is a characteristic of technological process reproducibility. The best results are typical of the achievable level of TBC properties on optimizing all the TBC deposition process parameters. It is noteworthy that each product type should be manufactured using its optimal technological process, because even insignificant variations of a sample or a blade shape or mass entail some changes in the conditions of its heating and vapor flow condensation. The test results for the best samples of the previously mentioned work, together with the results known from references, are given for comparative analysis in Fig. 8.18. Ceramic layer spalling from the samples at their testing is caused by TGO (Al2O3) that forms and grows on the bond coat surface (Fig. 8.19). In the author’s opinion, the optimal TBC characteristics for the said deposition technique, bond coat, and ceramics compositions are as follows: bond coat thickness of 110 to 135 lm, ceramic layer thickness of 135 to 140 lm, and condensation rate of 4 lm/min. In the case of thicker or thinner TBCs, their service lives shorten. This effect of ceramic layer thickness on its service life under different test conditions has already been mentioned in Ref 51. The TBC-EB service life on JS6U-type superalloy versus ceramic layer thickness is presented in Fig. 8.20. The bond coat layer 90 to 120 lm thick was deposited from the Co22Cr12Al0.4Y alloy; the ceramic layer was deposited from the ZrO2-


8%Y2O3 pellet. Cycling conditions were as follows: heating to 1100 C for 4 min, holding at this temperature for 20 min, and cooling to 40 to 50 C for 6 min. The TBC service life is determined as a period to ceramic layer spalling from 50% of its surface. Maximum service life was demonstrated by the TBC including the ceramic layer 100 to 140 lm thick. General decrease of the number of cycles to ceramic layer spalling in comparison with the previously mentioned data is caused by yttrium content of the bond coat (0.1%Y). Test results for the samples 10 mm in diameter with TBCs deposited by the same technique are presented in Ref 2. The TBC bond coat was deposited sequentially from two alloys in a one-step process. First, an inner layer was deposited from the Ni20Co20Cr5AlY alloy, and then, an outer layer was deposited from the Ni20Co20Cr12AlY alloy. Yttrium content of the bond coat was 0.2 to 0.26%. The outer ZrO2-8%Y2O3 ceramic layer was 120 to 130

Table 8.22

lm thick. The samples were tested under the cycling conditions of 50 ↔ 1100 C, heating for 10 min, and cooling for 6 min. The test results are presented in Fig. 8.21. As compared to the earlier tests, it may be concluded that both reducing the temperature and the number of cycles (30 cycles) and enlarging the sample diameter had an effect on TBC durability. Ceramics spalling from the best samples occurred after 2500 h. An investigation compared two variants of coatings deposited by different techniques (Ref 52). The previously mentioned technique was used for TBC-4 coating deposition. Some minor distinctions in its deposition technique from the previously mentioned were concerned with the deposition of a thin layer (1.5 to 2 lm thick) on the bond coat surface before ceramics evaporation onset. The thin layer composition was identical to the bond coat composition. The RT31/ RT33 coating was used for comparison. The Hastelloy X samples 10 and 6 mm in di-

Composition of bond coat on CMSX-4 superalloy after testing at 1100 C for 120 h Chemical composition, wt%

Analyzed zone (Fig. 8.24) 1 2 3 4 5

Fig. 8.26

Al

Ti

Cr

Co

Ni

Mo

Ta

W

Re

2.5 6.6 9.6 6.5 17.4

0.4 0.6 0.5 0.2 0.2

12.4 10.4 6.7 19.9 8.0

4.5 5.9 2.6 4.2 2.3

29.3 64.1 74.1 65.0 71.7

1.1 0.3 0.1 0.2 0

2.8 6.0 4.8 1.1 0.6

24.7 4.7 1.6 1.2 0.2

20.5 1.4 0.1 0.5 0

Appearance of samples after adhesive strength testing at 1050 C for (a) 200 h and (b) 500 h


ameter were tested. The test conditions were as follows: 20 ↔ 1100 C, heating for 5 min, holding at the maximum temperature for 55 min, and cooling for 15 min. Ceramic layer damage was reported at 20% spalling from the surface. The characteristics of the coatings and the test results are presented in Table 8.20 and in Fig. 8.22. In all cases, ceramics spalling was caused by TGO layer formation between the ceramic layer and the bond coat. Independently of the initial state of a bond coat surface, TGO is forming on its surface under high-temperature test conditions. The thickness and the growth parameter of this layer differ for the coatings deposited by different techniques (Table 8.21). Ceramic layer spalling took place along the ceramics/TGO boundary for the samples with long service lives and along the bond coat/TGO boundary for the samples with short service lives. Short service lives of the samples with the TBC-4 coating may be caused by low yttrium content (0.1%Y). Similar ceramics spalling along the TGO/ bond coat interface was detected on testing the samples of CMSX-4 alloy at 1100 C. The ZrO28%Y2O3 ceramic layer spalled after 250 cycles (cycle of 60 min) of 20 ↔ 1100 C, which was much worse than other test results. The largest part of the spalling surface area from the ceramics inside consisted of Al2O3 (Fig. 8.23a), while from the Ni20Cr12AlY (Y 0.1%) bond coat side, spalling took place along its surface. No oxides were detected on that surface (Fig. 8.23b). The changes in the bond coat are the same as the changes in the coating without the ceramic layer, as discussed previously (Fig. 8.24). A great deal of aluminum from a bond coat is spent on TGO layer formation (Table 8.22). In addition to high-temperature areas of ap-

Fig. 8.27

Variation of adhesive strength of TBC ceramic layer versus test conditions

plication, TBCs may also be advantageous for improving service lives of blades operating in media containing sulfur compounds. In this case, a wholesome effect of the ceramics may be attributed to both blade surface temperature increase and less condensation of aggressive compounds on it as well as to bond coat surface protection from aggressive compound attack. The test results for a few coating modifications with outer ceramic layers are presented in Fig. 8.25 (Ref 51), together with test results for the coating deposited from the Co22Cr12AlY alloy for comparison. In this case, the procedures of bond coat surface preparation and coating deposition are the same as that mentioned previously. The bond coat was vapor deposited sequentially from two alloys, first from Co24Cr5Al0.4Y, then from Co24Cr12Al0.4Y. Bond coat thickness was 90 to 120 lm. The ZrO2-8%Y2O3 ceramic layer thickness varied from 30 to 160 lm. A synthetic ash of gas-turbine fuel was used in the tests. Its composition was as follows: Na2SO4, 66.2%; Fe2O3, 20.4%; V2O5, 1.8%; NiO, 80.3%; and CaO, 3.3%. The ash was applied to the sample surfaces as 120 g/m2 every 250 h. The samples were tested at 750, 800, and 850 C to the moment of corrosive damage of its coating (both the ceramic layer and the bond coat). Ceramic layer spalling on testing for 20,000 h (at 750 C) and for 7500 h (at 800 C) was not reported. In all cases, service lives of the coatings with ceramic layers were longer than service lives of the coatings without ceramic layers. Testing TBC Ceramic Layer for Adhesive Strength. For TBC service life prediction, one may use the adhesive strength characteristic of the ceramic layer/bond coat interface (Ref 53). To do so, strength variations at this interface should be determined as ra ⳱ f(T,s), depending on test time and temperature. Testing for adhesive strength may be carried out by different methods used for evaluation of this property (Ref 54). A commonly used “glue” method is suitable. To study ceramic layer adhesion, a TBC was deposited on a flat surface of a cylindrical sample 20 mm in diameter. The coating consisted of the Ni20Cr12AlY bond coat 80 lm thick and the ZrO2-8%Y2O3 ceramic layer 120 lm thick. Sample heating temperature at ceramics deposition was 850 C. The ceramic layer condensation rate was 10 lm/min. The samples and their TBCs were held under the furnace air atmosphere at temperatures of 1050, 1100, and 1150 C for periods of up to


Table 8.23 Blade test results on gas-dynamic test rig. Test cycling conditions: 200↔1100 oC, heating time is 30 s, cooling time is 30 s Blade lot

Number of cycles at testing

Injury description

1265 650 805 805

Ceramics started to spall after 1100 cycles Insignificant ceramics spallation along a leading edge Insignificant ceramics spallation along a leading edge Undamaged ceramics

A B C D Note: Ceramic layer thickness 120 to 160 lm

500 h. After this, the ceramics surface of one sample glued to the respective metallic surface of the other sample with the same geometrical characteristics. Then, the glued sample was tested for breaking at room temperature. The samples, after testing for adhesive strength, are shown in Fig. 8.26. In their initial state, the samples after ceramic layer deposition always break down along the glued joint, which has a strength of 70 MPa. On testing after holding at high temperatures, the samples break down partially

Fig. 8.28

Turbine blades with EB-TBC after testing in aircraft engine for 1200 h

along the ceramics/bond coat interface and partially along the glue joint (Fig. 8.26b). The higher the test temperature and the longer the test time, the larger the breakage area along the ceramics/bond coat interface. The adhesive strength data presented in Fig. 8.27 are a little lower than their real level, because calculations were carried out on the base of test sample crosssectional areas without due account of the surfaces where ceramic layer breakage occurred along the glue. Ceramic layer phase composition did not change after holding at high temperatures. The oxides with ␣-Al2O3 structure and spinels were detected on the inside ceramics surface adjacent to the ceramics/bond coat interface. The general trend in adhesive strength behavior of the ceramic layer is its drop as test time and temperature increase. TBC Testing on Gas-Dynamic Test Rigs and Engines. Testing on gas-dynamic test rigs makes interpretation and comparison of different test data even more difficult. In addition to tearing stresses generated during slow cooling, thermal stresses are generated on the test rig, due to nonuniformity of the heat pattern along the height and across the cross section of the samples and blades. Thermal stresses and strains result in ceramics fragmentation and facilitate oxygen access to the bond coat surface. The main cause of ceramic layer spalling is bond coat oxidation and the corresponding decrease in adhesive strength of the ceramic layer. Test results on gas-dynamic test rigs for TBCEB with a bond coat 60 to 80 lm thick deposited from the Ni20Cr12AlY alloy and the ZrO28%Y2O3 ceramic layer 120 to 140 lm thick are given in Table 7.18 (Chapter 7, “The Effect of Protective Coatings on the Mechanical Properties of Superalloys”). Thermal cycling is 1100 ↔ 200 C, time required for heating the sample up to the maximum temperature is 60 s, and cooling time is 60 s. Ceramics spalling was detected on the edge with the radius of 1 mm after 1500 cycles (25 h).


Similar results were obtained on testing turbine blades on this gas-dynamic test rig. In Table 8.23, test results are presented for a few turbine blade lots of the MAR-M-002 alloy on cycling that included heating the blades in the flow of combustion products to 1100 C and cooling them to 200 oC with compressed air. Heating and cooling periods were the same (30 s). The examination of the coatings after their testing has revealed the TGO layer formed at the ceramic layer/bond coat interface. The thickness of this layer increases with the increase in the number of cycles. For the best blades, ceramic layer spalling started after 1100 cycles. Maximum information on life and efficiency of TBCs is obtained through testing in engines. However, this has specific features connected with the peculiarities of blade design and testing conditions. To determine the general features of TBC behavior, the results of testing of different engines should be systematized. Tests of aircraft and power-turbine blades with TBCs are dealt with in Ref 55–59. Blade examination after long-term testing in engines reveals that the main TBC damage types are as follows: ceramic layer spalling because of bond coat oxidation in the high-temperature zones, ceramic layer spalling at the punched holes because of bond coat oxidation from the inside perforation space, erosion wear of the layer on the blade shroud due to unfavorable ceramics structure on these surfaces, and cracking in the zones where high thermal stresses arise. Phase transformations in the bond coat are similar to those in the coatings without ceramics, provided that less aluminum content is spent on oxidation. The TBCs (consisting of the Ni20Cr12AlY bond coat 80 to 120 lm thick and the ZrO28%Y2O3 or ZrO2-8%Y2O3-5%NiO ceramic layer 120 to 160 lm thick) were tested in different aircraft engines. Thermal barrier coating lifetime is closely allied with testing conditions, such as maximum temperature, time, and number of transient situations. A blade, after 1200 h of testing in an aircraft engine, is shown in Fig. 8.28. Under extreme conditions, the temperature on some blade surface areas is as high as 1120 to 1150 C. Within the said time, ceramic layer damage was detected only on a shroud surface. Under testing conditions equivalent to 10,000 h of engine operation, TBCs did not show any damage on the coated blades of MARM-200 alloy. The same coating was used. Examination revealed that it was undamaged even

on the surfaces where maximum temperatures were as high as 1100 to 1150 C. As in all hightemperature tests, the TGO layer formed and grew on the ceramics/bond coat interface. All the durability tests of samples and blades demonstrate that the vapor-deposited ceramic layer has some unique properties that allow it to hold on turbine blade surfaces for a long time. Its columnar structure compensates thermal stresses and strains arising at the operating temperatures in the cooled blade walls and on their surfaces. High adhesive strength of the ceramics/bond coat interface keeps ceramics on the bond coat surface on blade cooling. The problems that set limits to a TBC service life are caused primarily by the high-oxygenpermeable ZrO2 lattice. To a large extent, a TBC ceramic layer service life depends on its deposition technique. Optimization of the technological cycle of TBC deposition, with reference to technological equipment available and the TBC composition and design chosen, is the crucial consideration in producing TBCs with long service lives. REFERENCES

1. J. Stringer, The Reactive Element Effect in High-Temperature Corrosion, High Temperature Corrosion 2, R. Streiff, J. Stringer, R. Krutenat, and M. Caillet, Ed., Elsevier Science Publishers, 1989, p 129–137 2. I.S. Malashenko, N.P. Vashchilo, V.A. Belotserkovsky, and K.Y. Yakovchuk, Effect of Yttrium on Functional Characteristics of Vacuum Condensates and Protective Coatings MCrAlY and MCrAlY/ZrO2-8%Y2O3 at Thermocyclic Loading, Adv. Spec. Electrometall., Vol 1, 1997, p 24–33 3. R.A. Miller, R.G. Garlick, and J.L. Smialek, Phase Distribution in Plasma-Sprayed Zirconia-Yttria, Am. Ceram. Soc. Bull., Vol 63 (No. 12), 1983, p 1355–1358 4. R.P. Ingel, D. Lewis, B.A. Bender, and R.W. Rice, Physical, Microstructural and Thermomechanical Properties of ZrO2 Single Crystals, Advances in Ceramics, Vol 12, American Ceramics Society, Inc., 1984, p 408–414 5. S. Stecura, Optimization of the Ni-Cr-Al-Y/ ZrO2-Y2O3 Thermal Barrier System, Adv. Ceram. Mater., Vol 1, 1986, p 68–76 6. D.L. Ruckle and D.S. Duvall, QuenchCracked Ceramic Thermal Barrier Coatings, U.S. Patent 4457948, 3 Aug 1984


7. T.E. Strangman, Columnar Grain Thermal Barrier Coatings, U.S. Patent 4321311, 23 March 1982 8. H.G. Scott, Phase Relationship in the Zirconia-Yttria System, J. Mater. Sci., Vol 10 (No. 9), 1975, p 1527–1535 9. D.S. Rutman and Y.S. Toropov, High-Temperature Zirconium Dioxide-Base Materials, Moscow, Metallurgy, 1985, 137 p 10. V.S. Bakunov, V.L. Balkovich, and A.S. Vlasov, Ceramics of Highly Refractory Oxides, Moscow, Metallurgy, 1977, 304 p 11. I.D. Abramson, Ceramics for Aviation, Oborongiz, Moscow, 1993, p 240 12. A. Karklit, A. Larine, et al., Producing of High-Temperature Strength Materials by Semi-Dry Molding, Moscow, Metallurgy, 1982, p 368 13. K. Fritscher, M. Schmucker, C. Leyens, and U. Schulz, TEM Investigation on the Adhesion of YPSZ EB-PVD TBCs, High Temperature Corrosion 4, R. Streiff, J. Stringer, R. Krutenat, M. Caillet, and R. Rapp, Ed., Trans Tech Publication, 1997, p 965–970 14. I.S. Malashenko, G.C. Marijnissen, V.A. Belotserkovsky, A.N. Shelkovoj, and V.V. Orlenko, Microstructural Peculiarities of Metal/Ceramic Condensation Coatings Under Thermocyclic Loading Conditions, Adv. Spec. Electrometall., Vol 2, 1997, p 32–43 15. T.E. Strangman and P.A. Solfest, Ceramic Barrier Coating with Alumina Interlayer, U.S. Patent 5015502, 14 May 1991 16. B.A. Movchan, Y.E. Rudoj, and I.S. Malashenko, Method of Producing on a Substrate of Protective Coats with Chemical Composition and Structure Gradient Across the Thickness and with Top Ceramic Layer, Ukraine Patent 17473A, 6 May 1997 17. B.A. Movchan and Y.E. Rudoj, Gradient Heat-Resistant Coatings, Produced by Vacuum Electron-Beam Deposition of Vapour Phase, Adv. Spec. Electrometall., Vol 2, 1997, p 25–31 18. I.S. Kulikov, Thermodynamics of Oxides, Moscow, Metallurgy, 1986, 384 p 19. Y.A. Tamarin and S.V. Zherzdev, Technological Aspects of Forming TBC-EB Ceramic Layers, High Temperature Corrosion 5, R. Streiff, I.J. Wright, R. Krutenat, M. Caillet, and A. Cailerie, Ed., Trans Tech Publication, 2001, p 587–593 20. A.V. Demchishin, “Structure and Properties of Thick Vacuum Condensate of Metallic

21.

22.

23.

24.

25.

26.

27.

28.

29.

30. 31. 32.

and Nonmetallic Materials,” Ph.D. dissertation, Kiev, 1980 B.A. Movchan and A.V. Demchishin, Research into Structure and Properties of Thick Vacuum Condensate Layers of Titanium, Tungsten, Aluminum Oxide, Physics of Metals and Metals Science, Vol 28 (No. 4), 1969, p. 653–660 Y.A. Tamarin, E.B. Kachanov, and S.V. Zherzdev, Thermophysical Properties of Ceramic Layers in TBC, High Temperature Corrosion 4, R. Streiff, J. Stringer, R. Krutenat, M. Caillet, and R.A. Rapp, Ed., Trans Tech Publication, 1997, p 949–955 B.A. Movchan and I.S. Malashenko, Twoand Three-Layer Vacuum-Deposited Coatings for Gas Turbine Blades Protection, Automat. Weld., Vol 2, 1994, p 30–38 U. Schulz, K. Fritscher, H.-J. RatzerScheibe, et al., Thermocyclic Behavior of Microstructurally Modified EB-PVD Thermal Barrier Coatings, High Temperature Corrosion 4, R. Streiff, J. Stringer, R. Krutenat, M. Caillet, and R. Rapp, Ed., Trans Tech Publication, 1997, p 957–964 W.J. Parker, R.J. Jenkins, C.P. Bulter, and G.L. Abbott, Flash Method of Determining Thermal Diffusivity, Heat Capacity and Thermal Conductivity, J. Appl. Phys., Vol 32 (No. 9), 1961, p 1679–1684 R.D. Cowan, Pulse Method of Measuring Thermal Diffusivity at High Temperature, J. Appl. Phys., Vol 34 (No. 4), 1963, p 926– 927 R.F. Blumer and R.E. Taylor, Measurement by the Flash Method of Thermal Diffusivity in Two-Layer Composite Sample, High Temp.—High Press., Vol 6, 1974, p 491– 497 L. Pawlowski, D. Lombard, A. Mahlia, et al., Thermal Diffusivity of Arc Plasma Sprayed Zirconia Coatings, High Temp.— High Press., Vol 16, 1984, p 347–359 P. Morell and R. Taylor, Thermal Diffusivity of Thermal Barrier Coatings of ZrO2 Stabilized with Y2O3, High Temp.—High Press., Vol 17, 1985, p 79–88 AGARDS-R-823, NATO workshop on Thermal Barrier Coatings, (Aalsborg, Denmark), 1998 A. Maesono, “Measurement of Thermal Constants by Laser Flash Method,” SinkuRiko Co. LTD, April, 1983 J.A. Cape and G.W. Lehman, Temperature and Finite-Pulse Time Effect in the Flash


33.

34.

35.

36.

37. 38. 39.

40.

41.

42.

43.

Method for Measuring Thermal Diffusivity, J. Appl. Phys., Vol 34 (No. 7), 1963, p 1909–1913 R.C. Hekman, Finite-Pulse Time and Heat Loss Effects in Pulse Thermal Diffusivity Measurements, J. Appl. Phys., Vol 44 (No. 4), 1973, p 1455–1460 L.M. Clark and R.E. Taylor, Radiation Loss in the Flash Method for Thermal Diffusivity, J. Appl. Phys., Vol 46 (No. 2), 1975, p 714–719 R.F. Blumer and R.E. Taylor, Measurement by the Flash Method of Thermal Diffusivity in Two-Layer Composite Samples, High Temp.—High Press., Vol 6, 1974, p 491– 497 K.B. Larson., K. Koyama, Correction for Finite-Pulse Time Effects in Very Thin Samples Using the Flash Method of Measuring Thermal Diffusivity, J. Appl. Phys., Vol 38 (No. 3), 1967, p 465–473 R.E. Krzhinhanovsky and Z.J. Shtern, NonMetallic Material Thermophysical Properties, Leningrad, Energy, 1973, 336 p R. Berman, Thermal Conduction in Solids, Claredon Press, Oxford, 1976, 283 p J.R. Nicholls, K.J. Lawson, A. Johnston, and D.S. Rickerby, Low Thermal Conductivity EB-PVD Thermal Barrier Coatings, High Temperature Corrosion 5, R. Streiff, I.J. Wright, R. Krutenat, M. Caillet, and A. Cailerie, Ed., Trans Tech Publication, 2001, p 595–606 J.R. Nicholls, K.J. Lawson, D.S. Rickerby, and P. Morrel, “Advanced Processing of TBC’s for Reduced Thermal Conductivity,” Paper 6, AGARDS-R-823, NATO workshop on Thermal Barrier Coatings (Aalsborg, Denmark), 1998 K.J. Lawson, J.R. Nicholls, and D.S. Rickerby, “The Effect of Coating Thickness on the Thermal Conductivity of CVD and PVD Coatings,” Fourth International Conf. on Advances in Surface Engineering (Newcastle, U.K.), 1996 S. Suhr, T.E. Fitchell, and R.J. Keller, Microstructure and Durability of Zirconia Thermal Barrier Coatings, Advances in Ceramics, Vol 12, American Ceramics Society, Inc., 1984, p 503–517 L. Lelait, S. Alperine, and R. Mevrel, Alumina Scale Growth at Zirconia-MCrAlY Interface: A Microstructural Study, J. Mater. Sci., Vol 27 (No. 1), 1992, p 5–12

44. O. Kubaschewski and B.E. Hopkins, Oxidation of Metals and Alloys, Butterworths, London, 1962, 424 p 45. R.A. Miller, Oxidation-Based Model of Thermal-Barrier Coating Life, J. Am. Ceram. Soc., Vol 67 (No. 8), 1984, p 517–521 46. Y.A. Tamarin, V.G. Sundyrin, and V.Y. Kanaev, Stress Calculation of the Ceramic Thermal Barrier Coatings for the Cooled Turbine Blades, Mechanics of Creep Brittle Materials-2, A.C.F. Cocks and A.R.S. Ponter, Ed., Elsevier, 1991, p 308–312 47. D. Zhu and R.A. Miller, “Sintering and Creep Behavior of Plasma-Sprayed Zirconia and Hafnia-Based Thermal Barrier Coatings,” NASA/TM-1998–208406/ REV1, Aug 1998 48. J.R. Nicholls, Y. Jasler, and D.S. Rickerby, Erosion and Foreign Object Damage of Thermal Barrier Coatings, High Temperature Corrosion 4, R. Streiff, J. Stringer, R. Krutenat, M. Caillet, and R.A. Rapp, Ed., Trans Tech Publication, 1997, p 935–948 49. S. Alperine, “Molten Salt Induced High Temperature Degradation of Thermal Barrier Coatings,” Preprint AGARD, High Temperature Surface Interaction, Conference Proceedings N461 50. I.S. Malashenko, G.C. Marijnissen, V.A. Belotserkovsky, A.N. Shelkovoj, and V.V. Orlenko, Life of Condensed Coatings NiCoCrAlY/ZrO2-8%Y2O3 at Thermocyclic Loading, J. Adv. Spec. Electrometall., Vol 1, 1997, p 34–47 51. B.A. Movchan, I.S. Malashenko, K.Y. Yakovchuk, A.I. Rybnikov, and A.A. Tchizik, Two- and Three-Layer Coatings Deposited Under Vacuum for Gas Turbine Blades Protection, J. Surf. Coatings Technol., Vol 67, 1994, p 55–63 52. Y. Jaslier and S. Alperine, EB-TBC Coatings: A Comparative Evaluation of Competing Deposition Technologies, Preprint AGARD, 85th Meeting of the Structures and Materials Panel, 15–16 Oct 1997, p 81 to 8-10 53. Y.A. Tamarin, E.B. Kachanov, and S.V. Zherzdev, Properties of Ceramic Coatings for Turbine Blades Deposited by ElectronBeam Technique, Probl. Mach. Build. Reliab., Vol 1, 1994, p 74–80 54. P.R. Chalker, S.J. Bull, and D.S. Rickerby, A Review of the Methods for the Evaluation


of Coating-Substrate Adhesion, Mater. Sci. Eng. A., Vol 140, 1991, p 583–592 55. A. Bennett, Rolls-Royce Experience with Thermal Barrier Coating, Proc. Br. Ceram. Soc., Vol 34, 1984, p 207–217 56. F.C. Toriz and A. Bennett, A Philosophy for TBCs and Its Collaboration by 10,000 h Service Experience on RB-211 Nozzle Guide Vane, J. Surf. Coatings Technol., Vol 32, 1987, p 359–375 57. T.N. Rhys-Jones and F.C. Toriz, Thermal Barrier Coating for Turbine Application in

Aero Engines, High Temp. Technol., Vol 7 (No. 2), 1989, p 73 58. F.C. Toriz, A.B. Thakker, and S.K. Gupta, Flight Service Evaluation of Thermal Barrier Coating by Physical Vapor Deposition at 5200 h, J. Surf. Coatings Technol., Vol 39/40, 1989, p 161–172 59. I.S. Malashenko, V.V. Orlenko, S.V. Domoroslov, et al., Overlay Coatings on Turbine Blades for Gascompressor Units, J. Adv. Spec. Electrometall., Vol 2, 1990, p 57–70



CHAPTER 9

Some Principles of Strength Designing for Turbine-Blade Protective Coatings* WHEN USING different types of protective coatings for turbine blades, specific problems emerge that are related to the need for strength and lifetime ensurance of both a blade and a coating itself under the action of thermal and mechanical loads brought about by turbine service conditions. Basic principles of the design approach to aircraft-engine turbine blades are discussed elsewhere (Ref 1–4). In calculating the coated-blade service life, the following considerations should be taken into account: ●

The application of a coating has an effect on the main mechanical properties of the blade material. Coatings that have minimum effect on the substrate properties are considered to be preferable. ● A coating gives rise to additional centrifugal load for a blade. This can result in partial loss of the benefits of coating application. This problem is most critical for thermal barrier coatings (TBC). ● Because of distinctions between the thermal expansion coefficients as well as the mechanical and thermal-physical properties of the coating and substrate materials, additional thermal stresses can arise in the coatings, initiating cracks in the coating or resulting in spalling of its fragments from the blade surfaces. As a result, with further running, there is a risk of coating crack propagation into the blade material, causing its failure on the whole. ● In the elevated working temperature range (1100 oC), the reduction of the blade working cross-sectional area can take place in its

hot zone, because of diffusion interaction between the coating and the superalloy. When designing a turbine blade, the effect of the coating must be taken into account. Blademass gain resulting from coating application is allowed for by taking into account its contribution to the centrifugal load. Then, stresses and strains in the blade can be calculated using different model approximations, which are currently used for calculations of turbine blade stressed-strained state, neglecting the contribution of coating stiffness. In the calculations, due regard must be given to the effect of the coating on the mechanical properties of the blade material, involving the test results on the coated specimens. When applying a TBC, there is a need to take into account its influence on the blade temperature field.

