Sacmi Vol 1 Inglese - II Edizione

Applied Ceramic Technology

10


10

Applied Ceramic Technology Volume I

Copyright 2005 SACMI IMOLA Via Selice Provinciale 17/A - 40026 Imola (BO) Italy Tel. 0542/607111 - Fax 0542/642354 www.sacmi.it e-mail: [email protected] Not for sale All rights reserved. Translation, electronic storage, reproduction or adaptation by any means (microfilm and photocopy included), whole or partial, is prohibited. ISBN 88-88108-48-3 Editrice La Mandragora s.r.l. Via Selice 92 - Cas. Post. 117 - 40026 Imola (Bo) Italy Tel. 0039-0542/642747 - Fax 0039-0542/647314 e-mail: [email protected]

CONTENTS

To the reader................................................................................................................................. 9 Introduction ................................................................................................................................ 11 The production process ........................................................................................................... 22 Chapter I Identification and characterisation of ceramic raw materials ....... 25 Sampling ...................................................................................................................................... 26 Chemical analysis ....................................................................................................................... 27 Mineralogical (or crystallographic) analysis ...................................................................... 32 Thermal analyses ....................................................................................................................... 33 Plastic, semi-plastic and non-plastic raw materials .......................................................... 41 Chapter II Rocks .................................................................................................................. 45 The rock cycle ............................................................................................................................ 45 Magma .......................................................................................................................................... 47 Clays and rocks .......................................................................................................................... 47 Primary rocks ............................................................................................................................. 48 Main minerals in the earths crust ........................................................................................ 51 Structure of clayey minerals .................................................................................................. 52 Characteristics of the main clayey minerals ...................................................................... 58 Chapter III Raw materials for ceramic bodies.......................................................... 59 Kaolinite ....................................................................................................................................... 59 Illite ............................................................................................................................................... 67 Montmorillonite (smectite) ..................................................................................................... 74 Chlorite......................................................................................................................................... 80 Talc ................................................................................................................................................ 86 Non-plastic materials ............................................................................................................ 92 Silica ............................................................................................................................................. 92 Feldspathic minerals ................................................................................................................. 99 Pyrophyllite ............................................................................................................................. 105 Wollastonite ............................................................................................................................. 109 Carbonates ................................................................................................................................ 112 Accessory minerals present in ceramic raw materials .................................................. 118 Halloysite .................................................................................................................................. 118 Attapulgite, Sepiolite, Vermiculite ..................................................................................... 122 Micas .......................................................................................................................................... 126 Mineral hydroxides ................................................................................................................ 128 Soluble salts ............................................................................................................................. 129 Vegetable substances and reducers .................................................................................... 134

Sulphur and sulphides (alunite)........................................................................................... 136 Vitreous materials .................................................................................................................. 137 Chapter IV Raw materials for frits and glazes ...................................................... 139 Raw materials for melting of frits ..................................................................................... 139 Technological aspects of frits and glazes ........................................................................ 141 Standard raw materials and their influence on the characteristics of glass .......... 145 Types of frit ............................................................................................................................ 150 Classification of ceramic glazes .......................................................................................... 156 Traditional double-firing ...................................................................................................... 160 Fast double firing .................................................................................................................... 162 Porous single firing wall tiling ........................................................................................... 166 Silk-screen printing products .............................................................................................. 167 Single firing floor tiling ........................................................................................................ 169 Dry-application grains .......................................................................................................... 171 Chapter V Physical and structural properties of ceramic raw materials .... 175 Particle size distribution ....................................................................................................... 176 How water influences ceramic systems............................................................................. 186 Chapter VI Rheology: basic concepts ....................................................................... 191 Rheology of clays ................................................................................................................... 201 Mineralogy and rheology of clays .................................................................................... 202 Rheology of ceramic bodies ................................................................................................ 204 Influence of grinding water ................................................................................................ 205 The wet grinding process and spray drying of the ceramic bodies from a rheological standpoint ........................................................................................................ 206 Rheological additives ............................................................................................................. 213 Inorganic deflocculants ......................................................................................................... 215 Organic deflocculants ............................................................................................................ 217 Main additive classes ............................................................................................................. 220 Ceramic spray drying bodies ............................................................................................... 221 Ceramic glazes ......................................................................................................................... 223 Glazes: non-rheological side effects ................................................................................... 227 Most commonly used additives ........................................................................................... 228 Chapter VII The removal of water ............................................................................ 231 The water-clay system and how it affects moulding ..................................................... 231 Removing the water ............................................................................................................... 232 Chapter VIII Description of ceramic products .................................................... 245 Interpreting the behaviour of ceramic bodies ................................................................ 248 Chapter IX Wall tiles ...................................................................................................... 255 Aesthetic features ................................................................................................................... 258 Raw materials for bodies ....................................................................................................... 260

