Newsletter. Vol 4, Issue 1. (June 20, 2011) : 1-9. By Alejandro Vega, P.E. As Published in the PV Newsletter.
ABSTRACT: This paper reviews the continued use of ASTM A-212 steel pressure vessels. It provides a discussion of the embrittlement phenomena, methodology behind Minimum Design Metal Temperature, issues with continued use of A-212 pressure vessels, and limitations on use imposed on post-1988 ASME design rules.
ASTM A-212 Pressure Vessel Steel – A Case Against Continued Use ASTM A-212 was a specification for a high tensile strength, Carbon-Silicon steel plate for use in boilers and pressure vessels. The standard was withdrawn in 1967. In the case of pressure vessel manufacture it was specified as a Fire Box Quality steel and specified to be processed per ASTM A-300, Specification for Notch Toughness Requirements for Normalized Steel Plates for 1 Pressure Vessels (withdrawn in 1975) , if specified for service at low temperatures. Today it is widely recognized that the older carbon steels, such as A-212, have reduced low temperature toughness, meaning that these types of steels have a low resistance to low temperature brittle fracture. Because some pressure vessels constructed with A-212 are still found in service, or are considered for service, it is not considered good engineering practice to use original editions ASME Code that was in effect at the time of construction to determine minimum design temperatures when performing Fitness-for-Service (FFS) evaluations. Review of older ASME Code material specifications reveal that it was common for such older steels to have been exempted from impact testing to temperatures of -20 F, but these same steels evaluated to the current ASME Codes would actually have minimum design metal temperature
(MDMT) near 100 F. Up until the late 80s most carbon steels less than 4-inch thick were
considered good for use at temperatures down to -20 F. With the 1988 ASME Pressure Vessel Code changes, a re-evaluation of pressure vessels under FFS methods identifies carbon steels, such as A-212, as not adequate for service at ambient temperatures. Since the withdrawal of the A-212 Specification and the introduction of Specifications for ASTM A-516 (fine grain) and A517 (course grain), it has become a common practice, though not recommended, to compare these metals to A-212 in MDMT evaluations.
One problem with the continued use of pressure vessels made from A-212 is that some have been in operation for an extremely long amount of time and may have both a high pedigree and documented history and show no signs of temperature embrittlement. Every so often a vessel constructed with A-212 will be evaluated for continued use and the pressure vessel blog sphere will abound with questions as to how to approach the MDMT evaluation process. All of these types of vessels’ vessels ’ construction pre-date the revised 1988 Code requirement for MDMT and as stated previously, were designed with a default minimum temperature of -20ºF. The nameplate information on the vessels may even list a minimum temperature of -20ºF. For example, a case Page 1 of 11
of a 1960’s vessel investigated by the author calculated an MDMT 114ºF for the shell and 102ºF for the heads. To rerate such a vessel would drastically reduce its operating pressure and thus make it inapplicable for use. The concern for A-212 steel is the onset of brittle fracture and subsequent catastrophic failure which can cause injury to personnel or effect damage to infrastructure; therefore operation above the MDMT is instrumental to avoiding brittle fracture. Compounding operational issues associated with MDMT are unforeseen events during operation, such as isentropic freeze up or other upset conditions occurring that may introduce temperatures less than MDMT, which if not planned for may precipitate brittle fracture. There are generally three conditions that contribute 2 to brittle fracture : 1) the presence of a flaw, 2) stress at great enough levels to induce flaw growth, and 3) the presence of temperature below the Nil-Ductility Transition temperature (NDTT); the temperature at which the material will transition from being ductile to becoming brittle. The onset of brittle behavior need not have visible flaws, and any presence of flaws may be the result of upset conditions, construction, corrosion, or service fatigue, such as micro-cracking. This type of scenario combined with normal stress at operating conditions or elevated stress at upset conditions, in combination with temperature below the NDTT, can lead to catastrophic failure. Therefore, one of the goals of the MDMT Code changes is to control and monitor the low temperature operation within the ductile operating region of the metal. In other words, the rationale for MDMT is to gear pressure vessel design toward avoiding brittle fracture and restricting operating temperatures within a ductile material regime of the Impact Energy Transition Curve (Reference Fig 1).