Calculation of Stresses and Strains in Coatings For all heat resistant coatings (such as diffusion coatings and overlay coatings) whose thickness is significantly less (by a factor of 10 to 20) than the thickness of cooled turbine-blade walls, one can assume that the strains in the coating are defined completely by the strains in the blade; that is: eZc ⳱ eZ; eSc ⳱ eS

(Eq 1)

where eZc and eSc are strains in the coating, and eZ and eS are strains in the blade surface layer

*This chapter was written in cooperation with Dr. V.G. Sundyrin.




adjacent to the coating corresponding to longitudinal direction z and transverse direction along the surface line S in the blade cross section (Fig. 9.1). The strains eZ and eS in the blade surface layer are determined in advance as a result of the calculation of its stress-strain state with the use of one-, two-, or three-dimensional models, which are usually applied to blades. At this stage, the coating is taken into account only as an additional mass that increases the centrifugal load on a blade. With the use of a one-dimensional model (beam model), it is taken that eS 0. The strain characteristics of (eZc, eSc, eZ, and eS in equalities (Eq 1) are, correspondingly, the sums of strains generated by effective stresses and thermal strains in the coating and the substrate: eZc ⳱ efZc Ⳮ eTc

eSc ⳱ efSc Ⳮ eTc

eTc ⳱ ␣Tc • (T ⳮ T0) eT ⳱ ␣T • (T ⳮ T0)

where T is the blade surface temperature, T0 is the conventional initial temperature, and ␣Tc and ␣T are, correspondingly, thermal expansion coefficients of the coating and substrate materials in the temperature range of T to T0. The temperature characterized by zero thermal stress level in the coating and the blade is used as the initial temperature of T0. As a rule, it is a temperature of coating deposition or that of hightemperature annealing of coated blades. In addition to the previously mentioned strains, the strain enc also occurs in the coating in the direction on the normal, n, to the blade surface (Fig. 9.1), which also must be considered in the calculations. The coating strain enc can be defined from the expression:

(Eq 2a) enc ⳱ ⳮ

and eZ ⳱ efZ Ⳮ eT

eS ⳱ efS Ⳮ eT

(Eq 3)

(Eq 2b)

where efZc, efSc, efZ, and efS are strains being directly caused by the stresses involved, and eTc and eT are, correspondingly, the coating and substrate thermal strains. Thermal strains (eTc , eT) are calculated from the corresponding thermal expansion coefficients of the coating and superalloy materials:

l*c 1 Ⳮ l*c • (eZ Ⳮ eS) Ⳮ 1 ⳮ l*c 1 ⳮ l*c

• ␣TC • (T ⳮ T0)

(Eq 4)

In Eq 4, l*c is Poisson’s ratio for the coating material, in case of its elastic-plastic straining. The plastic strains in the coating can be taken into account as a result of a number of elastic strain calculations using the method of successive approximations discussed elsewhere (Ref 2). With the use of ordinary strain plasticity theory, l*c in Eq 4 is of the following form: l* c ⳱

vc • (1 Ⳮ lc) ⳮ (1 ⳮ 2lc) 2vc • (1 Ⳮ lc) ⳮ (1 ⳮ 2lc)

(Eq 5)

where lc is Poisson’s ratio for the coating in case of its elastic straining, and vc is the coefficient of plasticity: vc ⳱

3Ec 1 • 2(1 Ⳮ lc) EcS

(Eq 6)

In Eq 6, Ec is the elasticity modulus of the coating: ESc ⳱

Fig. 9.1

Stresses in the coating and blade wall

ric eic

(Eq 7)

the so-called “secant” modulus for the coating material, where ric and eic are the stress and strain intensities in the coating, respectively. When considering Eq 1 in Chapter 8, the strain intensity in the coating is defined by the expression:



Some Principles of Strength Designing for Turbine-Blade Protective Coatings / 197

eic ⳱

冪2

3 • 冪(ezc ⳮ esc)2 Ⳮ (ezc ⳮ enc)2 Ⳮ (esc ⳮ enc)2 (Eq 8)

In this expression, only the coating strain enc is the unknown term. For determining enc and, consequently, eic, it is best to use the iterative method, which is often called the method of variable elasticity parameters presented elsewhere (Ref 2). For cyclic loading, the strain-intensity range in the coating can be estimated from the following formula: Deic ⳱ |sign(emax 1c ) max min min • eic ⳮ sign(e1c ) • eic |

(Eq 9)

min In this formula, emax and eic are, correspondic ingly, maximum and minimum values of the strain intensity in the coating during cycle of min loading; sign (emax 1c ) and sign (e1c ) are signs of the first principal strain e1c for, correspondingly, min maximum (emax ic ) and minimum (eic ) values of strain intensity in the coating. Stresses in the coating in the case of its elastic-plastic straining can be estimated, if necessary, with the use of the formulae:

冤1 ⳮ l* E* c

rZc g

2

• (eZ ⳮ l*e c S)

c

Ⳮ rSc

E* c • ␣Tc • (T ⳮ T0) 1 ⳮ l*c

冥

冤

冥

␾

(Eq 10)

3Ec 2(1 Ⳮ lc) • vc Ⳮ 1 ⳮ 2lc

g⳱

1 Ec • h 1Ⳮ E•H

(Eq 11)

(Eq 13)

where N0 is the centrifugal load in the given blade cross section, neglecting the coating mass, and b is the coefficient of the coating-mass effect on the centrifugal load in the blade:

b

where E*c is the reduced modulus of elasticity in the strain theory of plasticity, which is also calculated in the process of successive approximations described elsewhere (Ref 2).

(Eq 12)

As a rule, protective coatings ␾ 2 ⳰ 5% are quite typical. For plastic straining of the coating, the value of this error is significantly less. Value Correction for Centrifugal Load in a Blade. Coating application results in an increase in the value of centrifugal load in a blade. The value of the total centrifugal load in the blade with the consideration of coating mass in a certain specified blade cross section can be estimated with the use of the approximate formula:

Here E* c ⳱

Ec • h • 100% E•H

N ⳱ bN0

E* c g • (eS ⳮ l*e c Z) 1 ⳮ l*c 2 E* • ␣Tc • (T ⳮ T0) Ⳮ c 1 ⳮ l*c

is the coefficient that allows one to take into account tensile stiffness of coating Ec • h, where h and H are, correspondingly, the coating and blade-wall thickness values. In the case of elastic straining of the coating, it should be taken that E*c ⳱ Ec and l*c ⳱ lc for Eq 9 and 10. In this case, the calculation is performed once without recourse to the process of successive approximations. With the use of the previously described model, the stresses and elastic-plastic strains in a protective coating can be calculated for any blade surface points under transient and steadystate conditions. It allows the determination of the elastic-plastic strain range in the coating under cyclic thermomechanical blade loading. The error of the calculations performed for the given model using elastic strains is:

冤

qc • h q•H Ec • h 1Ⳮ E•H 1Ⳮ

(*)

冥

(Eq 14)

where qc and q are, correspondingly, densities of coating and blade materials. The parameter values from the midheight blade portions are used for calculations in this formula. Equations 13 and 14 are obtained from the consideration of the coated blade wall as the sys-




tem of two bars having different tensile stiffness values, E • H and Ec • h, in the case of their joint elastic straining, when the strains of these bars are the same (Fig. 9.2). If b 1, then according to Eq 13, the stresses generated in the coated blade by centrifugal load are less than those in the uncoated blade. This is the case for “hard” coatings. If b 0, then the blade material is being additionally loaded by extra centrifugal load associated with the coating mass. This is the case for “soft” heat resistant metallic coatings at their plastic deformation. If b ⳱ 1, then the total extra centrifugal load generated by the coating mass is carried by the coating itself. This is the case for “neutral” coatings, typical of heat resistant metallic coatings at their elastic deformation and also, to a certain extent, of the ceramic layers of TBCs if the network of developed cracks has not appeared in these layers. Actually, for the condition of b ⱕ 1 to be met, wherein there is no undesirable extra blade loading, it is required, as follows from Eq 14, to fulfill the following condition: qc Ec ⱕ q E

(Eq 15)

The condition (Eq 15) is not met for the case of developed plastic deformation of a corrosion-resistant coating when Ec ⳱ Esc E (qc q), as well as for a TBC with a crack network in the ceramic layer when one can take Ec ⳱ 0. In such a case: b1Ⳮ

qc • h 1 q•H

It means that the coating contributes to bladematerial loading with the extra load.

Fig. 9.2

Centrifugal force in a coated blade

Presented dependencies (Eq 13 and 14) may be used when correcting the value of centrifugal load in the blade, in view of actual coating elastic-plastic properties and thickness in calculating stresses and strains in the blade. Coating Life Estimation. One of the most common types of failures of turbine-blade protective coatings is the initiation of individual cracks or a network of multiple, closely spaced cracks (fragmentation) in the coating. As a rule, the mechanical coating failures occur after longterm operation under thermal cyclic loading. However, in some cases, cracking in the coating may be detected even after short-term cycling, such as 10 to 100 cycles of “start-maximumshut” conditions. It happens by choosing an unsuitable coating because of its physical-mechanical properties. The following criteria may be used for estimation of coating serviceability under the conditions of thermal cycling (Ref 5). Criterion of Ultimate Strain. High-temperature, corrosion-resistant aluminide coatings are characterized by a steep change in ductility within the blade working temperature range of 20 to 1100 oC. The ductile-brittle transition temperature (DBTT) range, in which a steep change in ductility and strength properties is taking place, depends on the coating structure and its chemical composition. If, within the reducedcoating ductility temperature range, considerable thermal stresses occur in the coating due to the difference between the thermal expansion coefficients of the coating and substrate materials, coating cracking may result after a small number of loading cycles (10 to 100 cycles). In this case, the failure mechanism is of a static (quasi-static) nature. Here, as the criterion for coating serviceability, the following condition may be used: eic dc

(Eq 16)

where dc is coating ductility and eic is the strain intensity in the coating. Equation 16 should be checked under transient cooling conditions at the blade surface temperatures, which are less than the DBTT of the coating material. In this case, the coating ductility is minimum, and thermal stresses in the coating caused by the difference between the thermal expansion coefficients of the coating and blade materials may be at their maximum levels. In this case, the residual temperature stresses in condensated coatings usually are found to be tensile, and in diffused coatings they are compressive. Under the service condi-




tions in the blade temperature range of 1000 to 1100 oC, the temperature stresses in the heat resistant coatings are actually at the zero level. As an example, Fig. 9.3 and 9.4 illustrate the plots of strain intensity eic (curves 1) and ductility dc (curves 2) versus temperature for two overlay coatings of Ni20Cr5AlY (Fig. 9.3) and Ni20Cr12AlY (Fig. 9.4) applied to the blade cast from the nickel-base directionally solidified superalloy of MAR-M-200 type. The final annealing temperature, once the coatings are applied, is T0 ⳱ 1050 oC. It follows from the presented plot that the Ni20Cr12AlY coating is more preferable by the criterion of ultimate strain (Eq 16) when compared with the Ni20Cr5AlY coating, despite the

fact that the Ni20Cr5AlY coating is more ductile. Actually, curves 1 and 2 in Fig. 9.3 are almost the same at T ⱕ 180 oC. It evidences that condition (Eq 16) for the Ni20Cr5AlY coating is not met; that is, there is a certain probability for crack nucleation in the coating after a small number of loading cycles. On the contrary, for the Ni20Cr12AlY coating, curves 1 and 2 in Fig. 9.4 are located far from each other, and they do not intersect over the entire temperature range; that is, condition (Eq 16) is met and a considerable margin is left. The main reason, because of which the Ni20Cr5AlY coating is found to be susceptible to cracking despite its rather high ductility, is an unfavorable combination of the coating and blade material thermal expansion coefficients, resulting in enhanced thermal stresses and strains in the coating. Thus, despite its high ductility, the Ni20Cr5AlY coating can provoke cracking on the blade surface rather than hinder it. Criterion of Thermomechanical Fatigue. With large cyclic running time, a network of microcracks (fragmentation) can occur in the turbine blade coatings, which are caused by thermomechanical fatigue (TMF) by nature. The criterion for the coating cyclic life, when applied to such a failure, can be presented as follows (Ref 6): N NRC

(1) Strain intensity, eic, and (2) ductility, dc, of Ni20Cr5AlY coating on a blade uniformly heated to temperature, T

Fig. 9.3

(Eq 17)

where N is the number of loading cycles and NRC is the number of cycles to crack detection under the given cyclic loading conditions. For experimental determination of the coating TMF characteristic (NRC), the approach can be used that is based on the determination of the number of loading cycles for the coated specimens to the first crack detection in the coating at their testing for thermomechanical fatigue (see Chapter 7, “The Effect of Protective Coatings on the Mechanical Properties of Superalloys”). As a result, an experimental TMF curve for a coated specimen can be obtained: De ⳱ De(NRc)

(1) Strain intensity, eic, and (2) ductility, dc, of Ni20Cr12AlY coating on a blade uniformly heated to temperature, T

Fig. 9.4

where De is the strain range of a specimen at cycling loading. In the case of thermomechanical cyclic loading of the coated specimen, the strain-intensity range in the coating, where a plane stress state is observed, can be approximately presented as




a sum of the mechanical component, DeM ic , caused by the external mechanical load applied to the specimen, and the thermal component, DeTic, caused by the difference between the coating and blade material thermal expansion coefficients under cyclic heating and cooling of the specimen: T Deic DeM ic Ⳮ Deic

(Eq 18)

Deic ⳱ 2D␣T • DT

where D␣T ⳱ ␣T ⳮ ␣Tc and DT ⳱ Tmax ⳮ Tmin.

As a result, the total range of elastic-plastic strains in the coating under its nonisothermal loading can be expressed as:

If it is taken that l ⳱ lc ⳱ l*c ⳱ 1/2 (where l is Poisson’s ratio for specimen material) for the developed plastic specimen and coating straining conditions, then, from Eq 1 to 9, it is obtained:

For the specimen and coating elastic straining (if l ⳱ lc 0.3 is used):

DeM ic ⳱ De

Deic 0.87 • De Ⳮ 1.24 • D␣T • DT

(Eq 19)

DeTic ⳱ 2|D␣T(Tmax)(Tmax ⳮ T0) ⳮ D␣T(Tmin)(Tmin ⳮ T0)|

where Tmin and Tmax are the minimum and maximum cycle temperatures: D␣T(Tmax) ⳱ ␣T(Tmax) ⳮ ␣Tc(Tmax) D␣T(Tmin) ⳱ ␣T(Tmin) ⳮ ␣Tc(Tmin)

If one takes, approximately, for the Tmin ⳮ Tmax temperature range that: D␣T(Tmin) D␣T(Tmax) ⳱ ␣T ⳮ ␣Tc const

then Eq 18 is of more simple form:

Thermomechanical fatigue (LA-TMF test) for JS6U superalloy (T ⳱ 100 } 950 oC, t ⳱ 1.5 min). 1, uncoated; 2, Ni10Cr17AlY coating (80 lm); 3, Ni10Cr17AlY coating to detection of the first crack 0.5 mm long; 4, calculated relationship Deic ⳱ Deic(NRC)

Fig. 9.5

Deic De Ⳮ 2D␣T • DT

Thus, if the experimental TMF curve of De ⳱ De(NRC) for the coated specimen is available, then, with the use of dependencies (Eq 18 and 19), the TMF curve of Deic ⳱ Deic(NRC) may be plotted for the coating strain-intensity ranges, which might be directly applied in combination with the coating stressed-strained model discussed earlier to predict its life. Figure 9.5 presents the results of strain control thermomechanical fatigue tests (when De ⳱ const) of standard cylindrical specimens 6.5 mm in diameter with the Ni10Cr17AlY coating. The specimens are cast from the JS6U superalloy and have equiaxial structures. Coating application reduces the lives of the specimens at standard tests in air by a factor of 1.5 to 2, depending on strain-range value, De. The figure also illustrates the relationship for the Ni10Cr17AlY coating that is plotted on the basis of measurements of the number of cycles to detection of the first crack 0.5 mm long in the coating. The period to detection of such a crack appearing in the coating is 2 to 3 times shorter than the time to coated specimen fracture. The calculated plot of the Ni10Cr17AlY coating service life against its strain-intensity range, Deic ⳱ Deic(NRC), is obtained with the use of Eq 18 and 19, based on the experimental relationship of De ⳱ De(NRC). It can be used to predict the thermomechanical life of the given coating applied to a blade, if, when defining a stressed-strained state of the coating, the calculation method discussed previously is used. For approximate estimations of protective coating cyclic lives, when there is a lack of experimental data for coatings, the empirical dependencies used for corrosion-resistant alloys may be applied. In particular, with this purpose,




Manson’s formula (Ref 7) may be used as applied to a coating as follows: De ⳱

3.5 • rBc ⳮ0.12 • NRc Ec

冤冢

Ⳮ 1n

0.6

冣冥

1 1 ⳮ wc

• Nⳮ0.6 Rc .

(Eq 20)

where rBc ⳱ rBc(T) is coating material ultimate strength: wc ⳱

dc 1 Ⳮ dc

is the parameter of coating ductility (transverse cross-sectional area reduction when ruptured).

Stress intensity (ric) in coatings at minimum (1, 3, 5) and maximum (2, 4, 6) gas temperatures. 1 and 2, Ni20Co20Cr12AlY coating; 3 and 4, Ni20Cr12AlY coating; 5 and 6, Ni8Co20Cr12AlY coating

Use of Manson’s formula enables one to estimate only approximately the number of cycles to coating cracking. At the same time, this formula is found to be rather convenient for comparison of the calculated assessments of coating cyclic durability at the preliminary stage of selection of the coatings, which would be best matched to the given blade by the TMF criterion (Eq 17). As an example, Fig. 9.6 to 9.8 show the results of calculated estimates of the stressedstrained state of different condensated coatings deposited on turbine blades, and Fig. 9.9 shows the assessed values of their cyclic lives performed from Manson’s formula (Ref 7). The blade was cast of the nickel-base directionally

Fig. 9.6

Strain intensity (eic) in coatings at minimum (1, 3, 5) and maximum (2, 4, 6) gas temperatures. 1 and 2, Ni20Co20Cr12AlY coating; 3 and 4, Ni20Cr12AlY coating; 5 and 6, Ni8Co20Cr12AlY coating

Fig. 9.7

Strain-intensity range (Deic) in coating at thermocyclic loading of blade. 1, Ni20Co20Cr12AlY coating; 2, Ni20Cr12AlY coating; 3, Ni8Co20Cr12AlY coating

Fig. 9.8

The calculated number of cycles to coating failure (NRc) at thermocyclic loading of blade. 1, Ni20Co20Cr12AlY coating; 2, Ni20Cr12AlY coating; 3, Ni8Co20Cr12AlY coating

Fig. 9.9




solidified superalloy of MAR-M-200. The calculation is performed for mean blade cross-sectional area for two stationary regimes at minimum and maximum gas temperatures. The blade surface temperature under these conditions is shown in Fig. 9.10. The coating strain-intensity range was defined approximately as the difference of the coating strain intensities at maximum and minimum gas temperatures from Eq 9. Stresses and strains in the blade itself were calculated based on the generalized plane-strainstate model and the finite-element method (FEM), with due regard to anisotropy of the mechanical properties of the directionally solidified blade material. The minimum calculated cyclic lives of the coatings correspond to points 27 to 50 on the outer blade surface. For the blade under con-

Fig. 9.10

Temperature of blade surface at minimum (Tmin) and maximum (Tmax) gas temperatures

Table 9.1 Intensities of strains (eic) and stresses (ric) on trailing edge of turbine blade made of MAR-M-002 alloy with Ni20Cr12AlY coating within start-maximum-shut cycle Temperature, oC 13 216 322 401 726 915 1009 893 785 630 392 248 127 13

eic • 103, %

ric, MPa

0.979 1.048 1.054 1.098 1.106 0.341 0.386 0.638 0.840 0.960 0.977 1.013 1.031 0.979

163.5 165.9 162.1 164.9 150.5 43.0 46.1 81.3 112.3 134.5 147.1 159.1 167.1 163.5

sideration, the best coatings, from the cycliclife standpoint, are Ni20Cr12AlY and Ni8Co20Cr12AlY, ones whose minimum calculated cyclic-life values are similar: NRcmin ⳱ 3100 cycles for the Ni20Cr12AlY coating and NRcmin ⳱ 3800 cycles for the Ni8Co20Cr12AlY coating. The test results for blades with the Ni20Co20Cr12AlY and Ni20Cr12AlY coatings run in the engine have confirmed the calculationbased conclusions. For the blade under consideration, preference should be given to the Ni20Cr12AlY coating over the Ni20Co20Cr12AlY coating, by the coating TMF criterion considerations. After the first 80 h of engine tests, many cracks (fragmentation) have appeared on the Ni20Co20Cr12AlYcoated blades, while no cracks have been detected on the Ni20Cr12AlY-coated blades within 380 running hours. As follows from Fig. 9.9, the coating servicelife characteristics in the different points of the blade surface differ greatly because of different stresses, strains, and temperatures in these points. That is why coating service lives may be calculated only for the most loaded points on the blade surface. The calculation results of stress and strain intensities arising in the Ni20Cr12AlY coating on the blade of MAR-M002 alloy in the most loaded point on the trailing edge are presented as an example in Table 9.1 (Ref 8). The calculations are based on the data of temperature and elastic strains arising at different moments of the start-maximum-shut cycle of turbine operation. For this blade point, the calculations of stressed-strained state have been carried out, together with estimation of service lives for different coating variants. For calculations, the data on coating-material properties discussed in previous chapters were used. The thickness of all the coating variants was 100 lm. The calculated number of cycles to coating failure, NRc, was determined from Manson’s formula, Eq 20. Strain-intensity range in the coating was determined from Eq 9. For calculations of NRc values, the coating mechanical properties corresponding to the maximum blade surface temperature within the loading cycle were used. The calculation results are presented in Table 9.2. As follows from Table 9.2, variations in the coating chemical composition open up possibilities of considerable changes in its service life under thermomechanical loading conditions.




The minimum calculated cyclic life of NRc ⳱ 1800 cycles is typical of the Ni20Co20Cr12AlY coating, while the maximum cyclic life of NRc ⳱ 141,000 cycles is demonstrated by the Ni7Cr12AlY coating. Low calculated value of NRc for the Ni20Co20Cr12AlY coating, despite its high ductility, is due to considerable difference between thermal expansion coefficients of this coating and the MAR-M-002 alloy used as a blade material. Among the large number of coatings studied, it is the Ni20Co20Cr12AlY coating whose thermal expansion coefficient differs to the greatest extent from the thermal expansion coefficient of the blade material in the temperature range of 20 to 800 oC, which is characterized by the highest stresses and strains in the coating. Long cyclic life (NRc) of the Ni7Cr12AlY coating is due to quite similar values of the coating and blade material thermal expansion coefficients, especially in the temperature range of low coating ductility (20 to 600 oC), as well as to high yield and ultimate strength values of the coating over its full-service temperature range. As a result, the Ni7Cr12AlY coating deposited on the blade is capable of withstanding a considerable number of cycles without cracking, despite its relatively low ductility. Under service conditions, only elastic strains arise in this coating. It is noteworthy, however, that the choice of coating should be guided by the required combination of coating properties. First of all, the protective properties should be considered. The Ni7Cr12AlY coating, as has been mentioned in Chapter 6, “Turbine Blade Coating Protective Properties,” has relatively low resistance to hot corrosion. In choosing it for application in constructions exposed to hot corrosion attack, its special alloying with such elements as tantalum and hafnium is required.