Figure 1. Charpy Ductile-to-Brittle Ductile-to-Brittle Temperature Transition Curve for A-212B
Page 2 of 11
To address the impacts that changes in Code requirements have had on regulating agencies, the National Aeronautics and Space Administration (NASA) has specifically addressed MDMT with 3 the following example, within one of its mechanical integrity, pressure systems guidelines : Code PV/S [Pressure Vessel/Systems] for which current Code requirements have changed from the original fabrication Code shall be reassessed and re-rated as necessary to assure an acceptable risk level. Note: For example, the 1988 changes to fracture toughness rules for prevention of brittle fracture could significantly increase the assessed risk of continued operation at the original design limits. Thus, a 4 inch thick vessel fabricated from A-212 Grade B (Firebox) material is now known to have an allowable minimum design material temperature (MDMT) of 118 degrees F. The vessel nameplate likely shows an MDMT of – 20 20 degrees F. If the vessel normally receives ambient compressed gas at 60 degrees F, the vessel would require risk reassessment and likely additional hazard mitigation to assure continued safe operation. Origin of MDMT Code Changes
In the 1988 Code revisions, ASME Section VIII, Division 1, Section UG-20 requires the establishment of an MDMT based on the lowest temperature experienced during operation, upset 4 conditions, auto-refrigeration, ambient conditions, and any other ancillary sources of cooling . Sections UCS-66, 67, and 68 grant exemptions for impact testing based on operating conditions. Service failures due to brittle fracture in pressure vessels are the result of tri-axial states of stress and temperature acting at the notch tip. It has been empirically determined that steels having near identical properties when tested in tension or torsion at slow rates of strain will show marked variances in their propensity for brittle fracture during notched-impact testing. From this it can be understood why attempting to use the MDMT of replacement steels, such as A-516 and A517, for A-212 would be strongly discouraged. MDMT criteria is indirectly indirectl y derived from Impact Energy Transition Curves. The Charpy VNotch curves have been used in engineering for the purpose of identifying materials that are resistant to brittle fracture over a large range of temperatures by determining the amount of energy a material absorbs during fracture. The results of the Charpy test plots provide the material temperature curve to identify the Ductile-to-Brittle Transition. This locus defines the Ductile-to-Brittle Ductile-to-Bri ttle Transition-Temperature Transition-Temperatu re curve (DBTT). This high strain-rate test correlates the absorbed energy to the material’s toughness and trends the material’s material’s temperature dependent 5 brittle-to-ductile brittle-to-duct ile transition. One specific example of a Charpy DBTT curve for A-212 is shown
Page 3 of 11
in Figure 1, above. In this particular example published in 1960s, the results of the data locus appear to show ductile behavior down to the 20 F, with the NDTT below this temperature.
Standard engineering design approach is to identify materials for use that can be subjected to severe in-service loading conditions, as determined through strength of materials methods, without considering stress concentrations due to cracks or the use of fracture mechanics. Notch toughness is then defined as the material’s ability to absorb energy in the presence of a flaw, 6 usually under dynamic loading . Figure 2 shows a typical Impact Energy Transition Curve with its associated NDTT. The material selection process would use this curve and NDTT to identify the operating temperature where all service related failure will be through purely plastic deformation, within the purely ductile region of the curve. This area corresponds to the upper section of the graph shown and denoted as the Plastic region.
Figure 2. Impact Toughness-Temperature and Brittle to Ductile Transition
Ductile failure will occur in metals which can sustain substantial plastic deformation before exhibiting fracture, usually at their breaking strength. Ductile failure will require the material to plastically deform, requiring more strain energy and slowing the process of fracture, thereby 6 providing time and visible indicators to allow for the correction of the problem . Brittle fracture on the other hand, results from material cleavage through slip planes, grain boundaries, and more specifically in regions of flaws or cracks; usually occurring at much lower stresses levels than those associated with ductile failure ( well below the material’s yield strength). strength). Brittle fracture in Page 4 of 11
materials with near-ambient NDTT, such as A-212, are even more susceptible to failure if drivers such as fatigue micro-cracking or surface damage, such as grinding or machining marks are induced over time. Brittle fracture failure modes suggest that the changes in the Code were driven to require the MDMT to correspond at, or above the yield criterion (YC), the point beyond which only ductile failure of steel is possible, usually taken as a temperature 60ºF higher than the NDT 2,7 temperature . This is the rational as used in ASME’s Fracture Toughness Requirements for 7,8
Materials, NB-2300 , for identifying the Reference Temperature, RT NDT = TNDT + 60 F.