Thermal Barrier Coatings The main causes of a TBC failure are the thermal stresses, which arise at the ceramic layer/ Table 9.2 Calculated number of cycles to crack appearance, NRc, in protective coatings on trailing edge Coating Ni20Co20Cr12AlY Ni20Cr12AlY Ni7Cr12AlY Ni8Cr18AlY Ni5Cr25Al

NRc, number of cycles 1800 12,500 141,000 76,500 15,500

bond-coat interface, because of the difference between the thermal expansion coefficients of the ceramics and the superalloy. Both these stresses and oxidation at the ceramic/bond-coat interface result in ceramic layer spallation. To estimate stresses and strains in different (ceramic and bond coat) TBC layers, the calculation model described previously for heat resistant metallic coatings can be used. In this case, the calculation for each TBC layer with the use of the previously mentioned model is performed separately (independently of the other layers). This results in determination of the rZc and rSc stresses and the eZc and eSc strain components averaged over the thickness for each TBC layer. When calculating for the ceramic layer, the temperature averaged over its thickness should be used. It can differ by 30 to 50 oC or more from the blade-metal temperature. The calculation model described previously is applicable to TBCs, if the error (␾) for this model determined by Eq 12 for elastic stresses does not exceed some reasonable limit, for example, 10 to 15%. If the TBC ceramic-layer stiffness is found to be comparable with the blade-wall stiffness (which is possible only for thick coatings, when ceramic-layer thickness is 0.3 to 0.4 mm), then one has to use more complicated calculation models. In particular, the two-dimensional generalized plane-strain-state models based on FEM application are effective for this case. Such models allow natural consideration of the combined effect of the coating and blade-wall stiffness characteristics, especially if specific finite elements are involved, for example, finite elements in the form of curvilinear strips intended for the calculation of stresses and strains in thin-surface blade layers (Ref 9). Thermal Barrier Coating Life Estimation. Criterion of TBC Ceramic-Layer Cracking. Spalling of ceramic-layer fragments from a blade surface is the main type of TBC mechanical failure. The occurrence of a network of microcracks in the ceramic layer (fragmentation) in some blade-surface zones is considered to be the initial stage of TBC failure. At the same time, ceramic cracking does not yet mean its failure. Safe operation for a long time period can be possible for blades having such cracks formed in the TBC ceramic layer usually within the first hours or tens of hours of turbine operation. The condition for TBC cracking can be presented as follows (Ref 9):




rmax ⳱ rcr

(Eq 21)

where rmax is the maximum tensile stress in the center of the fragment of cracked ceramic layer/ metallic bond-coat interface (point B in Fig. 9.11 and 9.12), and rcr ⳱ rcr(T) is the cohesive ultimate strength of the ceramic layer (minimum tension stress bringing about instant ceramic cracking at the temperature of T). The subscript “cr” symbol stands for a TBC ceramic layer. Figure 9.12 illustrates the plot of the maximum relative tensile thermal stress in the center of the ceramic-layer fragment at the moment of its cracking versus the relative size, d/hcr, of these fragments:

rmax ⳱

At the moment when the ceramic layer is being cracked, the stress (rmax) is as high as the cohesive ultimate strength (rcr), that is rcr ⳱ rmax ⳱ f (d/hcr) Ecr • (␣T ⳮ ␣Tcr) • (T ⳮ T0) • 1 ⳮ lcr

(Eq 22)

Using this formula and the plot given in Fig.

rmax • (1 ⳮ l) ⳱ f (d/hcr) Ecr • (␣T ⳮ ␣Tcr) • (T ⳮ T0)

where d is fragment spacing and hcr is thickness of a TBC ceramic layer. This dependence is obtained from calculations using the two-dimensional FEM for elastic thermal stresses in the three-layer plate, whose top (ceramic) layer has a regular network of cracks with a fragment size designated as d. The calculation is performed for a two-layer TBC of ZrO2-8%Y2O3/Ni20Cr12AlY (100/80 lm) deposited by electron beam technique.

Fig. 9.11

Fig. 9.12

Maximal relative stress (r*max) versus relative size (d/hcr) of ceramic layer fragment

Thermal stresses in cracked ceramic layer of TBC. (a) Finite-element grid. (b) Thermal stress isolines (MPa). (c) Thermal stress variations along blade-wall thickness in the center of the cracked ceramic-layer fragment




9.12, one can determine the value of the ceramic-layer cohesive ultimate strength, rcr, for the specified temperature conditions, if the mean size of the cracked fragments of the ceramic layer and the ceramic temperature, T, at its cracking are defined. Obviously, cracking of the ceramic layer is possible if tensile thermal stresses are generated in it. It can take place only when the coating temperature, T, is higher than the annealing temperature, T0, (as for ceramic layers ␣Tcr ␣T), and in cases where considerable creep strains are accumulated in the blade-metal surface layer and in the metallic bond coat, which are capable of giving rise to tensile stresses in the ceramic layer. If the rcr values for different service conditions are known, then, with the use of Eq 22, the plot in Fig. 9.12, and Eq 21, it is possible to

Fig. 9.13

Model of bond-coat surface oxidation

Fig. 9.14

Maximal principal stress (r1) in ZrO2-8%Y2O3 ceramics in the vicinity of Al2O3 ridge (MPa)

estimate by calculations the probability of TBC cracking and even to define the mean size of the ceramic layer fragments that appear on cracking. Criterion of TBC Ceramic-Layer Spallation. Thermal barrier coating failure results from the decrease in adhesive strength at the ceramic/ bond-coat interface during long-term service at high temperatures, causing ceramic-layer spalling (Ref 10). This effect is due to the formation of the oxide of Al2O3 type at the ceramic/bondcoat interface. Aluminum oxide not only reduces adhesive strength but also changes the stressed state at the ceramic/metallic bond-coat interface, because the properties of this oxide are different from those of the ceramic and bond-coat layers. To simulate the damaged ceramic layer, it may be thought that Al2O3 formation does not occur uniformly over the ceramic/bond-coat interface, but by the formation of the specific “ridges” intruding into the ceramic layer (Fig. 9.13). The dimensions of these ridges and their spacing depend on ZrO2 • Y2O3 ceramic-layer crystal sizes, ceramic porosity, blade-surface roughness, and so on. However, the most intensive growth of these ridges is observed at the base of fragmentation cracks in the ceramic layer. Formation of Al2O3 occurs at a high temperature, and while cooling, additional internal thermal stresses are generated in the ceramic layer near the ridges, which are caused by the difference between the values of the thermal expansion coefficients for the TBC ceramic layer and Al2O3 oxide. As an example, Fig. 9.14 illustrates the results of calculations done using a FEM for the twodimensional problem of joint straining of two semiinfinite bodies of ZrO2 and Al2O3 materials uniformly cooled to the same temperature of 20 o C and joining over the contact surface. At the interface of these materials, in the ZrO28%Y2O3 body, there is a ridge of Al2O3 oxide in the form of a wedge 1 lm high and 2 lm wide at its base. The calculation used the following initial data: T0 ⳱ 1000 oC; for ZrO2-8%Y2O3 ⳮ E ⳱ 10 GPa, ␣T ⳱ 9.3 • 10ⳮ6 1/oC; lcr ⳱ 0.3; for Al2O3 ⳮ E ⳱ 40 GPa; ␣ ⳱ 8.0 • 10ⳮ6 1/oC; l ⳱ 0.3. As is seen from the offered model, extra concentration of thermal stresses in the ceramic layer near the Al2O3 oxide ridge facilitates both ceramic-layer spalling from the substrate in the vicinity of point A at the ridge base, and the opening of a crack formed earlier in the ceramics or a new crack nucleating between the adjacent columnar crystallites of ZrO2-8%Y2O3 in the vicinity of point B at the ridge top (Fig. 9.14).




As an integral characteristic describing the failure mechanics of the TBC ceramic layer, the adhesive strength value racr ⳱ racr(T, s) defined experimentally can be used. Here, racr is the minimum stress of ceramic-layer separation from the substrate at room temperature after holding the specimen in a furnace at the temperature of T for a period of time, s. In order to take into account the effect of thermocyclic loading on TBC adhesive strength, the experiments on determination of the racr characteristic

can be performed with simulation of blade-service thermocycles by means of periodic specimen cooling to room temperature followed by heating to the test temperature. The criterion of TBC adhesive strength can be presented as: s sRcr

where sRcr is time of operation at high temperature before spalling of TBC ceramic layer from the blade surface after cooling to room temperature, and s is time of blade operation at high temperature. Time to ceramic-layer spalling, sRcr, can be defined from the experimental relation of ceramic-layer adhesive strength to its test time and temperature. They are shown in Fig. 9.15. The TBC adhesive strength curves can cross the time axis. Ceramic spalling caused by formation of Al2O3 oxide at the ceramic/bond-coat interface can also take place at zero stress level in the ceramic layer at low blade temperatures. The stresses rncr, which facilitate TBC ceramic-layer separation from the blade surface (Fig. 9.16), may be approximately defined by the formula: rncr

Fig. 9.15

Fig. 9.16

(Eq 23)

hcr • rscr R

(Eq 24)

Adhesion strength curve

(a) Stresses of spalling (rncr) and (b) rule of signs for these stresses

where R is a radius of curvature of the outer surface of the blade airfoil in its cross-sectional area. The values of rncr stresses are the greatest in the blade leading-edge zone, where the value of radius of curvature, R, is minimum. Ceramic layers of TBC are capable of withstanding high compressive thermal stresses caused by the difference between the thermal expansion coefficients of the ceramics and the blade material. However, with the occurrence of even minor tensile thermal stresses in the ceramic layer, this layer experiences immediate cracking, resulting in tensile stress relaxation. Thus, if the calculation performed gives the tensile stresses rZcr or rScr, in the ceramic layer, these stresses should be taken as zero. In the temperature range below the temperature of ceramic-layer deposition on the blade surface, only compressive stresses, rScr, in convex surface layers of the blade (Fig. 9.16), such as its suction face and leading edge, can give rise to spalling of ceramic-layer fragments. In concave blade-surface layers (such as its pressure face), these compressive stresses press the ceramic layer to the blade.




The following rule of signs for the stresses of layer separation rncr determined by Eq 24 can be used. Under the compressive stresses, rncr, in the ceramic layer, the stresses of separation, rScr, in convex blade-surface layers are positive, and those in concave blade-surface layers are negative. If the coating is affected by the positive stress, rncr, (which separates the coating from the blade surface), time to coating failure (spalling of ceramic layer), sRcr, can be found from the adhesive strength curves racr ⳱ racr(T, s) at the specified stress of separation, rncr (point A, Fig. 9.15). If the coating is affected by the negative stress, rncr, (which presses the coating to the blade), then, in this case, one can assume that rncr ⳱ 0 and define the time to ceramic spalling, sRcr, as the point where the appropriate adhesive strength racr(T, s) curve crosses the axis s (point B, Fig. 9.15). If no effect of rncr stress on the ceramic layer is observed, that is, rncr ⳱ 0, then sRcr is also defined by point B in Fig. 9.15. Stresses in the TBC ceramic layer vary with time, mainly due to the accumulation of creep strains in the blade. This process is most intensive in the leading- and trailing-edge zones, where the blade material temperature is maximal. The residual thermal stresses in the ceramic layer also vary with the blade running time. As an example, the results of durability calculations for a TBC ceramic layer on a turbine

Fig. 9.17

blade are presented subsequently. The CMSX-4 single-crystal alloy was used as a blade material. The ceramic layer was deposited by electron beam technique. The TBC consists of two layers: Ni20Cr12AlY bond layer (80 lm) and ZrO2-8%Y2O3 ceramic layer (100 lm). The steady-state temperature field at the blade leading edge is shown in Fig. 9.17 (a) and (b). The stressed-strained state calculation in the blade under stationary thermal conditions is carried out with the use of a generalized plane (twodimensional) strain-state model and the FEM. Triangular finite elements with a cubic approximation of form functions for displacements were used. The calculation is carried out with account of anisotropy of elastic, plastic, and creep properties of the single-crystal alloy. The calculations for the coating bond coat are carried out using the models described in this chapter. The calculation results are shown in Fig. 9.18, 9.19, and 9.20. Maximum thermal stresses in TBC ceramic layer arise on cooling the blade to room temperature (20 oC) immediately after coating deposition and subsequent high-temperature annealing of the coated blade (T0 ⳱ 950 oC). In this case, compressive residual stresses of rres ⳱ rZcr ⳱ rScr ⳱ ⳮ400 MPa arise in the TBC ceramic layer at its interface with the metallic bond coat, and tensile residual stresses of rres ⳱ Ⳮ560 MPa arise in the bond coat. In this case,

(a) Stationary thermal field and (b) temperature distribution over blade leading edge




Fig. 9.18

Actual stresses (rScr) in ceramic layer of TBC on the blade leading edge

Fig. 9.19

Residual stresses (rScr) in ceramic layer of TBC on the blade leading edge

Adhesion strength, racr, (full lines) and stress of spallation, rncr, (dotted line) versus time, s, in blade with TBC after cooling to room temperature

Fig. 9.20

the stresses separating the ceramic layer from the blade surface are maximum and account for rncr⳱ Ⳮ50 MPa. At blade heating, the internal elastic compressive stresses in the TBC ceramic layer decrease; under the steady-state operating conditions, they become tensile in the blade leading-edge zone and reach the maximum values of rScr ⳱ rZcr ⳱ Ⳮ70 ⳰ 75 MPa in points 6 and 7 of the blade surface at the initial time moment of s ⳱ 0 (Fig. 9.18, dotted lines). When taking, for example, the racr ⳱ 15 MPa value as the TBC ceramic layer cohesive strength for the blade leading-edge thermal conditions, then the indicated TBC surface zone should be covered with a network of microcracks. Their spacing is 100 lm in accordance with the plot given in Fig. 9.15. Under these conditions, the actual tensile stresses, rscr and rzcr, in the ceramic layer on the blade leading edge (points 4 to 9, see Fig. 9.17 and 9.18) relax. In this case, the plots for actual stress, rscr, in Fig. 9.18 between points 4 and 9 are full straight lines, rscr ⳱ 0 parallel to the abscissa. At long-term running, creep strains are accumulated in the blade body, resulting in the decrease in residual stresses arising in the TBC ceramic layer at blade cooling. In the example being considered, on reaching running time of s 300 h, these residual stresses in the leadingedge zone in points 6 and 7 are reduced to rScr ⳱ rres ⳮ200 MPa (Fig. 9.19). With decreasing residual stresses, rScr, the stresses of rncr determined by Eq 24, which separate the ceramic layer from the blade, are also being reduced. The plot of ultimate adhesive strength racr ⳱ racr(T, s) variation versus running time, s, under the service conditions in the most hazardous point, 6, of the highest blade-metal temperature of T ⳱ 1120 oC is presented in Fig. 9.20, together with the plot of the stress of separation, rncr, in this point at cooling the blade to room temperature (dotted line). This plot shows that the rate of reducing the stress of separation, rncr, in point 6 versus running time is less than the rate of reducing the adhesive strength, racr, in this point. The curves racr(s) and rncr(s) in Fig. 9.20 intersect at the time moment of s ⳱ 265 h. Heat resistant bond-coat life may be estimated using the approaches discussed in this chapter. It is noteworthy that the calculation results are approximate. For more reliable calculation of estimate of time to TBC spalling, one should have experimental data on modulus of elasticy, Ecr,




for the ceramic layer at its compression. The value has a noticeable effect on the level of calculated stresses of separation, rncr, and on the calculated service life of the TBC ceramic layer in the leading-edge zone of the blade. The experimental studies of the dependence of adhesive strength of the TBC ceramic layer on test time and temperature are also required.

6.

REFERENCES

1. Y.A. Nozhnitsky, R.A. Doulnev, and V.G. Sundyrin, Damage Mechanisms for Thermomechanical Fatigue of Aircraft Engines Material, 81st Meeting of the AGARD SMP (Banff, Canada), 2–4 Oct 1995, and published in CP-559 2. I.A. Birger and B.F. Chorr, Thermal Fatigue of Engine Details, Moscow, Mashinostroene, 1975, 455 p 3. I.A. Birger and N.I. Koterov, Ed., Strength Calculations for Aircraft Gas Turbine Engines, Moscow, Mashinostroene, 1984, 208c 4. V.K. Kostege, V.A. Khalturin, and V.G. Sundyrin, “Mathematical Models of Gas Turbine Engines and Their Components,” AGARD-LS-198, Advisory Group for Aerospace Research and Development, Dec 1994 (France), North Atlantic Treaty Organization 5. Y.A. Tamarin, E.B. Kachanov, and V.G. Sundyrin, Gas Corrosion and Thermal Fatigue of Protective Coatings for Turbine

7. 8.

9.

10.

Blades, High Temperature Corrosion and Protection, H. Guan, W. Wu, J. Shen, and T. Li, Ed., Liaoning Science and Technology Publishing House, China, 1991, p 161– 166 Y.A. Tamarin, V.G. Sundyrin, and N.G. Bychkov, Thermo-Mechanical Fatigue Tests of Coatings for Turbine Blades, High Temperature Surface Engineering, J. Nicholls and D. Rickerby, Ed., IOM Communications Ltd., 2000, p 587–593 S.S. Manson, Thermal Stress and Low-Cycle Fatigue, New York, 1964, p 187–213 Y.A. Tamarin, H.C. Low, Thermal Fatigue of Protective Coatings on CMSX-4 Superalloy, High Temperature Corrosion and Protection of Materials—5, R. Streiff, I.G. Wright, R.C. Krutenat et al., Ed., Trans Tech Publication, Switzerland, 2001, p 647–654 V.G. Sundyrin, Y.A. Tamarin, and V.J. Kanaev, Stress Calculation of the Ceramic Thermal Barrier Coatings for the Cooled Turbine Blades, Mechanics of Creep Brittle Materials 2, A.C.F. Cocks and A.R.S. Poner, Ed., Elsevier App. Science, 1991, p 308–313 Y.A. Tamarin and V.G. Sundyrin, Adhesive Strength and Life of TBC-EB Ceramic Layer, Proc. Intern. Symposium on High Temperature Corrosion Protection (Abstracts Booklet), SEFRACOR, Universite de Provence, EPRI, Les Embier, France, May 1996



Product code


06738G








APPENDIX

Regression Equation FOR SIMULATION of the relationship and plotting models for the properties of the alloys for the coatings depending on their chemical compositions, the methods of regression analysis were used (Ref 1). They included methods of experiment planning and an application package of “Stark” (statistical analysis regression correlation). The models for a Ni-Co-Cr-Al system were plotted using the data on the properties of the alloys, whose chemical compositions corresponded to the two-power plan matrix of B4 (Table A.1). To plot models for Co-Ni-Cr-Al and Ni-Cr-Al systems, the passive statistic data on Table A.1

the properties of the alloys, whose chemical compositions are given in Tables A.2 and A.3, were used. Standardly normalized magnitudes of variables (x1 … xn) are given in Tables A.1, A.2, and A.3. They are correspondingly used in calculated regressions: xi ⳱

xie ⳮ xia Dxi

where xi is normalized magnitude of variable i; xie is actual magnitude of variable;

Ni-Co-Cr-Al alloys Chemical composition, wt%

Normalized number of variable

No.

Alloy

Ni

Co

Cr

Al

Y

x1

x2

x3

x4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Ni10Co15Cr8Al Ni30Co15Cr8Al Ni10Co30Cr8Al Ni30Co30Cr8Al Ni10Co15Cr14Al Ni30Co15Cr14Al Ni10Co30Cr14Al Ni30Co30Cr14Al Ni10Co15Cr8AlY Ni30Co15Cr8AlY Ni10Co30Cr8AlY Ni30Co30Cr8AlY Ni10Co15Cr14AlY Ni30Co15Cr14AlY Ni10Co30Cr14AlY Ni30Co30Cr14AlY Ni10Co22Cr11AlY Ni30Co22Cr11AlY Ni20Co15Cr11AlY Ni20Co30Cr11AlY Ni20Co22Cr8AlY Ni20Co22Cr14AlY Ni20Co22Cr11Al Ni20Co22Cr11AlY Ni20Co22Cr11AlY

Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base

10 30 10 30 10 30 10 30 10 30 10 30 10 30 10 30 10 30 20 20 20 20 20 20 20

15 15 30 30 15 15 30 30 15 15 30 30 15 15 30 30 22.5 22.5 15 30 22.5 22.5 22.5 22.5 22.5

8 8 8 8 14 14 14 14 8 8 8 8 14 14 14 14 11 11 11 11 8 14 11 11 11

0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.3 0.3 0.3 0.3 0.3 0.3 0.1 0.5 0.3

ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 0 0 0 0 0 0 0

ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 0 0 ⳮ1 Ⳮ1 0 0 0 0 0

ⳮ1 ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 ⳮ1 ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 0 0 0 0 ⳮ1 Ⳮ1 0 0 0

ⳮ1 ⳮ1 ⳮ1 ⳮ1 ⳮ1 ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 0 0 0 0 0 0 ⳮ1 Ⳮ1 0

The alloys include 1.5% Hf, 1.0% W, 0.5% Nb, and 0.5% Si.


Table A.2

Co-Ni-Cr-Al alloys Chemical composition, wt%


No.

Alloy

Co

Ni

Cr

Al

Y

x1

x2

x3

x4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Co10Ni15Cr6AlY Co30Ni15Cr6AlY Co10Ni30Cr6AlY Co30Ni30Cr6AlY Co10Ni15Cr12AlY Co30Ni15Cr12AlY Co10Ni30Cr12AlY Co30Ni30Cr12AlY Co10Ni22Cr9AlY Co30Ni22Cr9AlY Co20Ni15Cr9AlY Co20Ni30Cr9AlY Co20Ni22Cr6AlY Co20Ni22Cr12AlY Co20Ni22Cr9AlY Co20Ni22Cr9AlY

Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base Base

10 30 10 30 10 30 10 30 10 30 20 20 20 20 20 20

15 15 30 30 15 15 30 30 22.5 22.5 15 30 22.5 22.5 22.5 22.5

6 6 6 6 12 12 12 12 9 9 9 9 6 12 9 9

0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.6 0.6 0.2 0.2 0.1 0.1 0.5 0.3 0.3

ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 ⳮ1 Ⳮ1 0 0 0 0 0 0

ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 0 0 ⳮ1 Ⳮ1 0 0 0 0

ⳮ1 ⳮ1 ⳮ1 ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 Ⳮ1 0 0 0 0 ⳮ1 Ⳮ1 0 0

0.2 ⳮ0.2 ⳮ0.2 ⳮ0.2 ⳮ0.2 ⳮ0.2 ⳮ0.2 Ⳮ1 Ⳮ1 ⳮ0.6 ⳮ0.6 ⳮ1 ⳮ1 ⳮ0.2 ⳮ0.2 ⳮ1

xia is average magnitude of variable in the range under consideration; and Dxi is range of varying the variable during experiment. Processing of experimental results makes it possible to obtain mathematical models connecting system properties and alloying element contents (regression equations) in the form of square polynomials, which may be presented in general as a series: n

y ⳱ bon Ⳮ

n

兺 bi xi Ⳮ i1j 兺 bij xi xj i⳱1 n

Ⳮ

兺

bii x i2 Ⳮ . . . .

n

S2R ⳱

兺 (yci ⳮ yei)2 i⳱1

where yei is actual value (average experimental magnitude of the property) in i-point; yci is magnitude of the property under study calculated from the model equations (response value) in this point; and n is number of experiments in the plan matrix. In addition to the above, the following criteria were used for selection and statistic analysis of regression equations. Multiple correlation coefficient (Rf) corrected on the basis of degree of freedom:

i⳱1

The procedure of increment inclusion of the terms in the model was used to select the best regression because it allowed maximal reduction of the residual sum of squares:

Table A.3

Ni-Cr-Al alloys Chemical composition, wt%


No.

Alloy

Ni

Cr

Al

Y

x1

x2

1 2 3 4 5 6 7 8 9 10 11 12

Ni22Cr11AlY Ni14Cr3AlY Ni18Cr5AlY Ni16Cr9AlY Ni21Cr9AlY Ni18Cr16AlY Ni19Cr12AlY Ni8Cr18AlY Ni8Cr15AlY Ni7Cr12AlY Ni7Cr10AlY Ni20Cr6AlY

Base Base Base Base Base Base Base Base Base Base Base Base

22.3 14.7 17.7 16.0 20.8 18.3 19.5 8.0 8.0 7.0 7.2 20.0

11.2 2.8 5.5 9.1 8.8 15.7 12.6 18.0 15.0 12.0 10.0 6.0

0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3

Ⳮ0.97 0 Ⳮ0.38 Ⳮ0.16 Ⳮ0.78 Ⳮ0.45 Ⳮ0.61 ⳮ0.87 ⳮ0.87 ⳮ1 ⳮ0.97 Ⳮ0.67

Ⳮ0.07 ⳮ0.97 ⳮ0.71 ⳮ0.18 ⳮ0.18 Ⳮ0.73 Ⳮ0.21 Ⳮ1 Ⳮ0.60 Ⳮ0.21 ⳮ0.05 ⳮ0.57

Rf ⳱

冪

1 ⳮ (1 ⳮ R2) • (n ⳮ 1) n ⳮ k

where R is assessment of multiple correlation coefficient for the effects of the factors included in the model at j-increment; k is number of independent assessments of coefficients in the model of j-increment calculation; and n is number of experiments included in the plan matrix. Accumulated cancelled ratio (ACR) of variances of the given increment of model selection to the sum of variances cancelled at the increments from the first to the current one. The regressions, for which the variance of nonadequacy (S2na) seemed to be minimal, multiple correlation coefficient was close to 1, and the current increment variance cancelled was comparable with an experiment error, were chosen as the best ones.