To better understand the dynamics associated with brittle to ductile transitions and the temperature effects promoting brittle fracture, a correlation between changes in tensile stress and yield strength as a function of temperature is shown in Figure 3. The plot is used for the purpose of illustration and does not represent an actual curve for A-212, but is based on the result of 2 study of Pellini, et al , for carbon steels. The Fracture Transition Plastic (FTP) point in Figure 3 corresponds to the temperature where the mechanism of fracture changes from strictly ductile to brittle and where the probability of fracture below this point is no longer negligible. At temperatures above the FTP, only fully ductile tearing will occur. The Fracture Transition Elastic (FTE) point is the temperature where elastic fracture no longer propagates, or the highest 7 temperature of fracture propagation by purely elastic stresses . The Crack Arrest Curve shown is the plotted relationship between the stresses required at an associated temperature for a crack to 7 propagate . This Stress-Temperature Stress- Temperature Diagram is the basis for Pellini’s Fracture Assessment Diagram and fracture initiation curves. Bas ed on Pellini’s work allowable minimum service 9 temperatures have been defined for structures containing sharp cracks as limited by an allowable stress ( ) as a function of the yield stress ( ys) and the NDTT: 1. T min min ≥ NDT: Permissible when applied stress s is less than 5-8 ksi. 2. T min min ≥ NDT + 30 F: Permissible when
≤
ys
3. T min min ≥ NDT + 60 F: Permissible when
≤
ys
/2
4. T min min ≥ NDT + 120 F: Permissible since failure will not occur below the ultimate tensile strength of the material
One specific characteristic that should be noted from Figure 3 is the relative change in NDTT for an un-cracked specimen versus a specimen with a small flaw, which is rather drastic with respect to changes in temperatures. In this example it is a 50 F higher temperature shift for a specimen with a small flaw.
Page 5 of 11
Figure 3. Stress-Temperature Diagram for Crack Initiation and Arrest
7
The yield strength in the figure is the stress corresponding to the transition from elastic to plastic behavior (provided certain conditions are maintained: presence of cracks, high stresses in the presence of flaws, or low temperatures below the NDTT). The diagram correlates stress, temperature, material strength, and crack arrest properties. It can be noted that as temperature is decreased the relative material strength increases (both tensile and yield). The rate of increase in the tensile strength is less than the rate of increase in the yield strength, ultimately leading to a convergence. 7
In carbon steels this convergence temperature may be in the order of 10 F , meaning that below this temperature there is no initiation of yielding prior to failure, causing the failure mechanisms
to be entirely brittle. This point is at the NDTT. For A-212 the NDTT is listed as 20 F and
40 F, depending on the study results referenced
10,11
.
Based on this relationship, the crack arrest curve shows that crack propagation will favor smaller flaw size, higher stresses, and lower temperature. For A-212 carbon steel pressure vessels that have been in service for a long period of time and operating near the NDT temperature, this can be especially critical. Further, the total energy absorbed during a standard V-notch test at a given temperature, may not be representative of the energy required to re-initiate growth in a sharp Page 6 of 11
crack already present in the material, which can be quite low. Therefore, the initiation energy for a fatigue pre-cracked specimen may be much smaller than the total energy of a standard V-notch test result. More important to pressure vessel design is the fact that the results of Charpy impact testing have an inherent deficiency due to the relatively small size of the specimen used in testing. A standard Charpy test taken at a given temperature may show a high shelf energy, but the same material in service and with a thick section structure may exhibit a low toughness at the same test 6 temperature ; therefore the Charpy Impact Test is not always a realistic model of actual inservice situations. It can be stated that in cases where very thick sections are exposed to large through-thickness stresses, the triaxial state-of-stress occurring ahead of a crack will reduce the 11 apparent ductility and notch toughness . An example of this is shown in Figure Figure 4.