Appendix / 213

Ni-Co-Cr-Al alloys Variability of values Variable Co Cr Al Y T, oC

Symbol

min

max

X1 X2 X3 X4 X5

10.10 14.40 7.80 0.04 …

30.20 30.00 14.10 0.45 …

Liquidus temperature Regression equation: Tliq ⳱ 1366.6 ⳮ 9.8 X2 Ⳮ 65.6 X3 Ⳮ 42.7 X12 ⳮ 14.6 X1 X2 X3, (oC) 2 S na ⳱ 239.8; (DSj2) ⳱ 0.94; Rf ⳱ 0.97 Solidus temperature Regression equation: Tsol ⳱ 1275.2 Ⳮ 6.6 X3 ⳮ 10.7 X12 Ⳮ 18.4 X32 Ⳮ 21.3 X2 X3 ⳮ 12.8 X1 X2 X3, (oC) 2 S na ⳱ 93.7; (DSj2) ⳱ 0.88; Rf ⳱ 0.94 Crystallization interval Regression equation: DT ⳱ 90.9 ⳮ 10.1X2 Ⳮ 60.7X3 Ⳮ 52.0X12 ⳮ 16.0X32 Ⳮ 3.6X1X3 ⳮ 27.2X2 X3, (oC) 2 S na ⳱ 73.0; (DSj2) ⳱ 0.98; Rf ⳱ 0.99 NiAl (b-phase) quantity

S2na

0.62X12 ⳮ 0.20X22 Ⳮ 0.14X32 Ⳮ 0.11X1X2X3, (⳯10ⳮ6, Kⳮ1) ⳱ 0.01; (DSj2) ⳱ 0.93; Rf ⳱ 0.96

Thermal expansion coefficient, range 100– 500 oC Regression equation: ␣ ⳱ 15.0 Ⳮ 0.14 X1 ⳮ 0.22 X2 ⳮ 0.54 X12 ⳮ 0.20X22 ⳮ 0.08X2X3 Ⳮ 0.09 X1 X2 X3, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.03; (DSj2) ⳱ 0.88; Rf ⳱ 0.93 Thermal expansion coefficient, range 100– 600 oC Regression equation: ␣ ⳱ 15.5 Ⳮ 0.22 X1 ⳮ 0.31 X2 ⳮ 0.47 X12 ⳮ 0.25 X22, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.03; (DSj2) ⳱ 0.89; Rf ⳱ 0.94 Thermal expansion coefficient, range 100– 700 oC Regression equation: ␣ ⳱ 16.3 Ⳮ 0.27 X1 ⳮ 0.40 X2 ⳮ 0.17 X3 ⳮ 0.64 X12 ⳮ 0.31 X22 ⳮ 0.08 X1 X3, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.06; (DSj2) ⳱ 0.89; Rf ⳱ 0.94 Thermal expansion coefficient, range 100– 800 oC

Regression equation: Qb ⳱ 60.3 Ⳮ 3.8X1 Ⳮ 13.0X2 Ⳮ20.8X3 Ⳮ 9.5X12 ⳮ 6.6X32 Ⳮ 2.5X1X3 ⳮ 11.2X2X3, (vol%) S2na ⳱ 14.5; (DSj2) ⳱ 0.97; Rf ⳱ 0.98

Regression equation: ␣ ⳱ 17.4 Ⳮ 0.19 X1 ⳮ 0.50 X2 ⳮ 0.28 X3 ⳮ 0.88 X12 ⳮ 0.43 X22, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.09; (DSj2) ⳱ 0.89; Rf ⳱ 0.94

NiAl (b-phase) quantity, annealing 1200 oC, 200 h

Thermal expansion coefficient, range 100– 900 oC

Regression equation: Qb ⳱ 63.6 Ⳮ 3.9X2 Ⳮ 25.0X3 Ⳮ 7.3X12 ⳮ 8.6X22 ⳮ 7.6X32 ⳮ 3.5X1X3 ⳮ 11.3X2X3, (vol%) S2na ⳱ 19.4; S (DSj2) ⳱ 0.96; Rf ⳱ 0.98

Regression equation: ␣ ⳱ 18.3 ⳮ 0.64 X2 ⳮ 0.48 X3 ⳮ 0.99 X12 ⳮ 0.59 X22 ⳮ 0.17 X1X2 ⳮ ⳮ 0.20 X1 X3 Ⳮ 0.31 X1 X2 X3, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.07; (DSj2) ⳱ 0.95; Rf ⳱ 0.97



Regression equation: ␣ ⳱ 14.01 ⳮ 0.15 X2 ⳮ 0.06 X3 ⳮ 0.72 X12 Ⳮ 0.22 X32 ⳮ 0.08 X1 X3 Ⳮ 0.16 X1 X2 X3, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.01; (DSj2) ⳱ 0.92; Rf ⳱ 0.96

Regression equation: Y ⳱ 19.3 ⳮ 0.31X1 ⳮ 0.80X2 ⳮ 0.65X3 ⳮ 0.96X12 ⳮ 0.66X22 ⳮ 0.19X1X2 ⳮ ⳮ 0.21X1X3 Ⳮ 0.54X1 X2 X3, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.03; (DSj2) ⳱ 0.98; Rf ⳱ 0.99


Elasticity modulus, 20–800 oC

Regression equation: ␣ ⳱ 14.5 Ⳮ 0.10X1 ⳮ 0.17X2 ⳮ 0.07X3 ⳮ

Regression equation: E ⳱ 1515 Ⳮ 118.7X1 Ⳮ 107.2X2 ⳮ 41.7X3 ⳮ 134.6X5 Ⳮ 85.6X22 ⳮ 100.6X32 ⳮ


ⳮ 37.2 X52 Ⳮ 50.6 X1 X2 Ⳮ 69.7 X1 X3 ⳮ 20.6 X1 X5 ⳮ 2.8 X2 X3 Ⳮ 7.4 X2 X5 Ⳮ Ⳮ 52.5X3X5 Ⳮ 25.9X1 X2X3 Ⳮ 14.0X1X2X5 ⳮ 12.0X2X3X5 (⳯100, MPa) S2na ⳱ 220; (DSj2) ⳱ 0.99; Rf ⳱ 0.99 Hot corrosion, 850 oC, 200 h Regression equation: qⳮ ⳱ 200.6 Ⳮ 22.1X2 Ⳮ 223.3X12 ⳮ 75.3X22 ⳮ 81.5X32 ⳮ 77.5X42 Ⳮ 123.2X1X2 Ⳮ Ⳮ 45.4 X1 X3 Ⳮ 130.9 X2 X3 Ⳮ 77.7 X2 X4 Ⳮ 38.1 X3 X4 Ⳮ 66.2 X1 X2 X3 Ⳮ Ⳮ 87.8 X1 X2 X4 Ⳮ 167.5 X1 X2 X3 X4 (g/m2) S2na ⳱ 653; (DSj2) ⳱ 0.98; Rf ⳱ 0.99 o

Hot corrosion, 900 C, 200 h Regression equation: log qⳮ ⳱ 1.868 ⳮ 0.224 X12 ⳮ 0.189 X22 Ⳮ 0.303 X32 ⳮ 0.089 X1 X3, (g/m2) 2 S na ⳱ 0.015; (DSj2) ⳱ 0.70; Rf ⳱ 0.84 Heat resistance, isothermal test at 1200 oC, 100 h Regression equation: qⳭ ⳱ 30.1 ⳮ 6.56X1 ⳮ 2.83X2 Ⳮ 18.74X3 Ⳮ 20.74X4 Ⳮ 4.50X12 Ⳮ 0.07X22 Ⳮ Ⳮ 24.50 X42 ⳮ 9.52X1X2 ⳮ 12.38X1X3 ⳮ 5.91X1X4 ⳮ 35.16X2X3 Ⳮ 3.86X2X4 Ⳮ Ⳮ 14.29X3 X4 Ⳮ 13.32 X1 X2 X3 ⳮ 13.14 X1 X2 X4 ⳮ 22.29 X1 X3 X4 ⳮ ⳮ 25.92 X2 X3 X4 Ⳮ 29.31X1 X2 X3 X4, (g/ m2) S2na ⳱ 28; (DSj2) ⳱ 0.96; Rf ⳱ 0.97 o

Heat resistance, isothermal test at 1200 C, 300 h Regression equation: qⳭ ⳱ 62.6 ⳮ 13.60X1 ⳮ 4.69X2 Ⳮ 37.37X3 Ⳮ 37.80X4 Ⳮ 8.98X12 Ⳮ 28.71X22 ⳮ ⳮ 29.42X32 Ⳮ 34.84X42 ⳮ 11.05X1X2 ⳮ 2.39X1X3 ⳮ 22.28X1X4 ⳮ 59.72X2X3 Ⳮ Ⳮ 6.57X2X4 Ⳮ 10.21X1X2X3 ⳮ 21.88X1X2X4 ⳮ 47.49X1X3X4 ⳮ 25.68X2X3X4 Ⳮ Ⳮ58.37X1X2X3X4, (g/m2) S2na ⳱ 20; (DSj2) ⳱ 0.99; Rf ⳱ 0.99 b-phase-free zone after heat resistance testing at 1200 oC, 200 h Regression equation: Hb ⳱ 69 ⳮ 9.2X1 ⳮ 14.1X2 ⳮ 47.5X3 Ⳮ 65.5X12 ⳮ 28.9X22 ⳮ 46.2X32 Ⳮ 4.8X42 Ⳮ Ⳮ 7.1X1 X2 ⳮ 16.3 X1 X4 ⳮ 7.2 X2 X3 Ⳮ 11.7 X1 X2 X3 ⳮ 13.4 X2 X3 X4 Ⳮ

Ⳮ 28.3X1X2X3X4, (lm) S2na ⳱ 59; (DSj2) ⳱ 0.99; Rf ⳱ 0.99 Co-Ni-Cr-Al alloys

Variable

Symbol

Variability of values min max

Ni Cr Al

X1 X2 X3

10.60 14.30 5.90

31.00 30.20 12.00

Y T, oC

X4 X5

0.1 …

0.7 …

Thermal expansion coefficient, range 20– 400 oC Regression equation: ␣ ⳱ 14.5 ⳮ 0.21 X1 ⳮ 0.68 X2 ⳮ 0.32 X3 Ⳮ 0.21 X1 X2 Ⳮ 0.14 X1 X3 ⳮ ⳮ 0.56X2X3 ⳮ 0.26X1X2X3, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.04; (DSj2) ⳱ 0.95; Rf ⳱ 0.97 Thermal expansion coefficient, range 20– 500 oC Regression equation: ␣ ⳱ 14.8 ⳮ 0.22X1 ⳮ 0.76X2 ⳮ 0.36X3 Ⳮ 0.16X1X2 ⳮ 0.62X2X3 ⳮ 0.32X1X2X3, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.03; (DSj2) ⳱ 0.96; Rf ⳱ 0.97 Thermal expansion coefficient, range 20– 600 oC Regression equation: ␣ ⳱ 15.4 ⳮ 0.22X1 ⳮ 0.78X2 ⳮ 0.42X3 Ⳮ 0.31X1X2 Ⳮ 0.18X1X3 ⳮ 0.62X2X3 ⳮ ⳮ 0.27X1X2X3, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.03; (DSj2) ⳱ 0.97; Rf ⳱ 0.98 Thermal expansion coefficient, range 20– 700 oC Regression equation: ␣ ⳱ 16.0 ⳮ 0.27X1 ⳮ 1.01X2 ⳮ 0.37X3 Ⳮ 0.69X22 Ⳮ 1.05X1X2 Ⳮ 0.68X1X3 ⳮ ⳮ 0.29X2X3, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.28; (DSj2) ⳱ 0.93; Rf ⳱ 0.96 Thermal expansion coefficient, range 20– 800 oC Regression equation: ␣ ⳱ 16.0 ⳮ 0.96X2 Ⳮ 0.67X12 Ⳮ 0.83X22 Ⳮ 1.4X1X2 Ⳮ 0.75X1X3 ⳮ 0.35X2X3, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.52; (DSj2) ⳱ 0.88; Rf ⳱ 0.94

Appendix / 215

Thermal expansion coefficient, range 20– 900 oC Regression equation: ␣ ⳱ 16.1 ⳮ 0.99X2 Ⳮ 0.84X12 Ⳮ 0.73X22 Ⳮ 1.24X1X2 Ⳮ 0.64X1X3 ⳮ 0.32X2X3, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.51; (DSj2) ⳱ 0.87; Rf ⳱ 0.93 Thermal expansion coefficient, range 20– 1000 oC Regression equation: ␣ ⳱ 16.5 ⳮ 0.93X2 Ⳮ 0.52X12 Ⳮ 0.31X22 Ⳮ 0.62X32 Ⳮ 1.01X1X2 Ⳮ 0.62X1X3, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.31; (DSj2) ⳱ 0.89; Rf ⳱ 0.94 Elasticity modulus, 20–800 oC Regression equation: E ⳱ 1626 ⳮ 99.9X1 Ⳮ 98.4X2 Ⳮ 103.9X3 ⳮ 254.1X5 Ⳮ 113.4X12 Ⳮ 109.2X22 ⳮ ⳮ 48.4X32 ⳮ 67.8X52 ⳮ 70.9X1X2 ⳮ 48.3X1X3 Ⳮ 22.3X1X5 Ⳮ 125.1X2X3 Ⳮ Ⳮ 31.1X2 X5 Ⳮ 55.4 X3 X5 ⳮ 59.9 X1 X2X 3 ⳮ 11.9 X1 X2 X5 Ⳮ 24.0X2 X3 X5, (⳯100, MPa) S2na ⳱ 1639; (DSj2) ⳱ 0.96; Rf ⳱ 0.98 Ultimate strength, 700–1100 oC Regression equation: rb ⳱ 16.7 ⳮ 1.67X1 ⳮ 22.93X5 Ⳮ 2.52X22 Ⳮ 2.55X32 Ⳮ 9.90X52 ⳮ 1.41X3X5, (⳯10, MPa) S2na ⳱ 7.52; (DSj2) ⳱ 0.98; Rf ⳱ 0.99 Yield strength, 700–900 oC Regression equation: r0.2 ⳱ 24.3 ⳮ 2.50X1 ⳮ 4.29X2 Ⳮ 0.24X3 ⳮ 12.36X5 Ⳮ 1.46X12 Ⳮ 5.97X22 ⳮ ⳮ 16.81X32 Ⳮ 1.53X2X5 Ⳮ 36.36X1X2X5 ⳮ 33.31X1X3X5, (⳯10, MPa) S2na ⳱ 0.90; (DSj2) ⳱ 0.99; Rf ⳱ 0.99 Relative elongation, 700–1100 oC Regression equation: d ⳱ 55.1 Ⳮ 12.40X1 ⳮ 5.93X2 Ⳮ 16.89X5 ⳮ 10.05X12 ⳮ 10.43X52 Ⳮ 6.74X1X5 ⳮ ⳮ 46.30X1X2X5 Ⳮ 56.02X1X2X3X5, (%) S2na ⳱ 47.15; (DSj2) ⳱ 0.93; Rf ⳱ 0.96 Hot corrosion, 700 oC, 200 h Regression equation: qⳮ ⳱ 0.57 Ⳮ 0.13 X1 ⳮ 0.21 X2 Ⳮ 0.29 X12 Ⳮ 0.42 X42 ⳮ 0.28 X1X2 ⳮ

ⳮ 0.8424501 X2 X3 Ⳮ 0.56 X1 X2 X3 X4, (g/ m2) S2na ⳱ 0.08; (DSj2) ⳱ 0.85; Rf ⳱ 0.92 Hot corrosion, 800 oC, 200 h Regression equation: qⳮ ⳱ 9.56 Ⳮ 2.34X1 Ⳮ 2.93X2 Ⳮ 1.94X1X2 ⳮ 2.93X1X3 ⳮ 4.20X2X3 ⳮ ⳮ 3.36X1X2X3, (g/m2) S2na ⳱ 18.85; (DSj2) ⳱ 0.80; Rf ⳱ 0.89 Hot corrosion, 850 oC, 200 h Regression equation: qⳮ ⳱ 31.11 Ⳮ 8.73X1 Ⳮ 17.24X2 Ⳮ 7.96X3 Ⳮ 8.38X12 ⳮ 2.09X22 Ⳮ 13.91X1X2 Ⳮ Ⳮ 15.54X1X3 Ⳮ 11.73X2X3 Ⳮ 13.17X1X2 X3, (g/m2) 2 S na ⳱ 15.93; (DSj2) ⳱ 0.95; Rf ⳱ 0.97 Hot corrosion, 900 oC, 200 h Regression equation: qⳮ ⳱ 36.43 Ⳮ 3.88X1 ⳮ 8.31X2 Ⳮ 6.99X3 Ⳮ 20.36X22 ⳮ 6.73X32 Ⳮ 5.36X1X2 Ⳮ 16.29X1X3 ⳮ 7.05X1X2X3, (g/m2) S2na ⳱ 85.33; (DSj2) ⳱ 0.87; Rf ⳱ 0.93 Heat resistance, cyclic test 20 } 900 oC, 1000 h Regression equation: qⳭ ⳱ 8.48 (11.01 X2 Ⳮ 6.56 X12 ⳮ 3.78 X22 ⳮ 0.79 X32 ⳮ 5.40 X42 ⳮ 14.62 X2 X4, (g/m2) 2 S na ⳱ 0.06; (DSj2) ⳱ 0.99; Rf ⳱ 0.99 Heat resistance, cyclic test 20 } 900 oC, 5000 h Regression equation: qⳭ ⳱ 11.88 ⳮ 9.22X2 Ⳮ 1.82X4 Ⳮ 2.72X22 ⳮ 4.06X42 ⳮ 6.06X2X4 Ⳮ Ⳮ 3.59X1X2X3X4 , (g/m2) S2na ⳱ 0.65; (DSj2) ⳱ 0.99; Rf ⳱ 0.99 Heat resistance, cyclic test 20 } 900 oC, 8000 h Regression equation: qⳭ ⳱ 12.05 Ⳮ 23.48 X4 Ⳮ 16.08 X42 ⳮ 18.52 X1 X2 ⳮ 11.40 X1 X2 X3 ⳮ ⳮ 29.50X1X2X4 Ⳮ 8.85X2X3X4, (g/m2) S2na ⳱ 0.59; (DSj2) ⳱ 0.98; Rf ⳱ 0.99


S2na ⳱ 0.31; (DSj2) ⳱ 0.74; Rf ⳱ 0.86

Ni-Cr-Al alloys

Variable Cr Al T, oC

Symbol X1 X2 X3

Variability of values min max 7.00 2.80 …

22.5 18.00 …

Thermal expansion coefficient, range 100– 300 oC Regression equation: ␣ ⳱ 13.4 ⳮ 0.66 X1 ⳮ 1.14 X2, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.12; (DSj2) ⳱ 0.77; Rf ⳱ 0.87 Thermal expansion coefficient, range 100– 400 oC Regression equation: ␣ ⳱ 13.84 ⳮ 0.63 X1 ⳮ 1.08 X2, (x10ⳮ6, Kⳮ1) S2na ⳱ 0.13; (DSj2) ⳱ 0.73; Rf ⳱ 0.85 Thermal expansion coefficient, range 100– 500 oC Regression equation: ␣ ⳱ 13.8 ⳮ 0.63 X1 ⳮ 1.08 X2, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.13; (DSj2) ⳱ 0.73; Rf ⳱ 0.85 Thermal expansion coefficient, range 100– 600 oC Regression equation: ␣ ⳱ 14.3 ⳮ 0.59 X1 ⳮ 1.02 X2 Ⳮ 0.49 X22, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.26; (DSj2) ⳱ 0.68; Rf ⳱ 0.82 Thermal expansion coefficient, range 100– 700 oC Regression equation: ␣ ⳱ 14.5 ⳮ 0.70 X1 ⳮ 1.45 X2 Ⳮ 0.59 X12 Ⳮ 0.53 X22, (⳯10ⳮ6, Kⳮ1)

Thermal expansion coefficient, range 100– 800 oC Regression equation: ␣ ⳱ 14.7 ⳮ 1.01 X1 ⳮ 2.12 X2 Ⳮ 0.85 X12, (⳯10ⳮ6, Kⳮ1) S2na ⳱ 0.59; (DSj2) ⳱ 0.67; Rf ⳱ 0.82 Thermal expansion coefficient, range 100– 900 oC Regression equation: ␣ ⳱ 15.0 ⳮ 1.06 X1 ⳮ 2.17 X2 Ⳮ 0.90 X12, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 0.83; (DSj2) ⳱ 0.60; Rf ⳱ 0.77 Thermal expansion coefficient, range 100– 1000 oC Regression equation: ␣ ⳱ 16.5 ⳮ 1.25 X2 Ⳮ 0.63 X1 X2, (⳯10ⳮ6, Kⳮ1) 2 S na ⳱ 1.00; (DSj2) ⳱ 0.46; Rf ⳱ 0.68 Elasticity modulus, 20–800 oC Regression equation: E ⳱ 1557 ⳮ 87.5X1 ⳮ 183.1X2 ⳮ 122.8X3 Ⳮ 92.2X12 ⳮ 187.5X22 ⳮ 22.2X32 Ⳮ Ⳮ 414.5X1X2 Ⳮ 114.3X2X3 Ⳮ 66.8X1X2X3, (⳯100, MPa) S2na ⳱ 2754; (DSj2) ⳱ 0.94; Rf ⳱ 0.97 REFERENCE

1. Y.M. Doljansky, G.B. Stroganov, and R.E. Shalin, Optimization of Machine-Building Material Properties with a Computer, Moscow, Voenizdat, 1979, 240 p



Index A Accumulated cancelled ratio (ACR), of variances of the given increment of model selection to the sum of variances cancelled at the increments from the first to the current one ................................................ 212 ACR. See Accumulated cancelled ratio. Activity coefficients of each component ..........39 Adhesion, of overlay coatings, testing of ..........52 Adhesive strength of ceramic layer/bond coat ...................... 162 of ceramic layer of thermal barrier coatings ....... 187(F), 188–189(F), 190, 206 of thermal barrier coatings ....................21–22 of thermal barrier coatings, criterion of .. 206, 207 Aging, with aluminizing ..............................29 Aircraft turbine blades, superalloy chemical compositions ............................. 10, 11(T) Air plasma spraying of powders (TBC-APS) ..................................... 162 Alitizing .................................................18 Alkaline metal salts ................................... 6 Alloying factor (K) ...............................18–19 Alpha-aluminum oxide ................. 83, 101, 189 of diffusion coating/superalloys system surfaces, heat resistance testing results .............. 106 formation on superalloy surface ............. 81, 83 Alpha-aluminum oxide chromium oxide, formation with overlay coatings on JS6U superalloy ................................... 109(T) Alpha-chromium phase of nickel-chromium-aluminum system .. 87, 88(F) in nickel-cobalt-chromium-aluminum alloys ....70 Alpha-molybdenum phase, of NiAl ............. 101 Alpha-tungsten phase, of NiAl ................... 101 Alumina/chromia coatings, to protect superalloys from oxidation during 10,000 hours ...11–12 Alumina-type oxides, formation on ceramic-metal interface with thermal barrier coatings ........22 Aluminides formed, by aluminizing nickel alloys ............55 heat resistance testing ............... 100–106(F,T) Aluminizing ............................................. 8

with aging ............................................29 aluminum content decrease with reused mixture applications ....................................30 annealing of powders to increase stability under H2 atmosphere .................................29 causes .................................................26 circulation technique for diffusion coating application ............................. 31–32(F) cleaning for preparation before ...................29 halide activator amount affecting thickness .................................... 26(F) nickel alloys, aluminides formed .................55 outcomes .............................................26 powder method of application ....................30 in powder mixtures with activators ...............25 and silicification ...................................... 1 temperature range with halide activators ........25 time considerations .................................29 Aluminum amount the mixture feeds to the saturated surface in aluminizing .................................26 bonded ................................................. 1 coefficient of transfer in flux, electric arc evaporation .................................40(T) consumption variation with temperature level ............................... 16–17, 18(F) content effect of Ni-Co-Cr-Al alloys on thermal expansion coefficients ................ 122, 123 content effect on heat resistance of Co-Ni-Cr-Al alloys .......................................... 106 content effect on hot corrosion and phase composition of Ni-Cr-Al alloys ........... 111 content effect on hot corrosion resistance of Co-Ni-Cr-Al alloys ................... 112, 113 content effect on hot corrosion resistance of Ni-Co-Cr-Al alloys ......................... 112 content effect on protective nature of overlay coatings on superalloys .................... 109 content effect on thermal expansion coefficient of Co-Ni-Cr-Al alloys ......................... 124 content effect on thermal expansion coefficient of Ni-Cr-Al alloys .............................. 121 content effect on thickness of beta-phase-depleted outer zone ....................103–104, 105(F)



218 / Index

Aluminum (continued) content in diffusion coatings affecting ductilebrittle transition temperature .............. 120 content in NiAl compounds and thermal fatigue ......................................... 145 degree of ionization .............................40(T) free ..................................................... 1 as impurity of jet fuels .............................. 5 mass-spectrometric analysis data for material evaporation .................................40(T) Aluminum-boron alloys, slurry technique for depositing diffusion coatings ...................35 Aluminum chlorides, formation of .................25 Aluminum-chromium alloys circulation technique for depositing diffusion coatings .........................................32 as powder mixtures used in aluminizing ........27 slurry technique for depositing diffusion coatings .........................................35 Aluminum halides ..................................... 1 Aluminum iodides, use in circulation technique for diffusion coating application ...................32 Aluminum-iron alloys, as powder mixtures used in aluminizing ....................................27 Aluminum mass fraction in a definite zone of coating (Xi) ..................................17–18 Aluminum oxide as corrosion product penetrating the coating ....13 formation at ceramic/bond-coat interface ................................ 205, 206 formation due to interaction of overlay coating and superalloy .................................96 Aluminum powder, content effect on coatings deposited by slurry technique ..................35 Aluminum reserve calculation of value ............................17–18 as criterion for coating life expectancy ..... 17, 19 definition ........................................ 18, 69 of diffusion aluminide coatings ......... 15–16, 38 with gamma double-prime phase compound coatings .........................................81 and high-temperature testing ............. 84–85(F) parameter introduced in coating ..................99 significance ...........................................69 Aluminum-silicon alloys as base system for development of electric arc coating compositions .........................36 circulation technique for depositing diffusion coatings .........................................32 slurry technique for depositing diffusion coatings ..................................... 35(F) Aluminum-silicon-yttrium alloy, as diffusion layer, phase transformations during annealing of superalloys ............................ 62(F), 63 Aluminum-to-surface transfer .....................31 Alundum crucibles ................................. 100 Ammonium chloride, and activator distribution for use ..............................................29 Amylacetate .................................. 33, 34, 35 Angle of vapor incidence, and ceramics blackening ............................. 169, 170(T)

Annealing ................................... 130(F), 131 after electric arc technique of applying diffusion aluminide coatings ............................37 powder mixture of diffusion aluminide coatings .........................................30 Annealing temperature, elevation effect on aluminum concentration in diffusion aluminide coatings ............................................60 Approximate life expectancy of a new coating ........................................14–15 Arc plasma spraying .................................. 8

B Barium, as impurity of jet fuels ...................... 5 Beam model .......................................... 196 Bending stresses ........................................ 8 Beta-phase, in cobalt-nickel-chromium-aluminum alloys ...............................................71 Beta-phase-free zone after heat resistance testing at 1200C, 200h, regression equation for calculation of ................................... 214 Beta-phase nickel aluminide (NiAl) ........... 2, 55 alloying with molybdenum heat resistance effect .......................................... 101 alloying with silicon, heat resistance effect ... 101 alloying with tungsten, heat resistance effect .......................................... 101 alloying with zirconium, heat resistance effect .......................................... 101 aluminum content ............................. 119(T) chromium solubility in Ni-Cr-Al system alloys .......................................... 111 color ...............................................56(T) crystalline structure .............................56(T) density ............................................56(T) in diffusion coating formed in high-activity media ........... 59, 60(T), 61(F), 62(F,T), 63 in diffusion coating formed in moderate and lowactivity media ......... 63–69(F,T), 70, 72(F) disappearance in coating outer zone .............98 elongation ...................................... 119(T) elongation vs. temperature ................... 120(F) hardness ..........................................56(T) heat of formation ................................56(T) heat resistance at high temperatures101(F) high-temperature strength ........................ 120 homogeneity region .............................56(T) melting temperature .............................56(T) of Ni-Cr-Al system ........................ 87, 88(F) oxidation mechanism ............................. 101 oxidation time ................................. 100(T) oxide structure ................................. 100(T) phase transformations in Al5SiY diffusion coatings after high-temperature tests ..................................84(F), 85(F) in phase transformations of overlay coatings on superalloys ....... 89(F), 90(F), 91(F), 92(T), 93(F), 95(T)