Figure 4. Impact Toughness-Temperature Curve for Thick and Thin Specimens. ASME MDMT Impact Testing Exemption Curves
The ASME Impact Test Exemption Curves used to determine MDMT without the need for impact testing have been simplified in such a way as to allow for material identification to a specific curve in UCS-66 (reproduced as Figure 5), whereby MDMT can be determined as a function of material section thickness. For ASTM A-212, a Curve A material, the MDMT is 18 F for material sections slightly greater than 3/8-inch, and up to 120 F for a 6-inch thick section. It should also be noted that the nominal thickness is limited to 4-inch for welded construction. What this means is that for pressure vessels with section thickness greater than 1
inch the MDMT reaches 70 F.
Page 7 of 11
13
Figure 5. ASME MDMT Impact Test Exemption Curves MDMT Impacts and Subsequent Studies on A-212
The changes in the Code as they relate to the impacts for use of A-212 steel have generated various studies ranging from railroad tank cars to ship hull designs employing this type of steel. 10 One particular study of interest focused on A-212 Grade B due to it being identified as a fast fracture steel plate. In this study, plates of ¾-inch thickness were tested for fracture at temperatures between -100ºF and 75ºF. The results of testing indicated that at sufficiently low temperatures, fast fracture could be initiated in A-212 steel, if sufficiently sharp initiation areas were located within the field of relatively high applied and/or residual stresses. Of particular interest are the statements listed below: “The results may be summarized by noting that a very sharp ended defect was found to be a prime requirement for the occurrence of a brittle fracture. For instance, fracture did not result from a 12- inch central slot with ends sharpened by a jeweler’s saw nor from a test in which a specimen was subjected to arc strikes, gouges, slag, inclusions or porosity. Fatigue cracks, however, were a successful means of triggering fracture especially when a high residual tension stress was also present. Although the data showed a great deal of scatter in most cases, an applied stress of 90 percent of yield or greater was required. This means, of course, that the material could not be classed as Page 8 of 11
particularly notch sensitive. Also, a reduction in temperature had the expected effect of tending to lower fracture stresses for a given defect length.
The study then goes on to state that there was “...not a sufficient number [of test specimens] to constitute a statistical sampling. …repeatability does exist to the extent that fracture at a temperature near the NDT consistently be induced if an attempt is made to provide a sharp initiation site (such as a fatigue crack) located in the presence of a high residual strain field . Conversely, if either of the above requirements is purposely minimized, fracture at a temperature near the .” NDT consistently does not occur .” The results of these findings show the correlation that brittle fracture at temperatures near NDT is dependent on the existence of an initiation site, such as a flaw or service fatigue to initiate a micro-crack. Though based on the statements above, it would lead the responsible engineer to conclude that the continued use of a pressure vessel made out of A-212 is warranted if no initiation site is identified, the limitations of the study to subjecting the test specimens to only tensile loading would introduce concern that these results could not be readily correlated to pressure vessel induced tri-axial stress conditions, or that the presence of fatigue micro-cracking may not be readily identifiable through standard radiographic or other non-destructive testing techniques. This latter concern is what is at the core of continued use of A-212 pressure vessels, steel fatigue and the possibility of induced initiation sites. The goal of pressure vessel design is the dependence on conservative fatigue life for the purpose of avoiding fatigue fracture growth, due 12 to micro-cracking under high tri-axial stress distributions . Results of Stress vs. Flaw Studies for A-212
In another material properties study for A-212 steel performed by Irwin, et al, published in 11 Fracture and Fatigue Control in Structures , the calculation of nominal design stress was derived to satisfy a leak-before-break (LBB) criterion. Where, [T]he leak-before-break criterion was … a means of estimating the necessary toughness of pressure vessel steels so that a surface crack could grow though the wall and the vessel “leak” before fracturing. That is, the critical crack size at the design stress level of a material meeting this criterion would be greater than the wall thickness of the vessel so that the mode of failure would be leaking (which would be relatively easy to detect and repair) rather than catastrophic fractures.