Index / 219

quantity, annealing 1200C, 200h, regression equation for calculation of ................. 213 quantity, regression equation for calculation of .............................................. 213 specific volume ..............................128–129 two-way diffusion in diffusion coatings .... 80(F), 81(T), 82(F), 83(T), 84(F), 85(F) ultimate strength ............................... 119(T) ultimate strength vs. temperature ........... 120(F) yield strength .................................. 119(T) yield strength vs. temperature ............... 120(F) x-ray diffraction analysis ..................... 100(T) Black ceramics oxidation kinetics ............................. 169(F) phase composition of ......................... 169(F) Blade fatigue limit, thermal fatigue of overlay coatings ............................................. 9 Blade leading-edge zone ........................... 206 Blade pressure surfaces, preparation for overlay coatings ............................................45 Blade suction surfaces, preparation for overlay coatings ............................................45 Blade-wall thickness value ........................ 197 Blasting, as preparation of inner cavity for aluminizing ........................................31 Bond coats ............................................ 161 effect of technological factors on element contents of surface ................ 167, 168(T) oxidation ............................................ 182 plastic strain consideration ....................... 162 preoxidizing ........................................ 167 protecting blade from oxidation ..................21 surface preparation ................166–167, 168(T) surface roughness ................................. 166 Boron, as alloying element for diffusion coating ........................................... 116 Burners .................................................. 8 Burnouts ................................................. 7

C Calcium, as impurity of jet fuels ..................... 5 Carbide phase precipitation Cr7C3 in diffusion coatings on Ni20CrTi alloy ....................................80–81(T) Cr23C6 in diffusion coatings on Ni20CrTi alloy ....................................80–81(T) of diffusion coatings .......... 60(T), 65(T), 83(T), 84(F), 85(F) in interaction mechanism between overlay coatings and superalloys .....................93 on nickel superalloy Ni20CrTi ....................81 of overlay coatings in high-temperature tests ..87, 88(F), 89(F), 90(F), 92(T), 93(F), 94(F), 95(T) Carbides, effect on NiAl with alloying elements ......................................... 101 Carbon ................................................... 6 as impurity of jet fuels .............................. 5

Carbon dioxide ......................................... 6 Carbon hydroxide (CH4) ............................. 6 Carbon monoxide ...................................... 6 Cathode material transfer coefficient (l) ........40 Cathode spots ..........................................39 temperature of ...................................40(T) vapor pressure ...................................40(T) Centrifugal force, in coated blade ................ 198 Centrifugal load in a blade, value correction .. 197 Centrifugal load in a given blade cross section ........................................... 197 Centrifugal loading, blade-mass gain from coating application ............................. 195 Ceramic coating compressive stress ............ 184 Ceramic coatings. See also Ceramic condensates; Thermal barrier coatings. effects on blade-life expectancy ..................15 requirements .......................................... 8 Ceramic layer blackening ......................... 169 Ceramic layer modulus of elasticity ............ 184 Ceramic layer spalling. See also Spalling ...... 162 Ceramics for electron-beam evaporation ..... 162–165(F,T) spalling on turbine blades with thermal barrier coatings ................................ 20–22(F) Ceramics condensates, specific types ZrO2, phase composition ..................... 164(T) ZrO2-20%CeO2 color .......................................... 177(T) density ....................................... 177(T) EB, thermophysical properties ........... 178(T) manufacturing method ..................... 177(T) thickness ..................................... 177(T) ZrO2-7%Y2O3-5%NiO, thermal expansion coefficients ............................... 181(T) ZrO2-7%Y2O3-15%NiO, thermal expansion coefficients ............................... 181(T) ZrO2-8%Y2O3 ................................. 164(T) APS, thermophysical properties ......... 178(T) as ceramic layer of thermal barrier coating, service life .................186–187(F), 188 as ceramic layer tested for adhesive strength ................ 187(F), 188–189(F) as ceramic layer tested on gas-dynamic test rigs ................................... 189, 190 color .......................................... 177(T) density ....................................... 177(T) EB, thermophysical properties ........... 178(T) heat capacity versus temperature ........ 178(F) manufacturing method ..................... 177(T) maximum principal stress ................. 205(F) phase composition ......................... 164(T) powder, heat capacity versus temperature ............................ 178(F) spalling ...................................... 185(F) thermal expansion coefficient ............... 180, 181(T), 182(F) thickness ..................................... 177(T) ZrO2-8%Y2O3-5%NiO ceramic layer tested on gas-dynamic test rigs ......................................... 190



220 / Index

Ceramics Condensates (continued) color .......................................... 177(T) density ....................................... 177(T) EB, thermophysical properties ........... 178(T) heat capacity versus temperature ........ 178(F) manufacturing method ..................... 177(T) powder, heat capacity versus temperature ............................ 178(F) thickness ..................................... 177(T) ZrO2-11%Y2O3, emissivity, coating ........ 181(T) ZrO2-15%Y2O3 ................................ 164(T) phase composition ......................... 164(T) powder, heat capacity versus temperature ............................ 178(F) thermal expansion coefficients ........... 181(T) Chemical compositions diffusion aluminide coatings ............. 10, 11(T) of overlay coatings ..............................10(T) superalloys for aircraft turbine blades ... 10, 11(T) Chlorine, in deposits detected on gas-turbine engine blade surfaces ............................99 Chromium addition effect on beta-phase depleted outer zone ........................................... 104 addition effect on saturating mixtures for aluminizing ................................ 27, 29 as alloying element for diffusion coating ...... 116 as alloying element, retarding of aluminum diffusion ........................................57 alloying with aluminide to retard diffusion processes .......................................81 coefficient of transfer in flux, electric arc evaporation .................................40(T) consumption variation with temperature level ............................... 16–17, 18(F) content effect of Ni-Co-Cr-Al alloys on thermal expansion coefficients ................ 122, 123 content effect on heat resistance of Co-Ni-Cr-Al alloys .......................................... 106 content effect on hot corrosion resistance of Co-Ni-Cr-Al alloys ................... 112, 113 content effect on hot corrosion resistance of Ni-Cr-Al alloys .............................. 111 content effect on hot corrosion resistance of Ni-Co-Cr-Al alloys ...................111–112 content effect on protective nature of overlay coatings on superalloys .................... 109 content effect on thermal expansion coefficients of Co-Ni-Cr-Al alloys ...................... 124 content effect on thermal expansion coefficients of Ni-Cr-Al alloys ........................... 121 content effect on thickness of interaction zone of overlay coating and superalloy ..... 93, 95(F) in diffusion aluminide coatings .................... 1 diffusion in overlay coatings ......................88 diffusion into interaction zone with cobalt content addition in overlay coatings .................91 diffusion with interaction mechanism of overlay coating and superalloy at high temperatures ..................89, 91(F), 92(T)

Chromium-base carbides, stable phases formed from interaction mechanism ....................94 Chromium-base phases ..............................82 Chromium mass fraction (XCr) .....................19 Chromium oxide, as corrosion product penetrating the coating .........................................13 Chromizing-aluminizing method ................... 1 Circulation technique, for applying diffusion coatings ................................... 31–32(F) Circumferential elasticity modulus ............. 151 Circumferential thermal stresses ................ 151 Cleaning, before aluminizing ........................29 Coastal regions ......................................... 9 Coated-blade service life, consideration affecting calculations ...................................... 195 Coating cyclic life expectancy ......................20 Coating ductility .................................... 198 Coating ductility parameter ...................... 201 Coating emissivity ........................ 180, 181(T) Coating lifetime. See Coating protective lifetime. Coating-mass effect on blade centrifugal load, coefficient of .................................... 197 Coating protective lifetime ..................... 98, 99 related to aluminum reserve .......................19 techniques for determination .................... 100 Coatings efficiency showing coating effect on superalloy properties ..................................... 119 factors affecting mechanical properties of superalloys ................................... 119 multilayer systems ............................ 134(F) Coating serviceability criterion .................. 198 Coating service life. See also Service life. ......................................... 155, 162 definition .............................................14 and thermomechanical fatigue testing .......... 157 Coatings, specific types aluminized life of diffusion coatings on JS6U superalloys ............................ 115(F) low-cycle fatigue ................. 149(F), 150(T) thermal shock testing of coating/superalloy systems ................................ 148(T) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) aluminum oxide life on JS6U superalloy, cyclic testing .................................. 108(T) thickness of coatings on JS6U superalloy ............................. 109(T) aluminum palladium (AlPd) coating life (hot corrosion test in GZT ash) ..................................... 115(T) coating life on JS6U superalloys ........ 116(T) life coefficient (hot corrosion test on GZT ash) ..................................... 116(T) life of diffusion coatings on JS6U superalloy ............................. 115(F) oxide life on JS6U superalloy, cyclic testing .................................. 108(T)



Index / 221

thickness of coatings on JS6U superalloy ............................. 109(T) AlSiBY, coating life (hot corrosion test in GZT ash) ................................. 115(T) AlSiCrY, coating life (hot corrosion test in GZT ash) ................................. 115(T) Al5SiY phase transformations diagram ............. 84(F) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) Al5Si1.5Y creep-rupture strength ..... 134, 135(F), 136(F) electric arc coating, life of diffusion coatings on JS6U superalloy .................. 115(F) elongation of JS6U superalloy ..... 136, 137(F) heat resistance of JS6U superalloy with diffusion coating ..................... 108(F) Al10Si0.7Y heat resistance of JS6U superalloy with diffusion coating ..................... 108(F) oxide life on JS6U superalloy, cyclic testing .................................. 108(T) thickness of coatings on JS6U superalloy ............................. 109(T) Al14Si0.7Y electric arc coating, life of diffusion coatings on JS6U superalloy .......... 115(F) Al15Si0.5Y oxide life on JS6U superalloy, cyclic testing .................................. 108(T) thickness of coatings on JS6U superalloy ............................. 109(T) Coatings, specific types, Cobalt-chromiumaluminum alloys Co20Cr11AlY oxide life on JS6U superalloy, cyclic testing .................................. 108(T) thickness of coatings on JS6U superalloy ............................. 109(T) Co20Cr11AlY/ZrO2-15%Y2O3 oxide life on JS6U superalloy, cyclic testing .................................. 108(T) thickness of coatings on JS6U superalloy ............................. 109(T) Co20Cr12AlY chemical composition .......................10(T) density ....................................... 181(T) elongation vs. temperature ................ 126(F) heat capacity ................................ 181(T) temperature .................................. 181(T) thermal conductivity ....................... 181(T) thermal diffusivity ......................... 181(T) ultimate strength vs. temperature ........ 126(F) Co22Cr9AlY chemical composition .......................72(T) microstructure ................................ 73(F) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co22Cr11AlY coating life ...................................... 116

coating life, hot corrosion test in GZT ash .................................... 115(F,T) coating life on JS6U superalloys ........ 116(T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co22Cr11AlY/ZrO2-15%Y2O3 coating life (hot corrosion test in GZT ash) ................................... 115(F,T) coating life on JS6U superalloys ........ 116(T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) weight change (hot corrosion test in GZT ash) ..................................... 114(F) Co22Cr12Al0.4Y, bond coat service life .............................183(F), 186–187 Co22Cr12AlY, corrosion lives after sulfur attack from ash ............................ 186(F), 188 Co23Cr12AlY, chemical composition .......10(T) Co24Cr5Al0.4Y, for bond coat, service life ................................... 186(F), 188 Co24Cr12Al0.4Y, for bond coat, service life ................................... 186(F), 188 Co25Cr4AlY, chemical composition .........10(T) Co26Cr9AlY, chemical composition .........10(T) Co28Cr5AlY, chemical compositions of cathode and deposited coating ............... 40, 41(T) Co30Cr6AlY, chemical composition .........10(T) Cobalt-chromium-iron alloys Co28Cr10FeY, chemical composition .......10(T) Cobalt-nickel-chromium-aluminum alloys Co8Ni23Cr12AlY, chemical composition ...10(T) Co10Ni15Cr6AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) thermal expansion coefficients vs. temperature .............................. 74(F) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co10Ni15Cr12AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) normalized number of variable ........... 212(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) thermal expansion coefficients vs. temperature .............................. 74(F) weight loss (hot corrosion test in GZT ash) ..................................... 113(T)



222 / Index

Coatings, (continued) Co10Ni22Cr9AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) ultimate strength, various temperatures .......................... 127(T) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Co10Ni30Cr6AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) thermal expansion coefficients vs. temperature .............................. 74(F) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co10Ni30Cr12AlY chemical composition .............72(T), 212(T) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) thermal expansion coefficients vs. temperature .............................. 74(F) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co20Ni15Cr9AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) ultimate strength, various temperatures .......................... 127(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T)

yield strength, various temperatures ..... 127(T) Co20Ni22Cr6AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) ultimate strength, various temperatures .......................... 127(T) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Co20Ni22Cr9AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) ultimate strength, various temperatures .......................... 127(T) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Co20Ni22Cr12AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) normalized number of variable ........... 212(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) ultimate strength, various temperatures .......................... 127(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Co20Ni30Cr9AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T)



Index / 223

ultimate strength, various temperatures .......................... 127(T) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Co22Ni23Cr12AlY, chemical composition ................................10(T) Co25Ni20Cr12AlY, cyclic testing and fragmentation ......................... 20, 21(F) Co30Ni15Cr6AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) thermal expansion coefficients vs. temperature .............................. 74(F) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co30Ni15Cr12AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) thermal expansion coefficients vs. temperature .............................. 74(F) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co30Ni22Cr9AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) elongation, various temperatures ........ 128(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T) ultimate strength, various temperatures .......................... 127(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Co30Ni30Cr6AlY chemical composition .............72(T), 212(T) elasticity modulus values, various temperatures .......................... 125(T) normalized number of variable ........... 212(T) oxide life .................................... 106(T) phase composition ...........................72(T) thermal expansion coefficients, various temperatures .......................... 125(T)

thermal expansion coefficients vs. temperature .............................. 74(F) weight gain (heat resistance cyclic test) ..................................... 106(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co30Ni30Cr12AlY chemical composition .............72(T), 212(T) microstructure ................................ 73(F) normalized number of variable ........... 212(T) phase composition ...........................72(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Co32Ni21Cr8AlY chemical composition and other characteristics ......................... 153(T) coatings service life ............................ 157 cyclic life of coatings/superalloy CMSX-4 system ................ 155, 157(F,T), 158(F) EB-TBC, thermal barrier coating testing of MAR-M-002 superalloy on gas-dynamic test rigs and engines ................... 189(F), 190 Iron-chromium-aluminum alloys Fe22Cr11AlY, weight loss (hot corrosion test in GZT ash) .............................. 113(T) Fe25Cr5AlY, chemical composition .........10(T) Coatings, specific types, Iron-nickel-chromiumaluminum alloys Fe12 Ni20Cr12Al1SiY oxide life ........................................ 109 oxide life on JS6U superalloy, cyclic testing .................................. 108(T) Fe12Ni20Cr12Al1.5SiY coating life, hot corrosion test in GZT ash .............................115(F,T) coating life on JS6U superalloys ........ 116(T) life coefficient (hot corrosion test in GZT ash) .............................. 116(T) thickness of coatings on JS6U superalloy ............................. 109(T) Fe22Ni24Cr5AlY, chemical composition ...10(T) Coatings, specific types, Nickel-aluminum alloys Ni31Al, thermal expansion coefficients, various temperatures ............................. 121(T) NiAl (33% Al) elongation vs. temperature ................ 126(F) ultimate strength vs. temperature ........ 126(F) Nickel-chromium-aluminum alloys Ni5Cr8AlY elongation ................................... 129(F) ultimate strength ................127–128, 129(F) Ni5Cr25Al calculated number of cycles to crack appearance ............................ 203(T) chemical composition .......................65(T) elongation ................................... 126(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70



224 / Index

Coatings (continued) ultimate strength ............................ 126(T) Ni7Cr10AlY chemical composition .............65(T), 212(T) normalized number of variable ........... 212(T) phase composition ...........................65(T) Ni7Cr12AlY calculated number of cycles to crack appearance ............................ 203(T) chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) elasticity modulus .............................. 122 elasticity modulus values, various temperatures .......................... 122(T) hot corrosion resistance ....................... 111 maximum cyclic life ........................... 203 microstructure ................................ 66(F) normalized number of variable ........... 212(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70 weight loss due to hot corrosion ......... 110(T) Ni8Cr15AlY chemical composition .............65(T), 212(T) elasticity modulus values, various temperatures .......................... 122(T) elongation ................................... 126(T) microstructure ................................ 66(F) normalized number of variable ........... 212(T) phase composition ...........................65(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70 ultimate strength ...................... 124, 126(T) Ni8Cr18AlY calculated number of cycles to crack appearance ............................ 203(T) chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) elasticity modulus .............................. 122 elasticity modulus values, various temperatures .......................... 122(T) microstructure ................................ 66(F) normalized number of variable ........... 212(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) weight loss due to hot corrosion ......... 110(T) Ni10Cr17AlY service life vs. strain-intensity range ........ 200 thermomechanical fatigue tests .......... 200(F) Ni14Cr3AlY chemical composition .............65(T), 212(T) crystallization intervals, 49(T) elasticity modulus values, various temperatures .......................... 122(T) liquidus-solidus temperatures ..............49(T) microstructure ................................ 66(F) normalized number of variable ........... 212(T) phase composition ...........................65(T)

thermal expansion coefficients, various temperatures .......................... 121(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70 Ni15Cr3AlY, elasticity modulus ................ 122 Ni16Cr9AlY chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 122(T) hot corrosion resistance ....................... 111 liquidus-solidus temperatures ..............49(T) microstructure ................................ 66(F) normalized number of variable ........... 212(T) oxide life at 900C ......................... 102(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70 weight gain (cyclic heat resistance tested) .................................. 102(T) weight gain (isothermal heat resistance tested) .................................. 102(T) weight loss due to hot corrosion ......... 110(T) Ni16Cr11AlY, phase transformations, interaction mechanism with superalloy at high temperatures .....87–88(F), 89, 90(F), 91(F) Ni18Cr5AlY chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 122(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 212(T) oxide life at 900C ......................... 102(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70 Ni18Cr15AlY, thermal expansion coefficients, various temperatures .................... 121(T) Ni18Cr16AlY chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 122(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 212(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) weight loss due to hot corrosion ......... 110(T) Ni19Cr10AlY, thermal expansion coefficients, various temperatures .................... 121(T) Ni19Cr12AlY



Index / 225

chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 122(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 212(T) oxide life at 900C ......................... 102(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) weight gain (cyclic heat resistance tested) .................................. 102(T) weight gain (isothermal heat resistance tested) .................................. 102(T) weight loss due to hot corrosion ......... 110(T) Ni20Cr5AlY chemical composition .......................10(T) ductility vs. temperature .................. 199(F) strain intensity vs. temperature ........... 199(F) Ni20Cr6AlY chemical composition .............65(T), 212(T) density ....................................... 181(T) elongation ................................... 126(T) heat capacity ................................ 181(T) hot corrosion resistance, chromium-aluminum content ratio ............................... 111 normalized number of variable ........... 212(T) phase composition ...........................65(T) preparation of surface before coating deposition ...................................46 temperature .................................. 181(T) thermal conductivity ....................... 181(T) thermal diffusivity ......................... 181(T) thermal expansion coefficients, various temperature ............................ 121(T) ultimate strength ...................... 124, 126(T) Ni20Cr6AlY/Ni20Cr12AlY/ZrO2-8%Y2O3, fatigue strength .............................. 141 Ni20Cr7AlY elongation ................................... 129(F) elongation vs. temperature ................ 126(F) ultimate strength ................127–128, 129(F) ultimate strength vs. temperature ........ 126(F) Ni20Cr9.5AlY, phase transformations, interaction mechanism between overlay coating and superalloy at high temperatures .... 88, 89(F) Ni20Cr11Al, hafnium addition effect on service life of protective oxides .................... 101 Ni20Cr11Al1Hf, oxide life at 900C ....... 102(T) Ni20Cr11Al2Hf, oxide life at 900C ....... 102(T) Ni20Cr11AlY, cobalt effect on oxidation ..... 103 Ni20Cr12AlY adhesion strengths for various preparation and deposition techniques ..................51(T) blade elongation vs. service time .......... 22(F) bond coat for thermal barrier coating, bend testing .................................. 175(F) bond coat on thermal barrier coating .... 171(F) as bond coat, spalling of CMSX-4 superalloy surface ..................................... 188

as bond coat tested for adhesive strength ................ 187(F), 188–189(F) bond coat tested on gas-dynamic test rigs ................................... 189, 190 calculated number of cycles to crack appearance ............................ 203(T) chemical composition .......................10(T) coating thickness distribution across turbine blade middle section ................... 49(F) cyclic life ........................................ 202 cyclic testing and fragmentation ...... 20, 21(F) density ....................................... 181(T) ductility vs. temperature .................. 199(F) electric arc evaporation ..........................40 elongation ............................. 126, 129(F) elongation vs. temperature ................ 126(F) emissivity, coating ......................... 181(T) emissivity factors ............................... 180 evaporation rate ..................................47 external deformed appearance after 500h testing of turbine blades .......... 16, 17(F) fatigue strength compared to that of thermal barrier coating ............................ 141 fatigue strength of coated JS6U superalloy ..........140(T), 141(F), 143(F) fatigue strength of coated VJL12U superalloy blades ............................ 140, 142(T) fatigue strength of coated VLJ12U superalloy ....................... 140, 143(F) heat capacity ................................ 181(T) heat treatment effect on beta-phase quality ....................................77(T) high-temperature corrosion ................. 11(F) intensities of strains and stresses on coated MAR-M-002 blade ................... 202(T) on JS6U superalloy, heat resistance testing with oxide structural change ........................... 109, 110(F) low-cycle fatigue ........149(F), 150(T), 151(F) low-cycle fatigue testing ......... 149–150(F,T) microball peening .............................. 131 microstructure ................................ 75(F) microstructure and distribution of elements in ingot bath after 20 deposition processes ........................... 47, 48(F) multilayer coating system thermal stresses ................................. 134(F) no thermal fatigue cracking ........... 20, 21(F) number of cycles to coating failure ..... 201(F) oxidation depth of turbine blades .......... 16(F) phase transformations, interaction mechanism between overlay coating and superalloy at high temperatures ................. 88, 92(T) strain intensity .............................. 201(F) strain-intensity range ...................... 201(F) strain intensity vs. temperature ........... 199(F) stress intensity .............................. 201(F) surface preparation effect on adhesion strength deposited by electric arc technique ..46(T) temperature .................................. 181(T)



226 / Index

Coatings (continued) tensile stresses .................................. 130 thermal conductivity ....................... 181(T) thermal diffusivity ......................... 181(T) thermal shock testing of coating/superalloy systems ................................ 148(T) thermal stress distribution on JS6U superalloy blades ............................ 131, 132(F) thermal stresses ....................130(F), 131(F) thermal stresses on JS6F superalloy at different orientations ...................... 133(F), 134 thermal stresses on JS6U superalloy ....................... 132(F), 133 thermal stresses variations in coating and JS6U superalloy ...................... 133(F) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) threshold recrystallization temperature ..... 131 ultimate strength ................127–128, 129(F) ultimate strength vs. temperature ........ 126(F) weight change of overlay coating on JS6U superalloy (isothermal test) ......... 110(F) working temperature 1100 to 1150C, protection capability reduction ...........16 x-ray pattern of overlay coating surface ... 76(F) Ni20Cr12AlY/Al, thermofatigue testing of coating/superalloy systems .. 147(T), 148(F) Ni20Cr12AlY/chromium carbide coating life, hot corrosion test in GZT ash .................................... 115(F,T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) thickness of coatings on JS6U superalloy ............................. 109(T) Ni20Cr12AlY/Ni10Co20Cr12AlY low-cycle fatigue ................. 149(F), 150(T) thermal shock testing of coating/superalloy systems ................................ 148(T) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) Ni20Cr12AlY/ZrO2-8%Y2O3, as thermal barrier coating for CMSX-4 single-crystal alloy ........................................... 207 Ni20CrTi-Al system, diffusion layer .... 56, 57(F) Ni21Cr9AlY chemical composition .............65(T), 212(T) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 122(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 212(T) oxide life at 900C ......................... 102(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) weight gain (cyclic heat resistance tested) .................................. 102(T) weight gain (isothermal heat resistance tested) .................................. 102(T) weight loss due to hot corrosion ......... 110(T) Ni22Cr6AlY

elongation ................................... 126(T) ultimate strength ............................ 126(T) Ni22Cr11Al1Hf chemical composition ..................... 102(T) corrosion depth of hot corrosion in GZT ash ...................................... 111(F) cyclic testing, hafnium and silicon effects .................................. 102(F) weight gain (cyclic heat resistance tested) .................................. 102(T) weight gain (isothermal heat resistance tested) .................................. 102(T) weight loss due to hot corrosion ......... 110(T) Ni22Cr11Al2Hf chemical composition ..................... 102(T) corrosion depth of hot corrosion in GZT ash ...................................... 111(F) cyclic testing, hafnium and silicon effects .................................. 102(F) weight gain (cyclic heat resistance tested) .................................. 102(T) weight gain (isothermal heat resistance tested) .................................. 102(T) weight loss due to hot corrosion ......... 110(T) Ni22Cr11AlSi corrosion depth of hot corrosion in GZT ash ...................................... 111(F) weight loss due to hot corrosion ......... 110(T) Ni22Cr11Al1Si chemical composition ..................... 102(T) corrosion depth of hot corrosion in GZT ash ...................................... 111(F) cyclic testing, hafnium and silicon effects .................................. 102(F) weight loss due to hot corrosion ......... 110(T) Ni22Cr11Al2Si chemical composition ..................... 102(T) cyclic testing, hafnium and silicon effects .................................. 102(F) Ni22Cr11AlY chemical composition .............65(T), 212(T) corrosion depth in GZT ash .............. 110(F) crystallization intervals ......................49(T) cyclic testing, hafnium and silicon effects .................................. 102(F) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 122(T) elongation ................................... 126(T) liquidus-solidus temperatures ..............49(T) microstructure ................................ 66(F) normalized number of variable ........... 212(T) phase composition ...........................65(T) thermal expansion coefficients, various temperatures .......................... 121(T) thermal expansion coefficients vs. temperature ......................... 67(F), 70 ultimate strength ...................... 124, 126(T) weight gain (cyclic heat resistance tested) .................................. 102(T)



Index / 227

weight gain (isothermal heat resistance tested) .................................. 102(T) weight loss due to hot corrosion ......... 110(T) Ni22Cr12AlY, forces acting in coating on JS6U superalloy versus coating thickness ............................ 132(F), 133 Ni28Cr6AlY coating life ...................................... 116 coating life (hot corrosion test in GZT ash) ................................... 115(F,T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) thickness of coatings on JS6U superalloy ............................. 109(T) Ni28Cr11AlY chemical composition .......................65(T) corrosion depth in GZT ash .............. 110(F) hot corrosion resistance ....................... 111 oxide life at 900C ......................... 102(T) phase composition ...........................65(T) weight gain (cyclic heat resistance tested) .................................. 102(T) weight gain (isothermal heat resistance tested ................................... 102(T) weight loss due to hot corrosion ......... 110(T) Ni36Cr5AlY, chemical composition .........10(T) Ni40Cr10AlY, phase transformations, interaction mechanism with superalloy at high temperatures ..................... 90–91, 93(F) Coatings, specific types, Nickel-chromiummolybdenum-tungsten-cobalt alloys Ni10CrMoWCo, oxidation constant ........ 108(T) Coatings, specific types, Nickel-chromiumtungsten-molybdenum-cobalt alloys Ni10CrWMoCo, fatigue strength of coated JS6U superalloy ......................... 139(F) Coatings, specific types, Nickel-cobalt-chromiumaluminum alloys Ni6Co10Cr13AlTaReHf, thermal expansion coefficients, various temperatures .... 123(T) Ni6Co20Cr12AlY, thermal expansion coefficients, various temperatures .... 123(T) Ni6Co22Cr11AlY, weight loss (hot corrosion test in GZT ash) ......................... 113(T) Ni7Co12Cr17Al alloying element effects at high temperatures .............................. 116 chemical composition and other characteristics ......................... 153(T) coating service life ............................. 157 cyclic life, thermomechanical fatigue testing ..........................155, 157(F,T) thickness of coatings on JS6U superalloy ............................. 109(T) Ni7Co12Cr17Al(Ni8Co12Cr7Al/Al5Si1.5Y) coating life (hot corrosion test in GZT ash) ..................................... 115(T) electric-arc technique, microstructure ..... 75(F) Ni8Co12Cr7Al chemical composition and other characteristics ......................... 153(T)

coatings life (hot corrosion test in GZT ash) ..................................... 115(T) coating service life ............................. 157 cyclic life, thermomechanical fatigue testing ..........................155, 157(F,T) Ni8Co20Cr12AlY chemical composition .......................10(T) cyclic life ........................................ 202 density ....................................... 181(T) heat capacity ................................ 181(T) multilayer coating system thermal stresses ................................. 134(F) number of cycles to coating failure ..... 201(F) strain intensity .............................. 201(F) strain-intensity range ...................... 201(F) stress intensity .............................. 201(F) temperature .................................. 181(T) thermal conductivity ....................... 181(T) thermal diffusivity ......................... 181(T) Ni10Co15Cr8Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T) crystallization intervals ......................49(T) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1100C) ........................ 104(F) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni10Co15Cr8AlY chemical composition .............68(T), 211(T) ductility .......................................... 128 elongation, various temperature .......... 127(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) phase transformations after high-temperature testing ........................................71 thermal expansion coefficients vs. temperature .............................. 72(F) ultimate strength, various temperatures .......................... 127(T) weight change (heat resistance isothermal test at 1100C) ........................ 104(F) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) yield strength, various temperatures ..... 127(T) Ni10Co15Cr14Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T)



228 / Index

Coatings (continued) crystallization intervals ......................49(T) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1100C) ........................ 104(F) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni10Co15Cr14AlY chemical composition .............68(T), 211(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients vs. temperature .............................. 72(F) weight change (heat resistance isothermal test at 1200C) ............................. 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni10Co20Cr12AlY elongation ................................... 129(F) elongation, various temperatures ........ 127(T) elongation vs. temperature ................ 126(F) on JS6U superalloy, heat resistance testing with oxide structural change 109, 110(F) ultimate strength ................127–128, 129(F) ultimate strength, various temperatures .......................... 127(T) ultimate strength vs. temperature ........ 126(F) yield strength, various temperatures ..... 127(T) Ni10Co22Cr11Al elasticity modulus values, various temperatures .......................... 124(T) thermal expansion coefficients, various temperatures .......................... 123(T) Ni10Co22Cr11AlY chemical composition .............68(T), 211(T) crystallization intervals ......................49(T) elongation and ductility ....................... 127 elongation, various temperatures ........ 127(T) liquidus-solidus temperatures ..............49(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) ultimate strength, various temperatures .......................... 127(T) weight change (heat resistance isothermal test at 1200C) ........................ 105(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) yield strength, various temperatures ..... 127(T) Ni10Co30Cr8Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T)

crystallization intervals ......................49(T) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1100C) ........................ 104(F) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni10Co30Cr8AlY chemical composition .............68(T), 211(T) elongation, various temperatures ........ 127(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients vs. temperature .............................. 72(F) ultimate strength, various temperatures .......................... 127(T) weight loss (hot corrosion in GZT ash) ..................................... 112(T) yield strength, various temperatures ..... 127(T) Ni10Co30Cr14Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni10Co30Cr14AlY chemical composition .............68(T), 211(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients vs. temperature .............................. 72(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni18Co12Cr7Al, thickness of coatings on JS6U superalloy ......................... 109(T) Ni20Co15Cr11Al elasticity modulus values, various temperatures .......................... 124(T) thermal expansion coefficients, various temperatures .......................... 123(T) Ni20Co15Cr11AlY



Index / 229

chemical composition .............68(T), 211(T) crystallization intervals ......................49(T) liquidus-solidus temperatures ..............49(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) weight change (heat resistance isothermal test at 1200C) ........................ 105(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni20Co20CrAlY, cobalt effect on oxidation .. 103 Ni20Co20Cr5AlY in bond coat of thermal barrier coating ..... 187 cyclic testing of this inner layer alloy ..................................185–186 Ni20Co20Cr11AlY coating life, hot corrosion test in GZT ash .................................... 115(F,T) coating life on JS6U superalloys ........ 116(T) crystallization intervals ......................49(T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) liquidus-solidus temperatures ..............49(T) thickness of coatings on JS6U superalloy ............................. 109(T) Ni20Co20Cr11AlY/ZrO2-15%Y2O3 coating life (hot corrosion test in GZT ash) ................................... 115(F,T) coating life on JS6U superalloys ........ 116(T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) oxide life on JS6U superalloy, cyclic testing .................................. 108(T) thickness of coatings on JS6U superalloy ............................. 109(T) Ni20Co20Cr12Al, oxidation constant affected by alloying elements ..................101–102 Ni20Co20Cr12AlY alloying element content variation according to melting techniques ..................44(T) bond coat deformation ..................183–184 in bond coat of thermal barrier coating ..... 187 calculated number of cycles to crack appearance ............................ 203(T) chemical composition .......................10(T) creep-rupture strength ..... 134, 135(F), 136(F) creep-rupture strength in air of coated JS26-DS superalloy .................136–137, 138(F) cyclic life ........................................ 202 cyclic testing and fragmentation ...... 20, 21(F) cyclic testing of this outer layer alloy ..................................185–186 density ....................................... 181(T) elongation .......................127–128, 129(F) elongation of JS6U superalloy ..... 136, 137(F) elongation, various temperatures ........ 127(T) elongation vs. temperature ................ 126(F) evaporation rate ..................................47 fatigue strength of coated JS6U superalloy ............................. 140(T)

gas contents in ingots subjected to extra deoxidation with carbon and cerium ....................................44(T) gas contents in various zones of ingot without extra deoxidation .......................44(T) heat capacity ................................ 181(T) interaction zone of overlay coating with superalloy with high-temperature testing ............................... 88, 91(F) microstructure ................................ 75(F) microstructure after high-temperature testing for stress-rupture strength ........... 138(F) minimum calculated cyclic life .............. 203 number of cycles to coating failure ..... 201(F) strain intensity .............................. 201(F) strain-intensity range ...................... 201(F) stress intensity .............................. 201(F) temperature .................................. 181(T) thermal conductivity ....................... 181(T) thermal diffusivity ......................... 181(T) thermal shock testing of coating/superalloy systems ................................ 148(T) thermal stresses ....................130(F), 131(F) thermal stresses on JS6F superalloy at different orientation ....................... 133(F), 134 thermal stresses on JS6U superalloy ....................... 132(F), 133 thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) thermomechanical fatigue testing, hysteresis loop calculation ................. 152(F), 153 ultimate strength ................127–128, 129(F) ultimate strength, various temperatures .......................... 127(T) ultimate strength vs. temperature ........ 126(F) yield strength, various temperatures ..... 127(T) Ni20Co20Cr13AlY, phase transformations, interaction mechanism with superallloy at high temperatures ................. 91, 93(F) Ni20Co22Cr8Al elasticity modulus values, various temperatures .......................... 124(T) thermal expansion coefficients, various temperatures .......................... 123(T) Ni20Co22Cr8AlY chemical composition .............68(T), 211(T) elongation, various temperatures ........ 127(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) ultimate strength, various temperatures .......................... 127(T) weight change (heat resistance isothermal test at 1200C) ........................ 105(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) yield strength, various temperatures ..... 127(T) Ni20Co22Cr11Al chemical composition .............68(T), 211(T) elasticity modulus values, various temperatures .......................... 124(T)



230 / Index

Coatings (continued) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni20Co22Cr11AlHfSi chemical composition ..................... 102(T) oxide life .................................... 103(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight gain (heat resistance cyclic test) ..................................... 103(T) weight gain (heat resistance isothermal test) ..................................... 103(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Ni20Co22Cr11AlHf2Si chemical composition ..................... 102(T) oxide life .................................... 103(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight gain (heat resistance cyclic test) ..................................... 103(T) weight gain (heat resistance isothermal test) ..................................... 103(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Ni20Co22Cr11Al1Hf1Si chemical composition ..................... 102(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Ni20Co22Cr11Al2HfSi chemical composition ..................... 102(T) oxide life .................................... 103(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight gain (heat resistance cyclic test) ..................................... 103(T) weight gain (heat resistance isothermal test) ..................................... 103(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Ni20Co22Cr11Al2Hf2Si chemical composition ..................... 102(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Ni20Co22Cr11AlY chemical composition .............68(T), 211(T) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 124(T) elongation, various temperatures ........ 127(T) normalized number of variable ........... 211(T) oxide life .................................... 103(T) phase composition ...........................68(T)

thermal expansion coefficients, various temperatures .......................... 123(T) ultimate strength, various temperatures .......................... 127(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight change (heat resistance isothermal test at 1200C) ........................ 105(F) weight gain (heat resistance cyclic test) ..................................... 103(T) weight gain (heat resistance isothermal test) ..................................... 103(T) weight loss (hot corrosion in GZT ash) ..................................... 112(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) yield strength, various temperatures ..... 127(T) Ni20Co22Cr11AlY2HfSi chemical composition ..................... 102(T) oxide life .................................... 103(T) weight change (heat resistance cyclic test) ..................................... 105(F) weight gain (heat resistance cyclic test) ..................................... 103(T) weight gain (heat resistance isothermal test) ..................................... 103(T) weight loss (hot corrosion test in GZT ash) ..................................... 113(T) Ni20Co22Cr14Al, thermal expansion coefficients, various temperatures .................... 123(T) Ni20Co22Cr14AlY chemical composition .............68(T), 211(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) weight change (heat resistance isothermal test at 1200C) ........................ 105(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni20Co30Cr11Al, thermal expansion coefficients, various temperatures .................... 123(T) Ni20Co30Cr11AlY chemical composition .............68(T), 211(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) weight change (heat resistance isothermal test at 1200C) ........................ 105(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni22Co22Cr11AlY, weight change (heat resistance isothermal test at 1200C) ................................... 105(F) Ni22Co22Cr12AlY, microstructure .......... 74(F) Ni23Co28Cr9AlY coating life ...................................... 116 coating life, hot corrosion test in GZT ash .................................... 115(F,T) life coefficient (hot corrosion test in GZT ash) ..................................... 116(T) thickness of coatings on JS6U superalloy ............................. 109(T)



Index / 231

Ni30Co15Cr8Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T) crystallization intervals ......................49(T) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1100C) ........................ 104(F) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni30C015Cr8AlY chemical composition .............68(T), 211(T) ductility .......................................... 128 elongation, various temperatures ........ 127(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients vs. temperature .............................. 72(F) ultimate strength, various temperatures .......................... 127(T) weight loss (hot corrosion in GZT ash) ..................................... 112(T) yield strength, various temperatures ..... 127(T) Ni30Co15Cr14Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T) crystallization intervals ......................49(T) ductility during bend testing .............. 129(F) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1100C) ........................ 104(F) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni30Co15Cr14AlY chemical composition .............68(T), 211(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermalized expansion coefficients vs. temperature .............................. 72(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T)

Ni30Co22Cr11Al elasticity modulus values, various temperatures .......................... 124(T) thermal expansion coefficients, various temperatures .......................... 123(T) Ni30Co22Cr11AlY chemical composition .............68(T), 211(T) crystallization intervals ......................49(T) liquidus-solidus temperatures ..............49(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) weight change (heat resistance isothermal test at 1200C) ............................. 105(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni30Co30Cr8Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T) crystallization intervals ......................49(T) elasticity modulus values, various temperatures .......................... 124(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni30Co30Cr8AlY chemical composition .............68(T), 211(T) microstructure ................................ 69(F) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients vs. temperature .............................. 72(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni30Co30Cr14Al chemical composition .............68(T), 211(T) chemical composition of structural constituents ..............................69(T) crystallization intervals ......................49(T) liquidus-solidus temperatures ..............49(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) thermal expansion coefficients, various temperatures .......................... 123(T) weight change (heat resistance isothermal test at 1200C) ........................ 104(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni30Co30Cr14AlY chemical composition .............68(T), 211(T) normalized number of variable ........... 211(T) phase composition ...........................68(T) sigma-phase occurrence typical ................71



232 / Index

Coatings (continued) thermal expansion coefficients vs. temperature .............................. 72(F) weight loss (hot corrosion in GZT ash) ..................................... 112(T) Ni37Co15Cr11AlY, phase transformations, interaction mechanism with superalloy at high temperatures ............... 91–92, 94(F) Coatings, specific types, Thermal barrier coatings RT31/RT33 characteristics of coating ......184(T), 187–188 growth parameter ........................... 185(T) service life ................................... 185(T) thermally grown oxide thickness ........ 185(T) TBC, thermofatigue testing of coating/superalloy systems ......................... 147(T), 148(F) TBC-4 characteristics of coating ............ 184(T), 187 growth parameter ........................... 185(T) service life ................................... 185(T) thermally grown oxide thickness ........ 185(T) ZrO2-8%Y2O3 /Ni20Cr12AlY, ceramic-layer cohesive ultimate strength ....... 204–205(F) ZrO2-8%Y2O3 /Ni20Co20Cr11/6Al0.25Y, service lives of samples ................... 183(F), 187 ZrO2-8%Y2O3 /Ni23Co20Cr11/5Al0.26Y, service lives ....................................... 182(F) Coating strain ................................. 196, 197 Coating-thermal-stresses modeling ... 132–134(F) Coating thermomechanical fatigue characteristic .................................. 199 Coating thickness/specimen cross section ratio .............................................. 119 Coating thickness value ........................... 197 Cobalt addition effect on saturating mixtures for aluminizing .....................................27 content effect on beta-phase-depleted outer zone and oxidation ....................104–105 content effect on elongation and ductility ..... 127 content effect on hot corrosion resistance of Ni-Co-Cr-Al alloys ......................... 112 content effect on overlay coatings with hightemperature testing .. 91–93(F), 94(F), 95(T) content effect on thermal expansion coefficient of Ni-Co-Cr-Al alloys ................121–122 content effect on thickness of interaction zone of overlay coating and superalloy 93, 95(F) content growth during high-temperature testing of diffused coatings ..............................89 content increase in coating effects on coating/ alloy system ............. 92–93, 94(F), 95(T) effect on heat resistance and oxide-layer formation and growth ...................... 103 Cobalt-nickel-chromium-aluminum alloys. See also Coatings, specific types, cobalt-nickelchromium-aluminum alloys ..................... 3 chemical compositions .......... 71, 72(T), 212(T) elasticity modulus values, various temperatures ....................... 124, 125(T) elongation, various temperatures ........... 128(T)

heat resistance testing ..................105–106(T) hot corrosion resistance ................112–113(T) magnetic transformation .................. 71, 74(F) microstructures ............................. 71, 73(F) normalized number of variable .............. 212(T) as overlay coatings, and high-temperature testing ..........................91–93(F), 94(F) phase compositions ........................ 71, 72(T) phase compositions and structure changes .....87, 91–93(F), 94(F), 95(T) phase transformations ..............................71 regression equations for property values ...................................214–215 thermal expansion coefficients, various temperatures ....................... 124, 125(T) ultimate strength ............................... 127(T) variability of values ............................... 214 yield strength .................................. 127(T) Cobalt-nickel-chromium-aluminum-yttrium alloys, hot corrosion resistance in GZT ash. See also Coatings, specific types, cobalt-nickelchromium-aluminum alloys .............. 113(T) Coefficient of plasticity ............................ 196 Coefficients of transfer in flux for various materials, electric arc evaporation ........40(T) Colloxiline ......................................... 33, 34 Color, of nickel aluminide diffusion coatings ........................................56(T) Composition-property regression equations ..... 3 Composition-structure-property relationships .. 3 Compressive stresses ............................... 198 of ceramic coatings ............................... 184 in ceramic layer of thermal barrier coatings .................................206–207 Compressor pressure ratio .......................... 5 Concentration gradients .............................87 Condensation coatings. See Overlay coatings. Confidence coefficient .............................. 121 Copper cathode spot temperature and vapor pressure .....................................40(T) coefficient of transfer in flux, electric arc evaporation .................................40(T) as impurity of jet fuels .............................. 5 Corrosion penetration depth ..................... 100 Creep rates, of superalloys in air and GZT ash ............................................ 136(F) Creep-rupture strength of superalloys at high temperatures in air ..............................136–137, 138(F) of superalloys in media provoking hot corrosion ................134–136(F,T), 137(F) Creep strains, of thermal barrier coatings ....... 205 Cruising conditions .................................... 5 Crystalline structure, of nickel aluminide diffusion coatings ........................................56(T) Crystallization interval, regression equation for calculation of ................................... 213 Cyclic heat resistance testing ..................97–98 Cyclic service life calculation, of coating .........20



Index / 233

Cyclic temperature variations, and flaws in turbine engines .................................... 5 Cyclic testing, for heat resistance ...................97

D Databases ................................................ 3 DBTT. See Ductile-brittle transition temperature. Defects, causes in coatings ...........................10 Deformation patterns .............................. 151 Degreasing, to prepare blade surfaces for overlay coatings ........................ 45, 46–47 Degree of ionization, of electric arc evaporated products ........................................40(T) Delta-phase nickel aluminide (Ni2Al3) ............55 color ...............................................56(T) crystalline structure .............................56(T) density ............................................56(T) in diffusion coating formed in high-activity media .......56(T), 57(F), 58(F,T), 59, 60(T), 61(F), 62(F), 63 in diffusion coating formed in moderate and low-activity media ....................... 64, 67 hardness ..........................................56(T) heat of formation ................................56(T) homogeneity region .............................56(T) melting temperature .............................56(T) oxidation time ................................. 100(T) oxide structure ................................. 100(T) specific volume ..............................128–129 x-ray diffraction analysis ..................... 100(T) Density of evaporated substance ............................39 of nickel aluminide diffusion coatings .......56(T) of zone of coating (q) .........................17–18 Diffusion aluminide coatings ........................ 1 annealing of powder mixture ......................30 application ............................................. 8 application by electric arc technique under vacuum ....................................35–37 application in circulating gaseous phase ................................... 31–32(F) application in powder mixtures with halide activators .............................. 25–31(F) application using slurry ................... 32–35(F) beta-phase nickel aluminide aluminum content ........................................ 145 blade service conditions limitation ................ 2 changes in phase size ...................... 64(F), 66 characteristics dependent on temperature, time of annealing, and specific mass of aluminum layer .............................................61 chemical compositions .................... 10, 11(T) composition ranges .................................30 distribution patterns of alloying elements ..67–68 ductile-brittle transition temperature ....... 120(F) electric arc deposition below 640C ..............55 fatigue strength of coated superalloys ......................... 139–140(F)

filling saturation mixture, preventing outpour ..31 formation at temperatures above 640C ............................... 59–63(F,T) formation in low-activity media ...... 63–69(F,T) formation in moderate-activity media ................................ 63–69(F,T) formed in high-activity media ......... 55–63(F,T) fragmentation vs. time of testing ........ 20, 21(F) hot corrosion resistance .......................... 136 inspection of ....................................37–38 interaction mechanism with superalloys ............................ 85–86(F) intermetallic compound zones grow as in aluminizing .....................................80 kinetics of coating growth on nickel-base superalloys ............................ 27, 28(F) life expectancy increased with blade surface temperature ............................ 15–19(F) and low-cycle fatigue of superalloys ...................... 149–150(F,T) mass of deposited Al5Si1.5Y alloy vs. coating thickness .................................... 61(F) microstructure ........................... 55–63(F,T) microstructure changes during high-temperature tests .................................. 79–86(F,T) modified with silicon or yttrium ................... 2 on nickel, phase transformations .............. 80(F) phase composition ...................... 55–63(F,T) phase composition changes during hightemperature tests .................. 79–86(F,T) phase composition, factors determining .........55 preparation of blade inner cavity for coating ...31 specific mass of deposited layer ..................61 stages of technological process ...................30 on superalloys, heat resistance of coating/ superalloy system .............. 106–108(F,T) thickness dependence on temperature and time of aluminizing process ..................66–67 thickness related to applied slurry thickness ....34 thickness related to specific mass of aluminum alloy deposited .................................62 zones of successively arranged aluminides .....61 Diffusion annealing after aluminizing ....................................29 after diffusion coating application using slurries ........................32, 33–34, 35(F) temperature effect on hardness in slurry technique .............................. 34(F), 35 Diffusion layer ........................55, 56(T), 57(F) formation below 640C ...........55, 56(T), 57(F) Dilatometric analysis of cobalt-nickel-chromium-aluminum alloys ............................................71 to examine phase transformations of Ni-Cr-Al alloys ............................................70 Dimethyloxolate .................................. 33, 34 Directionally solidified superalloys, creep-rupture strength ..........................136–137, 138(F) Disproportionation reactions ..................25–26 Disulfides ................................................ 5



234 / Index

Dopants ......................................... 179, 180 erbium, for zirconium oxide crystal lattice .... 179 gadolinium as ...................................... 179 neodymium as ...............................179–180 nickel oxide as ............................... 179, 180 ytterbium, for zirconium oxide crystal lattice ....................................179–180 Ductile-brittle transition temperature (DBTT) of diffusion coatings .......................... 120(F) range ................................................ 198 Ductility, vs. temperature ....................... 199(F) Duga-90 unit ........................... 42, 43(F,T), 44

E EA. See Electric arc technique. EB. See Electron beam evaporation technique. EBR. See Electron beam remelting. Efficiency ................................................ 5 Elasomat instrument ............................... 121 Elastic deformation ................................. 151 Elasticity modulus of coating ........................................... 196 of Co-Ni-Cr-Al alloys .................. 124, 125(T) of Ni-Cr-Al alloys ............................. 122(T) of Ni-Co-Cr-Al alloys ........ 123, 124(T), 125(F) and thermal expansion coefficients of coating alloys ................................120–124(T) 20–800C, regression equation for calculation of ..................................213, 215, 216 Elastic-plastic strain range in coating under nonisothermal loading ....................... 200 Electric arc current (Iea) .............................40 Electric-arc (EA) evaporation technique ......... 2 to apply overlay coatings .................39–41(T) cycle stages for coating deposition ...............50 forming overlay coatings showing phase transformations followed by volume effects .......................................77(T) parameters controlled by units of deposition ...52 requirements on preparation of blade surfaces ................................46–47(T) units for coating deposition ............ 42–44(F,T) Electric-arc method ................................... 2 Electric arc (EA) technique for applying diffusion coatings under vacuum ....................................35–37 for depositing diffusion coatings, phase transformations after high-temperature testing ....................................... 84(F) deposition of bond coats ......................... 167 diffusion coatings above 640C ....... 59–63(F,T) diffusion coatings below 640C ....... 55–59(F,T) microstructure of diffusion coating ..........75–76 Electric arc vacuum deposition, as coating method for diffusion coatings ........................11(T) Electron beam (EB) evaporation technique ...... 2 applied to ceramic layer of thermal barrier coatings .........................................21

to apply overlay coatings .....................38–39 to apply overlay coatings, phase transformations ............................ 85(F) boundary temperatures at which condensate structure undergoes changes ........ 47–49(F) ceramic pellets for ................... 162–165(F,T) condensate structures deposited on substrates having different temperatures ........... 74(F) cycle stages for coating deposition ... 47–50(F,T) depositing condensates for TEC studies ....... 120 evaporation of special ceramics .................... 2 forming overlay coating showing phase transformations followed by volume effects .......................................77(T) forming overlay coating subject to heat treatment operation ...............................76–77(T) for manufacturing aircraft engine turbine blades ......................................... 162 parameters controlled by units of deposition .................................51–52 strict requirements to surface preparedness .......................... 45–46(F) and thermal shock testing of coating/superalloy systems .......................... 148–149(F,T) units for coating deposition ............ 41–42(F,T) Electron beam remelting (EBR), to melt alloys for coatings deposited by the EB technique ......44 Electron beam thermal barrier coatings ............................ 161–190(F,T) Elongation of Co-Ni-Cr-Al alloys ........................ 128(T) of Ni-Cr-Al alloys ........................... 126(F,T) of Ni-Co-Cr-Al alloys .............. 127(T), 129(F) Emissivity factor .......................... 180, 181(T) Engine-equivalent quick tests ......................10 Engine testing ..........................................10 Epsilon-phase nickel aluminide (NiAl3) ..........55 color ...............................................56(T) crystalline structure .............................56(T) density ............................................56(T) in diffusion coating formed in high-activity media ............... 56(T), 57(F), 58(F,T), 59, 60(T), 61(F), 62(F,T), 63 in diffusion coating formed in moderate and low-activity media ........................56(T) hardness ..........................................56(T) heat of formation ................................56(T) homogeneity region .............................56(T) melting point of compound ........................59 melting temperature .............................56(T) specific volume ..............................128–129 Equilibrium, between phases of a coating and a superalloy .........................................86 Equilibrium pressure of vapor of substance evaporated ........................................39 Equivalent quick tests ...............................10 Erbium, dopants for ZrO2 crystal lattice ......... 179 Erosion ................................................... 8 Erosion on gas flow attack .................. 183, 184 Error for model using elastic strains ........... 197



Index / 235

ES 1/3/60 (Leybold) unit ......................... 177 Evaporation coefficient of electron beam evaporation ........39 temperature of .......................................39 time of ................................................39 Evaporation rate, electron beam evaporation ...................................38–39 Evaporator thermal efficiency .....................38