Page 9 of 11
The results from this study are shown in Table 1, taken at two test temperatures, which are at 60 F and 120 F above the NDTT for the material. (Reference Pellini and Puzak NDTT
temperature correlations above.) Note that in this study the NDTT is defined as 20 F. What is evident from the results is that as the test temperature nears the NDTT and the sample thickness increases, the allowable design stress greatly decreases, in order to meet the LBB criterion. As expected, K ID, the critical-stress intensity factor, increases with temperature. K ID represents the material’s material ’s inherent ability to withstand a given stress-field intensity at the crack tip under plainstrain conditions.
Temp °F 80
140 *
y
KID
ksi
ksi-in
55.9
50.1
1/2
51.1
57
B in
ksi
1.0
36
2.5
19
5.0
13
1.0
55*
2.5
23
5.0
15
- greater than the yield stress.
K ID - material toughness for dynamic loading. B
- Steel sample thickness.
Table 1. Allowable Design Stresses Stresses to Satisfy LBB Criterion Criterion for A-212 Steel Having Cracks Equal to Wall Thickness. Conclusion
ASME post-1988 rules for construction imposed new impact testing exemption rules for determining MDMT. These new rules were based on the results of material research that better defined correlations between material embrittlement, service temperatures, section thickness, and the presence of flaws, with the overall goal to place pressure vessel design and any subsequent in-service failure to occur within the materials ductile range; where the damage could then be noticed and mitigated, instead of reaching a state of catastrophic and brittle fracture. Pressure vessels manufactured from ASTM A-212 are still in use and the re-assessment for use under Fitness-for-Service rules requires that MDMT be calculated under the newer ASME Code rules. Such evaluations will render the vessels, in all likelihood, not suitable for service under their original design specifications. specificatio ns. The use of A-212 vessels under these temperature conditions is therefore not recommended and should only be pursued through the use of extensive risk assessment and additional hazard mitigation practices, such as employing operational controls (engineering and administrative) by limiting personnel exposure to the area of probable hazard and containing the effects of such a hazard if it were to occur. The question of continued use should not be considered on the basis of need, but rather personnel safety and liability. Page 10 of 11
References: 1. ASTM Designation: A 212-64, Standard Specification for High Tensile Strength Carbon-Silicon Steel Plates for Boilers and Other Pressure Vessels. 2. Pellini and Puzak. “ Fracture Analysis Diagram Procedures for the Fracture-Safe Engineering Design of Steel Structures”. U.S. Naval Research Laboratory. March 15, 1963. 3. NASA-STD-8719.17, NASA Requirements for Ground-Based Pressure Vessels and Pressurize Systems (PV/S). 09/22/06. 4. ASME Section IX Welding Qualifications, CASTI Guidebook Series Vol 2. M.J. Houle. CASTI Publishing. 1999 5. R.G. Gerggren. “ Typical Impact Energy Transition Curve for Carbon Steel ASTM A-212B, Radiation Effects in Ferritic Steels”, U.S. Atomic Energy Commission Report, Office TID-7588, 1960. 6. Bhat, C.P. “ Seminar on Brittle and Ductile Fracture”, Department of Mechanical Engineering Manipal Institute of Technology. Technology. 2008. 7. DOE-HDBK-1017/2-93, DOE Fundamentals Handbook – – Material Science. January 1994. 8. ASME Section III, NB-2300, Fracture Toughness Requirements for Materials. 1974. 9. Hertzberg. Deformation and Fracture Mechanics of Engineering Materials. 4th Ed. R.W. Wiley & Sons. 1996. 10. SSC-204, Simulated Performance Testing for Ship Structure Components. Southwest Research Institute, 1970 11. Rolfe and Barsom. Fracture Fra cture and Fatigue Control in Structures: Applications of Fracture Mechanics. Prentice-Hall, Inc. 1977 12. SSC-170, Studies of Some Brittle Fracture Concepts, University of Illinois, September 1965. 13. ASME Section VIII, Div 1. Figure UCS-66.
Copyright Notice: Figures used in this publication are included und er the “Fair Use” clause of the Copyright Act of 1976 as amended in 1992, within the context of non-profit, educational purpose use only.
Page 11 of 11