F Fatigue life, overlay coating effect ............... 140 Fatigue strength coatings effect on superalloys ......138–142(F,T), 143(F), 144(F) factors having effect on coated specimens .. 138– 139 superalloys with diffusion coatings .. 139–140(F) superalloys with overlay coatings ... 140–141(F,T), 142(F,T), 143(F), 144(F) FEM. See Finite-element method. Fighters, temperature and stress fields nonuniform ......................................... 5 Finite-element method (FEM) ...... 202, 203, 204, 205(F), 207 Firing temperature ................................. 165 Flash method ........................................ 177 Force in coating (P) ................................ 133 Fractography, to evaluate thermomechanical fatigue test specimens ................... 155, 157 Fragmentation ....................19–20(F), 198, 199 causes .................................................19 on ceramic layer of thermal barrier coating .....21 as initial stage of thermal barrier coating failure ......................................... 203 thermomechanical fatigue criterion consideration ................................. 202 Free aluminum, effect on thickness in high-activity mixture for aluminizing ................ 26–27(F)

G Gadolinium, as dopant ............................. 179 Gamma double-prime phase nickel aluminide (Ni3Al) .............................................. 2 aluminum content ............................. 119(T) chromium solubility in Ni-Cr-Al system alloys .......................................... 111 color ...............................................56(T) controlling interaction mechanism between overlay coatings and superalloys ......93–96 crystalline structure .............................56(T) density ............................................56(T) in diffusion coating formed in high-activity media ........................ 61(F), 62(F,T), 63

in diffusion coating formed in moderate and lowactivity media .......63(F), 64(F), 65–66, 67, 69(T) elongation ...................................... 119(T) formation beneath oxide layer in NiAl oxidation ...................................... 101 hardness ..........................................56(T) heat of formation ................................56(T) high-temperature strength ........................ 120 homogeneity region .............................56(T) melting temperature .............................56(T) of Ni-Cr-Al system ....................... 87, 88)(F) in overlay coatings at high-temperature tests ...87 oxidation time ................................. 100(T) oxide structure ................................. 100(T) phase transformations in diffusion coatings after high-temperature tests ............84(F), 85(F) in phase transformations of overlay coatings on superalloys ............89(F), 90(F), 91(F), 92(T), 95(T) specific volume ..............................128–129 two-way diffusion in diffusion coatings .... 80(F), 81(T), 82(F), 83(T), 84(F), 85(F) ultimate strength ............................... 119(T) x-ray diffraction analysis ..................... 100(T) yield strength .................................. 119(T) Gamma-phase .........................................98 in aluminum-silicon-yttrium diffusion coatings ..................................... 85(F) in cobalt-nickel-chromium-aluminum alloys ..........................71, 72(T), 87, 91 in nickel-chromium-aluminum alloys .. 65(T), 70, 76(F), 77, 87(F), 88(F), 89(F), 90, 91(F), 92(T), 93(F) in nickel-cobalt-chromium-aluminum alloys .... 68(T), 69(T), 70, 71, 87, 91, 94(F), 95(T) Gamma-prime phase .................................97 in aluminum-silicon-yttrium diffusion coatings ..................................... 85(F) in nickel-chromium-aluminum alloys .. 65(T), 70, 76(F), 77, 88(F), 89(F), 91, 92(T), 93(F) in nickel-chromium-molybdenum-tungstencobalt alloys ....................................82 in Ni-Co-Cr-Al alloys ...68(T), 69(T), 70, 71, 91, 94(F), 95(F) Gamma-solid solution ................. 70, 76, 77, 87, 88, 89, 93–96 controlling interaction mechanism between overlay coatings and superalloys ......93–96 formation, and phase transformations and internal oxidation ............................... 17, 18(F) in overlay coatings at high-temperature tests ...87 Gas aluminizing methods ............................ 2 Gas corrosion resistance ............................97 Gas dynamic forces ................................... 8 Gas-dynamic test rig, schematic diagram ... 145(F) Gas-dynamic test rigs and engines ............................. 189–190(F,T) Gaseous halides, in application technique of diffusion coatings ....................... 31–32(F)



236 / Index

Gas-flow test benches (gas rigs) ...................99 Gas temperature, at turbine entry ................... 5 Gas turbine fuel synthetic (GZT) ash ................................. 99–100, 111(F) composition ......................................99(T) creep-rupture strength of superalloys having hot corrosion ................134–136(F,T), 137(F) for hot corrosion testing of Ni-Cr-Al alloys ............................. 110–111(F,T) for hot corrosion testing of overlay coatings ...........109–110, 112(F,T), 113(T), 114(F), 115(F), 116(F) GZT ash. See Gas turbine fuel synthetic ash.

H Hafnium alloying addition effect on Ni-Co-Cr-Al alloys ....................................101–103 content effect on hot corrosion resistance of Ni-Co-Cr-Al alloys ......................... 112 content effect on hot corrosion resistance of overlay coatings ............................. 111 effect on heat resistance of Ni-Cr-Al alloys ................................ 101, 102(F) as replacement for yttrium, effect on oxides of Ni-Cr-Al alloys .................... 101, 102(T) substitution for yttrium effect on Ni-Co-Cr-Al cyclic heat resistance ....................... 105 Halide activators coating formation from powder mixtures .......37 compounds used in aluminizing ..................26 diffusion coatings in powder mixtures .. 25–31(F) Halides ................................................ 1, 2 amount at aluminizing ..............................27 Hall end-face plasma accelerator ............. 40, 41 “Hard” coatings, case for .......................... 198 Hardness, of nickel aluminide diffusion coatings ........................................56(T) Hardness testing, of diffusion aluminide coatings ............................................38 Heat capacity, determination by pulse method .....................................175–176 Heat flow ................................................. 5 Heat of formation, of nickel aluminide diffusion coatings ........................................56(T) Heat resistance, metallographic determination ...98 Heat resistance, cyclic test 20↔900C 1000h, regression equation for calculation of .............................................. 215 5000h, regression equation for calculation of .............................................. 215 8000h, regression equation for calculation of .............................................. 215 Heat resistance, isothermal test at 1200C 100h, regression equation for calculation of .. 214 300h, regression equation for calculation of .. 214 Heat resistance parameter ........................ 107

Heat resistance testing ...................... 97–99(F) of aluminides ......................... 100–106(F,T) overlay coatings ...................... 100–106(F,T) Heterophase diffusion theory statement .........64 High-activity media description ............................................55 temperature range ...................................55 High-pressure (HP) turbines, nozzle vanes ....... 6 High-temperature corrosion ......................... 8 of overlay coatings .............................. 11(F) High temperatures blade working cross-sectional area reduced in hot zone ...................................... 195 effect on diffusion aluminide coatings, phase composition and microstructure ..................... 79–86(F,T) protection against .................................... 2 High-temperature strength .......................... 7 coating effects on superalloys ...... 134–138(F,T) High-temperature testing ............... 79–96(F,T) of overlay coatings, microstructural changes ............................. 87–96(F,T) of overlay coatings, phase composition changes ............................. 87–96(F,T) Homogeneity region, of nickel aluminide diffusion coatings ........................................56(T) Hot corrosion ................................ 1, 3, 8, 10 causing defects in coating .........................10 of coating/superalloy systems in GZT ash .......................134–136(F,T), 137(F) mechanism caused by aggressive compounds ..99 of Ni-Cr-Al alloys .................... 110–111(F,T) with salt addition to gas flow ................... 145 of superalloys, creep-rupture strength with GZT ash exposure ...........134–136(F,T), 137(F) temperature range 600 to 1200C resistance ..... 9 temperature regimes, depth of development ............................... 11(F) thermal barrier coatings for protection ...........20 of turbine blades ........................... 12–15(F) of ZrO2-8%Y2O3 ceramics ....................... 184 700C, 200h, regression equation for calculation of ................................ 215 800C, 200h, regression equation for calculation of ................................ 215 850C, 200h, regression equation for calculation of .......................... 214, 215 900C, 200h, regression equation for calculation of .......................... 214, 215 Hot corrosion resistance ........................ 10, 97 of coated superalloys ................ 113–116(F,T) of Co-Ni-Cr-Al alloys ..................112–113(T) effect of refractory elements, low solubility in aluminides ......................................63 of Ni-Cr-Al alloys .................... 110–111(F,T) of Ni-Co-Cr-Al alloys ................... 111–112(F,T), 113(T) of overlay coatings ................... 109–113(F,T) Hot corrosion resistance testing, 99–100(T) HP. See High-pressure turbines.



Index / 237

HT-1000 calorimeter (Setaram Company) .... 177 Hydroabrasive blasting, to prepare blade surfaces for overlay coatings ..................... 45(F), 46 Hydroabrasive processing ........................ 185 Hydrocarbon combustion ............................ 6 Hydrogen, as impurity of jet fuels ................... 5 Hydrogen gas ........................................... 6 Hydroxide method ............................ 164, 177 Hysteresis loops ........................... 152(F), 153

I Impurities industrial, as hazard ............................... 6, 9 of jet fuels ..........................................5–6 Induction transducer ............................... 121 Infrared pyrometers, in Duga-90 unit .............44 Inspection of diffusion aluminide coatings ...................37 of overlay coatings .............................51–52 Interaction zone thickness (d) ......................90 Intercrystalline defects ..................45(F), 46(F) Intercrystalline porosity ........................... 140 Ion transfer rate, of electric arc technique ........36 Iron addition effect on saturating mixtures for aluminizing ....................... 27–28(F), 29 in deposits detected on gas-turbine engine blade surfaces .........................................99 as impurity of jet fuels .............................. 5 Islets of lighter phase, at diffusion layer/aluminum interface .................................. 56, 57(F)

J Jet fuel combustion .................................... 6 product composition ................................. 5 product effect on turbine blade surface layers ... 5 temperature of ........................................ 6 Jet fuels, impurities .................................... 5 Jet fuel sulfur ........................................... 5

L Lamellar carbides precipitation ...................85 Laser flash method, to measure thermal conductivity of thermal barrier coating ceramic layer .......................... 175, 176(F) LA-TMF test. See Load-adjusted thermomechanical fatigue test. Lead, as impurity of jet fuels ......................... 5 Leaders ......................................45(F), 46(F) Life-expectancy calculations ............... 14–15(F)

Liquid penetrant inspection, of thermomechanical fatigue crack on coated superalloy ............................. 155, 157(F) Liquid-phase aluminizing method ................33 Liquidus temperature, regression equation for calculation of ................................... 213 Load-adjusted thermomechanical fatigue test (LA-TMF) ......... 150–151, 152–153(F), 156 for JS6U superalloy ........................... 200(F) Low-cycle fatigue, of superalloys, coating effect ......................... 149–150(F), 151(F) Low-cycle fatigue testing ....................... 149–150(F), 151(F) Low-pressure plasma spray (LPPS) method deposition of bond coats ......................... 167 for deposition of coatings on superalloys ..... 153 LPPS. See Low-pressure plasma spray method.

M Macrosegregation, in ingot with overlay coatings ........................................45(T) Magnesium, as impurity of jet fuels ................. 5 Manganese, as impurity of jet fuels ................. 5 Manson’s formula ..................... 200–201, 202 MAP unit ................................... 42–43(F,T) Margin of safety, reduction of ......................10 Mass, changes not related to coating life expectancies .......................................17 Material density (q) ..................................97 Material evaporated per second, electric arc evaporation technique ...........................40 Maximum admissible gas temperature ........... 7 Maximum cycle temperature ..................... 200 Maximum relative tensile thermal stress of thermal barrier coatings ................ 204(F) Mean mass temperature, of gas at turbine entry ................................................. 5 Mechanical mixing ....................... 163, 164(T) Melting temperature, of nickel aluminide diffusion coatings ............................56(T) Mercaptanes ............................................ 5 Metal heat resistant coatings .......................21 Metallographic examination, to determine coating thickness of diffusion aluminide coatings ....................................... 37, 38 Microball peening ........................ 130(F), 131 of as-deposited coating ........................... 115 of bond coats for thermal barrier coatings .... 166 of CMSX-4 superalloy with coating ........... 153 fatigue strength of overlay coatings affected 140 of overlay coatings ........................... 75, 140 of thermal barrier coatings ....................... 185 Microcracking .........................................21 Microdots, and low-cycle fatigue ................. 149 Microdroplet phase, in electric arc evaporation ........................................40



238 / Index

Microdrops ........................................... 140 on bond coat, caused by high yttrium contents of liquid bath .................................... 161 on bond coat, causing defects in TBC ceramic layer ........................................... 161 Microstructure, high-temperature test effects on diffusion aluminide coatings ....... 79–86(F,T) Minimum cycle temperature ..................... 200 Molecular mass of substance evaporated ........39 Molybdenum alloying NiAl with ................................ 101 solubility in gamma double-prime phase compound after high-temperature testing ...........................................82 Moment of the coating chemical and phase composition change .............................98 Multilayer coatings .................................... 9 Multiple correlation coefficient (Rf) corrected on the basis of degrees of freedom .............. 212 l-phase molybdenum-base as precipitate ..................82 in nickel-chromium-molybdenum-tungsten-cobalt superalloys ........................ 81(T), 82(F) two-way diffusion in diffusion coatings ............................ 81(T), 82(F)

N Neodymium, as dopant .......................179–180 “Neutral” coatings, case for ....................... 198 Nickel addition effect on saturating mixtures for aluminizing .....................................27 coefficient of transfer in flux, electric arc evaporation .................................40(T) content effect on heat resistance of Co-Ni-Cr-Al alloys .......................................... 106 content effect on thermal expansion coefficients of Co-Ni-Cr-Al alloys ...................... 124 degree of ionization .............................40(T) mass-spectrometric analysis data for material evaporation .................................40(T) Nickel aluminide diffusion coatings physical properties ......................... 55, 56(T) saturating media classification ........... 55, 56(T) NiAl2O4 spinel ..........................................83 formation with NiAl oxidation .................. 101 Nickel-chromium-aluminum alloys. See also Coatings, nickel-chromium-aluminum alloys ................................................ 3 bend testing ........................................ 128 chemical compositions .......... 65(T), 70, 212(T) elasticity modulus values, various temperatures ............................. 122(T) elongation, various temperatures ......... 126(F,T) heat resistance ........................... 101, 102(T) hot corrosion .......................... 110–111(F,T) microstructure ................................... 66(F) normalized number of variable .............. 212(T)

phase composition and structure changes ........................87–91(F), 92(T) phase compositions ........................ 65(T), 70 phase diagrams at high temperatures ......... 87(F) regression equations for property values ...... 216 thermal expansion coefficients, various temperatures ............................. 121(T) thermal expansion coefficients vs. temperature ............................ 67(F), 70 ultimate strength ............................. 126(F,T) variability of values ............................... 216 weight loss due to hot corrosion ............ 110(T) Nickel-chromium-molybdenum-tungsten-cobalt alloys .................................. 81(T), 82(F) Nickel-cobalt-chromium-aluminum alloys. See also Coatings, nickel-cobalt-chromiumaluminum alloys ................................... 3 alloying effect on specific mass loss of hot corrosion .................................. 112(F) alloying effect on thickness of beta-phasedepleted outer zone ............... 103, 105(F) bend testing ........................................ 128 beta-phase contents in alloy and coating structures .............................. 70, 72(F) chemical composition ........... 68(T), 70, 211(T) chemical composition of structural constituents ............................ 69(T), 70 elasticity modulus values, various temperatures ............. 123, 124(T), 125(F) elongation, various temperatures ........... 127(T) heat resistance testing ............... 101–105(F,T) hot corrosion resistance ... 111–112(F,T), 113(T) microstructures ............................. 69(F), 70 normalized number of variable .............. 211(T) as overlay coatings, and high-temperature testing ................. 91–93(F), 94(F), 95(T) phase composition and structure changes ...........87, 91–93(F), 94(F), 95(T) phase compositions ........................ 68(T), 70 phase transformations ..............................71 regression equations for property values ...................................213–214 strength levels of cast alloys lower than modern superalloys ................................... 128 thermal expansion coefficients ..... 122–124(F,T) thermal expansion coefficients, various temperatures ........................... 123(F,T) thermal expansion coefficients vs. temperature ............... 71, 72(F), 73, 74(F) ultimate strength, various temperatures .... 127(T) variability of values ............................... 213 weight gain (heat resistance cyclic testing) .................................... 103(T) weight gain (heat resistance, isothermal testing) .................................... 103(T) yield strength, various temperatures ................... 127(T), 129(F) Nickel-cobalt-chromium-aluminum-yttrium alloys, service life ............................ 15(F) Nickel oxide



Index / 239

as corrosion product penetrating the coating ....13 as dopants, effect on thermal conductivity ............................ 179, 180 formation with NiAl oxidation .................. 101 precipitate .................................... 178, 179 Nickel substrate, ductile-brittle transition temperatures for diffusion coatings 50lm thick .......................................... 120(F) Nickel superalloys, as turbine blade material ...... 1 Nitrogen, as impurity of jet fuels .................... 5 Nondestructive testing, to detect cracks in diffusion coatings ................................38 Normalized magnitudes of variables ..............................211–212(T) Nozzle guide vanes .................................... 6 Nozzle vanes of high-pressure (HP) turbine ...................... 6 temperature differences between vane back and trailing edges .................................... 7 thermal inhomogeneity .............................. 6 Number of cycles to coating failure at thermocyclic loading of blade .......... 201(F) Number of cycles to crack detection under given cyclic loading conditions .................... 199 Number of loading cycles ......................... 199

O Operating diagram of an airliner engine .. 5, 6(F) Oriented porosity fraction ........................ 178 Overlay coatings .................................2, 3, 9 application ............................................. 8 application techniques .................. 38–52(F,T) ceramic ............................................. 163 chemical composition of Ni-Cr-Al alloys with hafnium and silicon additions ......... 102(T) chemical compositions .................... 10(T), 74 cobalt-base, advantages .......................... 116 cobalt-nickel-chromium-aluminum ...... 71, 72(T) crystallization intervals .........................49(T) electric arc evaporation technique for application .............................39–41(T) electron beam evaporation technique for application .................................38–39 fatigue strength of coated superalloys ....... 140– 141(F,T), 142(F,T), 143(F), 144(F) fragmentation vs. time of testing ........ 20, 21(F) heat resistance testing ............... 100–106(F,T) heat treatment effect on beta-phase quantity ................................ 76, 77(T) heat treatment effect on phase transformations and volume effects ........................77(T) high-chromium or cobalt-base, orientation ......74 high-temperature corrosion .................... 11(F) high-temperature heat treatments ..... 76–77(F,T) hot corrosion resistance ....... 109–113(F,T), 136 inspection of ....................................51–52

interaction mechanism with superalloys ....93–96 interaction zones ....................................79 iron-base, hot corrosion resistance ............. 116 life coefficient values ......................... 116(T) life on turbine blades of JS6U superalloys ............................... 116(T) liquidus-solidus temperatures .................49(T) microstructure changes at high-temperature tests .................................. 87–96(F,T) microstructures ..................... 71, 74(F), 75(F) nickel-cobalt-chromium-aluminum alloys ........................ 69(F,T), 70–71(F) peening of finished coatings ......... 50–51(T), 52 phase composition changes at high-temperature tests .................................. 87–96(F,T) phase composition, factors determining .........55 phase compositions ..................... 69–77(F,T) preparing blade surfaces for coating 45–47(F,T) producing ingots and cathodes for coatings ................................44–45(T) recrystallization .................................... 130 recrystallization annealing .........................51 strain-hardening processing ........................51 structural constitution, factors determining .....55 substrate temperature effect on stability and hardness ........................................75 on superalloys, heat resistance of coating/ superalloy system .............. 108–109(F,T) thermomechanical fatigue testing .. 150–158(F,T) thickness of condensed substance ................39 x-ray structural pattern ................ 76(F), 77(T) Oxidation ............................................. 1, 8 blade damage ........................................17 causing defects in coating .........................10 during cooling in aluminizing process ...........29 effect on creep-rupture strength of superalloys ................................... 137 penetration depth ....................................98 resistance .....................................1, 10, 97 steady-state ...........................................98 of superalloys at testing and surface cracks ... 134 temperature range 600 to 1200C resistance ..... 9 temperature regimes, depth of development ............................... 11(F) thermal barrier coatings for protection ...........20 turbine blade operating at temperatures above 1000C ......................... 15–19(F) Oxidation activation energy (Q) ................. 107 Oxidation constant (Kp) ........................ 97, 98 for diffusion aluminized coated superalloys ............................... 108(T) Oxide lifetime ...................................... 98(F) Oxide protective life .............................. 98(F) hafnium addition effect on nickel-chromiumaluminum alloys ............................. 101 of overlay coatings on JS6U superalloys .. 108(T) Oxides, performing protective function on blades ..................................... 17, 18(F) Oxide spalling .........................................98



240 / Index

Oxygen ................................................... 6 as impurity of jet fuels .............................. 5

P Pack calorizing as coating method for diffusion coatings ....11(T) under vacuum, as coating method for diffusion coatings .....................................11(T) Palladium, as alloying element for diffusion coating ............................. 115(T), 116(T) Parabola exponent (n) .......................... 97, 98 Passenger aircraft engines, temperature and stress fields ................................................ 5 Peening. See also Microball peening. of Ni-Cr-Al and Ni-Co-Cr-Al overlay coatings ..................................... 75(F) overlay coatings ....................... 50–51(T), 52 Phase compositions determined from coating thickness ...............80 high-temperature test effects on diffusion aluminide coatings ................ 79–86(F,T) Phase transformations, in diffusion coating on nickel ........................................... 80(F) Phonon scattering ................................... 179 Pillars ..................................................... 8 Pit corrosion .................................. 12, 13(F) Plane-strain-state models, generalized ..........................202, 203, 207 Plasma flux ionization ...............................40 Plasma method ......................................... 2 Plasma transfer ........................................ 2 Plasticity, coefficient of ............................ 196 Poisson’s ratio for the coating in case of its elastic straining ................................ 196 Porosity, as result of diffusion processes ..........79 Potassium, in deposits detected on gas-turbine engine blade surfaces ............................99 Powder metallurgy ..............................1, 2, 8 aluminide coating application drawbacks .......30 aluminide coating characteristics .................30 diffusion coating applications with halide activators .............................. 25–31(F) Power of the electron beam (Pe) ...................38 Preoxidizing, of bond coat of thermal barrier coatings .......................................... 167 Pulse method ......................................... 177 to determine thermal diffusivity and heat capacity .................................175–176

Q Quantimet instrument, to determine NiAl phase amount in coating structure .....................70 Quartz resonators, in Duga-90 unit ................44

R Radiation heat exchange .......................... 176 Recrystallization .................................... 131 in overlay coatings ................................ 130 Recrystallization annealing overlay coatings .....................................51 of thermal barrier coatings ....................... 185 Reduced modulus of elasticity in the strain theory of plasticity ............................ 197 Refractory elements content growth during high-temperature testing of diffused coatings ...........................89 diffusive mobility ...................................68 low solubility in aluminides providing hot corrosion resistance to diffusion coatings .........................................63 not contained in NiAl compound of diffusion coating in low-activity powder mixtures ............................... 65(T), 67 phase precipitation of diffusion coatings on base of ...........................................66–67 Regression equations .........................212–216 for Ni-Co-Cr-Al alloys, specific weight measurements ................................ 103 Relative elongation, 700–1100C, regression equation for calculation of .................... 215 Reproducibility, of aluminizing coatings ..........29 Residual compressive stress, in aluminide coatings .......................................... 129 Residual sum of squares .......................... 212 Resonance vibrations .............................. 7, 8 Rigidity of coatings ................................. 144 Rinsing, to prepare blade surfaces for overlay coatings .................................. 45, 46–47 Roentgen (x-ray) radiation .........................38

S Sample specific weight gain ........................97 Sample specific weight loss .........................97 Scale removal method ............................. 100 SC-TMF test. See Strain-control thermomechanical fatigue test. Sea salts .......................................... 6, 9, 12 Secant modulus for the coating material ...... 196 Semidry molding technique ............ 164, 165(T) Service life .........................................2, 3, 8 of coated turbine blades ............................. 5 coating chemical composition effects on thermomechanical loading .......202–203(T) definition ........................................... 186 and deposition technique ......................... 190 determining coating composition and thickness ........................................10 1710BL unit .......................................... 186 Sigma phase .......................................70–71



Index / 241

in cobalt-nickel-chromium-aluminum alloys ................................... 71, 72(T) in nickel-cobalt-chromium-aluminum alloys .......................................70–71 Sign of first principal strain for maximum strain intensity in coating ........................... 197 Silicification, with aluminizing ....................... 1 Silicon addition effect on hot corrosion resistance of overlay coatings ............................. 111 addition effect on saturating mixtures for aluminizing ................................. 28(F) alloying addition effect on Ni-Co-Cr-Al alloys ....................................101–103 as alloying element for diffusion coating ...... 116 as alloying element, heat resistance benefit ... 101 alloying NiAl with ................................ 101 content effect on aluminide coatings ........... 108 content effect on diffused coating thickness with slurry technique ........................... 35(F) content effect on hot corrosion resistance of Ni-Co-Cr-Al alloys ......................... 112 content in diffusion aluminide coatings ................................ 62(F), 63 effect on heat resistance of Ni-Cr-Al alloys ................................ 101, 102(F) as impurity of jet fuels .............................. 5 substitution for yttrium effect on Ni-Co-Cr-Al cyclic heat resistance ....................... 105 Silicon-and yttrium-alloyed coatings, phase transformations after high-temperature testing ..................................84(F), 85(F) Siliconizing process ...................................28 Silver, coefficient of transfer in flux, electric arc evaporation ....................................40(T) Simulations .............................................. 3 Simultaneous deposition ................. 164, 165(T) Slurry aluminizing, as coating method for diffusion coatings ............................11(T) Slurry technique ....................................... 2 for application of diffusion aluminide coatings ................................ 32–35(F) for applying diffusion aluminide coatings, phases, structures, and compositions .....62(F,T), 63 composition of powders ............................33 diffused coating thickness .........................34 explosion hazard ....................................35 and mechanical finishing afterwards .............34 surface roughness, hardness and diffused coating thickness affected by silicon content ................................. 34–35(F) Sodium in deposits detected on gas-turbine engine blade surfaces .........................................99 as impurity of jet fuels .............................. 5 Sodium chloride (NaCl), causing hot corrosion ...........................................99 Sodium salt .............................................99 causing hot corrosion ...............................99 Sodium sulfate (Na2SO4), causing hot corrosion ...........................................99

Softening .............................................. 2, 8 “Soft” heat resistant metallic coatings ......... 198 Solidus temperature,, regression equation for calculation of ................................... 213 Spalling caused by unsatisfactory adhesive strength level ........................................... 133 of ceramic layer ................................... 169 of ceramic layer, black color after check annealing .........................170, 171(F,T) of ceramic layer of thermal barrier coatings ..........166, 182–188(F), 190, 203, 206(F) of ceramics on gas-dynamic test rig ................................. 189–190(F,T) of ceramics on turbine blades with thermal barrier coatings ................................ 20–22(F) of diffusion aluminide coatings ...................38 of diffusion coating/superalloy system during isothermal test ............................... 107 as result of diffusion processes between coating and alloy ........................................79 of thermal barrier coating ceramic layer, criterion of ......................... 205–209(F) of ZrO2-15%Y2O3 ceramic ...................... 109 Specific mass gain ....................................38 Specific power .......................................... 5 Specific weight .......................................... 5 Specific weight change .. 97, 98(F), 104(F), 105(F), 110(F), 114(F) Specific weight gain .................97, 100, 102(T), 103(T), 106(T) measurement at aluminizing to determine coating thickness ........................................38 Specific weight loss ..................97, 100, 110(T), 112(T), 113(T) Specimen mass loss ................................. 100 Spinels ............................................ 83, 189 formation on bond coat surface ................. 182 formation on diffusion coating of coating/ superalloy system ........................... 107 formation with NiAl oxidation .................. 101 in oxide layer of overlay coating/superalloy system ......................................... 109 proportion in Ni-Co-Cr-Al alloys increased with cobalt content ................... 103, 104–105 Splashing ...............................................51 during electron beam evaporation technique ....47 Stabilization time of evaporation ..................39 Stabilized zirconia oxide-base ceramics ........... 2 “Stark” (statistical analysis regression correlation) ..................................... 211 Steady-state oxidation ...............................98 Strain-control thermomechanical fatigue test (SC-TMF test) .................... 150–151, 152 Strain intensity .................................. 201(F) in coating .............................. 196–197, 198 maximum value of ................................ 197 minimum value of ................................ 197 vs. temperature ................................ 199(F)



242 / Index

Strain-intensity range in the coating ................... 197, 199–200, 202 at thermocyclic loading of blade ............ 201(F) Strain range ...................................... 154(F) in coating ................................. 153, 154(F) of specimen at cycling loading .................. 199 Strains .................................................... 2 of coating and blade ........................195–196 in nozzle vanes ....................................... 7 Stress concentrators ................ 140, 141, 143(F) causing low-cycle fatigue ........................ 149 coating defects in overlay coatings ....... 141, 142 Stresses. See also Thermal stresses ............... 2, 7 Stress intensity in the coating .......... 196, 201(F) Stress of separation ....................... 208(F), 209 Sulfides ................................................... 5 Sulfur ..................................................... 6 content effect on hot corrosion ............... 11, 12 content in corrosion products ................13–14 content range of aviation fuel ...................... 5 in deposits detected on gas-turbine engine blade surfaces .........................................99 as impurity of jet fuels .............................. 5 Superalloys for aircraft turbine blades, chemical compositions .......................... 10, 11(T) coated, hot corrosion resistance .... 113–116(F,T) coating effect mechanism on high-temperature strength ............................. 137–138(F) with coatings, heat resistance ...... 106–109(F,T) creep-rupture strength in air at high temperatures .................136–137, 138(F) creep-rupture strength in media provoking hot corrosion ................134–136(F,T), 137(F) with diffusion coatings, fatigue strength ............................. 139–140(F) with diffusion coatings, heat resistance of ................................. 106–108(F,T) fatigue strength, coating effects ......... 138–142(F,T), 143(F), 144(F) high-temperature strength coating effects ............................ 134–138(F,T) interaction mechanism with overlay coatings ....................................93–96 low-cycle fatigue, coating effect on .......... 149– 150(F), 151(F) nickel-aluminum, formation of diffusion layer below 640C ..................55, 56(T), 57(F) with overlay coatings, fatigue strength ... 140–141(F,T), 142(F,T), 143(F), 144(F) with overlay coatings, heat resistance of ................................. 108–109(F,T) service life .......................... 155, 157, 183(T) thermal fatigue and coating effects ............................ 142–150(F,T) thermomechanical fatigue coating effects on ................................. 150–158(F,T) Superalloys, specific types CMSX-4

ceramic layer spalling .....................185(F), 187(T), 188, 207 chemical composition .......................11(T) microstructure of thermal barrier coating bond coat on ........................... 186(F), 188 service life ....................................... 157 thermal expansion coefficients, various temperatures .......................... 121(T) thermomechanical fatigue testing, cyclic life with coatings .... 155(F), 157(F,T), 158(F) thermomechanical fatigue testing of coated superalloy systems .......... 153, 154(F,T), 155(T), 156(F,T), 157(F,T), 158(F) Hastelloy X cycles to spallation of thermal barrier coatings on superalloy ..............184(F), 187–188 physical properties of thermal barrier coated samples ................................ 183(T) service lifetime of thermal barrier coated samples ................................ 183(T) IN 625 physical properties of thermal barrier coated samples ................................ 183(T) service lifetime of thermal barrier coated samples ................................ 183(T) JS-type, ductile-brittle transition temperatures for diffusion coatings 50lm thick ........ 120(F) JS6F chemical composition .......................11(T) cyclic testing with diffusion and overlay coatings ............................. 20, 21(F) fatigue strength ................................. 140 multilayer coating system residual stresses ................................. 134(F) thermal fatigue cracking (fragmentation) .................... 19, 20(F) thermal stresses variations with different orientation ....................... 133(F), 134 JS6F-DS thermal shock testing of coating/superalloy systems ................................ 148(T) thermofatigue testing ...... 146–147(T), 148(F) JS6F-SC, thermal shock testing of coating/ superalloy systems ...................... 148(T) JS6K aluminizing ........................................27 aluminizing from slurries, phase and chemical compositions of outer zone ...... 62(T), 63 aluminizing in aluminum-silicon mixture ................................... 28(F) chemical composition .......................11(T) coating thickness related to aluminum reserve in coating ....................................86 diffusion coating characteristics, 98% Al-Fe alloy and 2% ammonium chloride .. 64(F), 65(T) fatigue strength ................................. 139 heat resistance and fatigue strength of coated blades ...................................... 142 interaction kinetics with overlay coatings and chromium content increase ...... 90, 92(T)



Index / 243

low-temperature aluminizing and composition of coating surface zone ...............57–58 oxidation constant with diffusion aluminized coatings, isothermal tests ........... 108(T) oxide life .................................... 106(T) phase and chemical compositions of diffusion coating formed in low-activity powder mixtures ............................ 65(T), 67 phase and chemical compositions of diffusion coatings after high-temperature testing ...............................82–83(T) thermal fatigue resistance of diffusion coated superalloy ............................. 145(F) time-to-failure in air and in GZT ash ................................ 135(T), 136 JS6U aluminizing by EA method from alloy Al5SiY, phase and chemical compositions ......59, 60(F,T), 61(F) ceramic layer thickness effect on service life ................................ 183(F), 186 chemical composition .......................11(T) coating life at 800C ....................... 116(T) creep rates ................................... 136(F) creep-rupture strength in GZT ash causing hot corrosion ................... 134, 135(F), 136 cyclic life of overlay coated vs. diffusion coated superalloy ......................... 146 cyclic testing with diffusion and overlay coatings ............................. 20, 21(F) elongation of coated and uncoated specimens in GZT ash ...................... 136, 137(F) fatigue strength of coated blades at 20C .. 140, 142(T) fatigue strength of overlay coating/superalloy systems ...140(T), 141(F), 142(T), 143(F) fatigue strength of thermal barrier coating vs. no coating ....................... 141, 144(F) fatigue strength with diffusion coatings ........................................... 139(F) forces acting in Ni22Cr12AlY coating vs. coating thickness ............... 132(F), 133 heat resistance and fatigue strength of coated blades ...................................... 142 heat resistance with diffusion coatings (isothermal and cyclic tests) .. 107, 108(F) high-temperature test results of interaction kinetics with overlay coatings, chemical composition effects ............... 93, 95(T) interaction kinetics between overlay coatings and superalloys, chemical composition effects ............................... 93, 95(T) interaction kinetics with overlay coatings and chromium content increase ...... 90, 92(T) life of diffusion coatings, hot corrosion testing in GZT ash ............................ 115(F) overlay coating changes of phase composition and structure after high-temperature tests ............... 87–88(F), 89(F), 90(F), 92(T) overlay coating thicknesses and deposition method ................................. 109(T)

oxidation constant with diffusion aluminized coatings, isothermal tests ........... 108(T) oxidation rates in diffusion coating .. 15–16(F) oxide life on uncoated superalloy, cyclic testing .................................. 108(T) oxide structural change with heat resistance testing of overlay coatings/superalloy system ............................ 109, 110(F) phase transformations after high-temperature testing ............................... 84, 85(F) phase transformations, aluminized in 98%Al-Fe alloy and 2% ammonium chloride at 1000C ................................ 64(F) phase transformations, interaction mechanism with overlay coating at high temperatures .......... 89, 90–91(F), 93(F) thermal stress distribution measured during chemical removal of coatings .......................... 131, 132(F) thermal stresses from coatings calculated ........................ 132(F), 133 thermal stresses in diffusion coating vs. temperature ...................... 129(F), 130 thermal stresses in overlay coatings at 20C .................................... 130(F) thermal stresses of coatings applied to .. 131(F) thermal stresses variations in Ni20Cr12AlY coating and superalloy .............. 133(F) thermofatigue testing ................146–147(T) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) thermomechanical fatigue tests of coated superalloy ............................. 200(F) thickness variations of diffusion coatings during high-temperature testing ...... 86(F) time to failure values in GZT synthetic ash ....................................137–138 uncoated, thermal fatigue present ............ 146 weight change of overlay coating/superalloy system (isothermal test) ............. 110(F) JS6U-DS, thermal shock testing of coating/ superalloy systems ...................... 148(T) JS26 chemical composition .......................11(T) high-temperature test results of interaction kinetics with overlay coatings, chemical composition effects ............... 93, 95(T) interaction kinetics between overlay coatings and superalloys, chemical composition effects ............................... 93, 95(T) interaction kinetics with overlay coatings and chromium content increase ...... 90, 92(T) phase transformations, interaction mechanism with overlay coating at high temperatures .................. 90–91, 93(F) JS26-DS creep-rupture strength in air ...........................136–137, 138(F) fatigue strength ................................. 140 low-cycle fatigue testing ............ 149–150(F)



244 / Index

Superalloys (continued) microstructure after high-temperature testing for stress-rupture strength ........... 138(F) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) JS30, chemical composition ...................11(T) JS30-SC, thermofatigue testing ..................... 146–147(T), 148(F) JS32 chemical composition .......................11(T) density ....................................... 181(T) JS32-DS fatigue strength ................................. 140 fatigue strength of coated superalloy with overlay coatings .......................... 141 thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) MAR-M-002 chemical composition .......................11(T) intensities of strains and stresses within startmaximum-shut cycle ................. 202(T) thermal barrier coating testing on gas-dynamic test rigs and engines ........... 189(F), 190 thermal expansion coefficient and cyclic life of coating ..................................... 203 thermomechanical fatigue testing, hysteresis loop calculation ................. 152(F), 153 MAR-M-200 chemical composition .......................11(T) coated blade surface temperature at minimum and maximum gas temperature .... 202(F) cycles to coating failure of coated blades ............................ 201–202(F) overlay coatings effect on strain intensity and ductility vs. temperature ............ 199(F) thermal barrier coating testing on gas-dynamic test rigs and engines ..................... 190 Nickel chemical composition of diffusion coatings ............................. 57, 58(T) diffusion aluminide coating characteristics, aluminized in 98% Al-Fe alloy and 2% ammonium chloride ........... 62(T), 63(F) phase transformation, aluminized in 98% Al-Fe alloy and 2% ammonium chloride at 1000C ................................ 64(F) phase transformation diagrams after EA deposition method ................ 57, 58(F) thickness variations of diffusion coatings during high-temperature testing ...... 86(F) Ni20CrTi chemical composition of diffusion coatings ............................. 57, 58(T) diffusion aluminide coatings of ...57, 58(F), 59 diffusion coating characteristics, 98% Al-Fe alloy and 2% ammonium chloride ...65(T) diffusion coating for ................... 57–58, 59 formation of diffusion layer below 640C .......................55, 56(T), 57(F) metallographic structure ................ 57, 59(F)

phase and chemical compositions on diffusion coatings after high-temperature testing ...............................80–81(T) phase composition changes at high-temperature tests ..................................80–81(T) phase transformation diagrams after EA deposition method ................ 57, 58(F) thickness variations of diffusion coatings during high-temperature testing ...... 86(F) Ni10CrMoWCo, phase and chemical compositions on diffusion coatings after high-temperature testing ........ 81(T), 82(F) Ni10CrWMoCo chemical composition ..........11(T), 57, 58(T) diffusion coating characteristics, 98% Al-Fe alloy and 2% ammonium chloride ...65(T) diffusion coating for ............... 57–58, 59(F) metallographic structure ................ 57, 59(F) VJL12U chemical composition .......................11(T) cobalt and chromium content effect on thickness of interaction zone with overlay coating .............................. 93, 95(F) fatigue strength with overlay coating ........................... 140, 142(F) heat resistance and fatigue strength of coated blades ...................................... 142 interaction kinetics between overlay coatings and superalloys, chemical composition effects ............................... 93, 95(T) interaction zone with overlay coatings after high-temperature testing ......... 88, 91(F) low-cycle fatigue ................. 149(F), 150(T) phase transformations, after high-temperature testing .................................... 84(F) phase transformations, interaction mechanism with overlay coating at high temperatures .......... 91–92, 93(F), 94(F) thermofatigue testing of coating/superalloy systems ...................... 147(T), 148(F) Surface area (S) .......................................97 Synthetic ash of gas turbine fuel (GZT) 99–100, 109–116(F,T), 134–137(F,T) composition ......................................99(T) gas medium composition at testing ...........99(T) Systematization ......................................... 3

T Takeoff power, effect on temperature field ........ 5 Tantalum, cathode spot temperature and vapor pressure ........................................40(T) TBC-APS. See Air plasma spraying of powders. TBC-EB. See Vapor condensation at EB evaporation of ceramic pellets. TBCs. See Thermal barrier coatings. TC-3000H unit (Sinku-Riko Company) ....... 176 Tearing stresses



Index / 245

magnitude of ....................................... 184 thermal barrier coating testing on gas-dynamic test rigs ....................................... 189 Tear-off stresses .......................................21 TEC. See Thermal expansion coefficient. Temperature characterized by zero thermal stress level ..... 196 effect on coating and superalloy fatigue strength .................................141–142 gas flow at turbine inlet ............................. 6 related to rates of diffusion processes ............79 Temperature gradient, of nozzle vanes ............ 7 Temperature of evaporation .......................39 Temperature rise, rate of .......................... 164 Tensile stiffness of coating ........................ 197 Tensile stresses ................................7, 8, 198 relaxation in thermal barrier coatings .......... 206 TGO. See Thermally grown oxide. Thermal barrier coatings (TBC) ............................2, 3, 203–209(F) actual stresses in ceramic layer on blade leading edge ........................ 207, 208(F) adhesive strength .................................. 182 adhesive strength curve ................ 206(F), 207 blade elongation vs. service time ............. 22(F) bond coat surface preparation ...166–167, 168(T) centrifugal load for blades ....................... 195 ceramic layer adjacent to bond coat, element contents analysis .................. 170, 171(T) ceramic layer condensation rate ....... 171, 172(F) ceramic-layer cracking criterion ...... 203–205(F) ceramic layer damage causes ....180–184, 206(F) ceramic layer durability ............. 180–190(F,T) ceramic layer microstructure .......... 172, 173(F) ceramic layer properties .......................... 162 ceramic-layer spallation criterion ..... 205–209(F) ceramic layer structure, factors governing .... 166 ceramic layer testing for adhesive strength .......................187(F), 188–189 ceramic layer thermal stresses ......... 172, 173(T) ceramic layer, thickness variation ................21 correction factors for irradiation losses ................................ 177, 178(T) cracking probability ............................... 205 deposition technique features ....... 165–175(F,T) deposition technique stages ...................... 166 design, typical ................................. 161(F) electron beam deposited ............. 161–190(F,T) emissivity ................................. 180, 181(T) failure causes ...................................... 203 fatigue strength on superalloys .................. 141 influence on blade temperature field ........... 195 inner aluminide heat-resistant bond coat ...... 161 lab tests of service life .............. 184–188(F,T) life estimation ............................ 203–209(F) life expectancy .................................. 21, 22 maximum relative tensile thermal stress ... 204(F) pellet evaporation and ceramic layer formation ........................ 168–175(F,T) phase transformations ............................ 182

quality control ........................... 174–175(F) recrystallization processes in ceramic layers .. 183 requirements .......................................... 8 residual stresses in ceramic layer on blade leading edge ................................. 207, 208(F) ridges intruding into ceramic layer ......... 205(F) service life ............................. 167, 172, 182, 183(F), 184–188(F,T) service life, characteristics governing .......... 166 sintering of ceramics .............................. 183 spalling of ceramic layer .......170, 171(F,T), 175 spalling of ceramics on turbine blades .. 20–22(F) steady-state temperature field at blade leading edge ....................................... 207(F) stresses facilitating ceramic-layer separation from blade surface ............................. 206(F) testing on gas-dynamic test rigs and engines ........................... 189–190(F,T) thermal conductivity ........177, 178–179, 180(F) thermal diffusivity .....................177,178, 179 thermal expansion coefficients ....... 180, 181(T), 182(F) thermophysical properties ...........175–180(F,T), 181(F,T) ultimate adhesion strength and stress of spallation vs. time ............................. 207, 208(F) Thermal barrier effect ...............................21 Thermal conductivity, of ceramic layer of thermal barrier coatings ....................... 175, 176(F) Thermal diffusivity, determination by pulse method .....................................175–176 Thermal energy .......................................38 Thermal expansion coefficients (TECs) ..... 19, 20 causing thermal stresses ............................20 of coatings, variations and their causes .... 133(F) for Co-Ni-Cr-Al alloys ................. 124, 125(T) difference between coating and blade materials ...................................... 198 elasticity modulus of coating alloys ..120–124(T) mismatch between coatings and substrates for compressive stresses ..................130–131 of nickel-chromium-aluminum alloys ...... 121(T) of nickel-cobalt-chromium-aluminum alloys ............................. 122–124(F,T) range 20–400C, regression equation for calculation of ................................ 214 range 20–500C, regression equation for calculation of ................................ 214 range 20–600C, regression equation for calculation of ................................ 214 range 20–700C, regression equation for calculation of ................................ 214 range 20–800C, regression equation for calculation of ................................ 214 range 20–900C, regression equation for calculation of ................................ 215 range 20–1000C, regression equation for calculation of ................................ 215 range 100–300C, regression equation for calculation of .......................... 213, 216



246 / Index

range 100–400C, regression equation for calculation of .......................... 213, 216 range 100–500C, regression equation for calculation of .......................... 213, 216 range 100–600C, regression equation for calculation of .......................... 213, 216 range 100–700C, regression equation for calculation of .......................... 213, 216 range 100–800C, regression equation for calculation of .......................... 213, 216 range 100–900C, regression equation for calculation of .......................... 213, 216 range 100–1000C, regression equation for calculation of .......................... 213, 216 Thermal fatigue ........................................ 7 coating effects on superalloys ...... 142–150(F,T) of overlay coatings ................................... 9 of turbine blades ...................19–20(F), 21(F) Thermal fatigue cracking ...... 1, 9, 10, 19–20(F), 21(F), 144 diffusion coatings .......................... 20, 21(F) of hollow samples of coated superalloys .. 147(T) overlay coatings ............................ 20, 21(F) on vane surfaces ...................................... 7 Thermal fluxes .........................................21 Thermally grown oxide (TGO) .................. 180 causing spalling between ceramic layer and bond coat ............................ 185(T), 188, 190 layer formation spending aluminum from bond coat .................................. 187(T), 188 thickness of, versus test time .......... 183(F), 186 Thermal shock, as thermal stress manifestation .................................... 148 Thermal shock testing ......................... 148(T) Thermal strains, calculation of ................... 196 Thermal stress cracking, in coatings ...........................19–20(F), 21(F) Thermal stresses ......................................21 calculation of ...................................... 128 in ceramic layer of thermal barrier coating, cracked .................................... 204(F) in diffusion aluminide coatings ....... 129–130(F) factors for generation of .......................... 128 initiating cracks or spalling ...................... 195 methods for studying those of coatings ........ 128 in nozzle vanes ....................................... 7 in overlay coatings ............................ 130(F) thermal barrier coating testing on gas-dynamic test rigs ....................................... 189 and thermal fatigue cracking .................... 144 Thermodynamic potential ..........................27 Thermofatigue testing ................ 142–150(F,T) equipment schematic diagram ............... 145(F) results ..................................... 146, 147(T) specimen sketches ....................... 145–146(F) test conditions (temperature/time) .......... 146(T) Thermomechanical cyclic loading ............... 151 Thermomechanical fatigue coating effect mechanism .............. 155–158(F) criterion ................................ 199–203(F,T)

of superalloys, coatings effects on ................................. 150–158(F,T) Thermomechanical fatigue testing 150–158(F,T) experimental method ............................. 152 specimen loading conditions ................ 154(T) test conditions ........................ 153–155(F,T) Thermomechanical life, prediction of ........... 200 Thickness of zone of coating (hi) ..................18 Thiophenes .............................................. 5 Thrust .................................................... 6 Time of evaporation ..................................39 Time of high-temperature test (s) .................90 Time of oxide lifetime (sp) ..........................98 Time to ceramic-layer spalling ............... 206(F) Time-to-failure to stress ........................... 135 Tin cathode spot temperature and vapor pressure .....................................40(T) as impurity of jet fuels .............................. 5 Titanium, as impurity of jet fuels .................... 5 Transfer coefficient (l), of electric arc technique ..........................................36 Tungsten alloying NiAl with ................................ 101 cathode spot temperature and vapor pressure ................................................40(T) coefficient of transfer in flux, electric arc evaporation .................................40(T) solubility in gamma double-prime phase compound after high-temperature testing ...........................................82 Turbine-blade coating protective properties ............................ 97–117(F,T) Turbine blade coatings aircraft-engine, causes ..................... 10–22(F) defects, causes .......................................10 principles for choosing for aircraft engines ...................................... 8–10 requirements for ...................................... 8 selection parameters for aircraft engines ..... 9(F) Turbine blade operation, parameters determining operating conditions .............................. 5 Two-way diffusion ....................................80

U UE-type coater diagram of central chamber ............... 41, 42(F) specifications .....................................41(T) Ultimate strain criterion ................ 198–199(F) Ultimate strength of coating material ................................ 201 of cobalt-nickel-chromium-aluminum alloys ...................................... 127(T) of nickel-chromium-aluminum alloys .... 126(F,T) of nickel-cobalt-chromium-aluminum alloys ............................ 127(T), 129(F) 700–1100C, regression equation for calculation of ................................ 215



Index / 247

Uniform corrosion .......................12, 13, 14(F) Uniform hot corrosion, of engine blades running at high temperature ..............................14

V Vacancy sinkers .......................................79 Vacuum arc remelting, for melting alloys for coatings deposited by the EB technique ......44 Vacuum-induction (VI) melting, to melt alloys for coatings deposited by the EB technique ......44 Vanadium content effect on hot corrosion ............... 11, 12 content in corrosion product .......................14 in deposits detected on gas-turbine engine blade surfaces .........................................99 Vanadium oxide ........................................ 5 causing hot corrosion ...............................99 Vapor condensation at EB evaporation of ceramic pellets (TBC-EB) ................... 162 VAR. See Vacuum arc remelting. Variable elasticity parameters method ......... 197 Variable stresses, in blades ........................... 8 VI. See Vacuum-induction melting. Vibration loads ......................................... 8 Vibrogrinding ....................................... 166 Vibropolishing ....................................... 166 to prepare blade surfaces for overlay coatings ................................ 45(F), 46 Visual inspection, to detect diffusion coating discontinuity ......................................38

W Warping ................................................. 7 Warplane engines, temperature and stress fields nonuniform ......................................... 5 Weight after corrosion product removal (mk) ..97 Weight with all its oxidation products (mox) ....97

X X-ray diffraction analysis of ceramic pellets and temperature changes ............................. 169–170(F) of crystallite texture of bond coats ............. 166 of thermal stresses ....................130, 131, 132

Y Yield strength of cobalt-nickel-chromium-aluminum alloys ...................................... 127(T)

of nickel-cobalt-chromium-aluminum alloys ............................ 127(T), 129(F) 700–900C, regression equation for calculation of .............................................. 215 Ytterbium, dopants for zirconium oxide crystal lattice .......................................179–180 Yttrium alloying addition effect on nickel-cobaltchromium-aluminum alloys .........101–103 content effect on formation of beta-phase-free zone ........................................... 105 content effect on heat resistance of Co-Ni-Cr-Al alloys .......................................... 106 content effect on hot corrosion resistance of Co-Ni-Cr-Al alloys ......................... 113 content effect on hot corrosion resistance of Ni-Co-Cr-Al alloys ......................... 112 content in diffusion aluminide coatings ................................ 62(F), 63 content requirements for bond coats of thermal barrier coatings .............................. 161 oxidation in melt of overlay coatings ............44 replacement with hafnium effect on oxides of Ni-Cr-Al alloys .................... 101, 102(T) Yttrium oxides, formation with hot corrosion .. 113

Z Zinc coefficient of transfer in flux, electric arc evaporation .................................40(T) as impurity of jet fuels .............................. 5 Zirconia, heat capacity increase with temperature rise ...................................... 177, 178(F) Zirconia • yttria oxide, temperature restrictions above 1150 C due to high oxygen-diffusion mobility ............................................22 Zirconium, alloying NiAl with .................... 101 Zirconium oxide dissociation, during ceramic pellet evaporation, products and degree of ............................................. 168(T) Zirconium oxide-yttrium oxide ........ 162–163(F) application techniques ............................ 162 ceramic pellets, chemical and phase compositions made by mechanical mixture of powders ............................. 163, 164(T) ceramic pellets, chemical and phase compositions, made by simultaneous deposition .......................... 164, 165(T) ceramic pellets, phase composition, volume % .......................................... 165(T) ceramic pellets, quality-control system ........ 165 color change of ceramic layers and phase composition ........................ 169, 170(T) phase composition changes, ceramic pellets ..................................... 164(T) solid-solution synthesis techniques ............. 163



Product code


06738G






Y. Tamarin-Protective Coatings for Turbine Blades (2002).pdf

Recommend Documents