Characteristics of raw materials for bodies ..................................................................... 261 Plastic raw materials .............................................................................................................. 261 Complementary raw materials ............................................................................................ 263 Body composition ................................................................................................................... 263 Product features ...................................................................................................................... 265 Raw materials for glazes ....................................................................................................... 266 Basic technological parameters ........................................................................................... 270 Plant engineering solutions ................................................................................................. 280 Machines ................................................................................................................................... 282 Weighing systems .................................................................................................................. 283 Mills ........................................................................................................................................... 283 Spray driers .............................................................................................................................. 285 Presses ....................................................................................................................................... 285 Driers ......................................................................................................................................... 286 Glazing machines ................................................................................................................... 287 Kilns ........................................................................................................................................... 288 Sorting ....................................................................................................................................... 290 Handling and storage systems ............................................................................................ 290 Conclusions .............................................................................................................................. 290 Chapter X Floor tiles ...................................................................................................... 293 The market ............................................................................................................................... 294 Product classification ............................................................................................................. 296 Technical features ................................................................................................................... 296 Aesthetic features ................................................................................................................... 298 Raw materials for bodies ....................................................................................................... 299 Body composition ................................................................................................................... 300 Raw materials for glazes ....................................................................................................... 303 Basic technological parameters ........................................................................................... 304 Plant engineering solutions ................................................................................................. 308 Machines ................................................................................................................................... 311 Chapter XI Porcelain tiles ............................................................................................. 321 Technical characteristics ...................................................................................................... 322 Commercial specifications .................................................................................................... 323 Raw materials for bodies ....................................................................................................... 327 Compositions ........................................................................................................................... 329 Basic technological parameters ........................................................................................... 333 Production technology .......................................................................................................... 336 The production process ........................................................................................................ 336 Technical outlook ................................................................................................................... 345 Aesthetic outlook .................................................................................................................... 346 Conclusions .............................................................................................................................. 347 Chapter XII Accessories and trims ............................................................................ 349 The market ............................................................................................................................... 350


Technology .............................................................................................................................. 352 Materials ................................................................................................................................... 356 Organisation ............................................................................................................................. 358 Trims ......................................................................................................................................... 358 Production line ........................................................................................................................ 360 Machines ................................................................................................................................... 365 The production process ........................................................................................................ 366 Decoration ................................................................................................................................ 366 Firing ......................................................................................................................................... 369 Sorting ....................................................................................................................................... 370 Cutting the material ............................................................................................................... 370 Appendix 1 Standards ..................................................................................................... 373 Appendix 2 Tables and figures .................................................................................... 379 Bibliography ........................................................................................................................... 443

8

Introduction

TO THE READER

The creation of a systematic compendium providing information on ceramic production technology, with a clear focus on pressed tiles, is something of challenge as readers of such a publication are likely to come from a wide range of fields and backgrounds, spanning both the academic and industrial. Such a work might be thought of as an introductory manual providing descriptions of work cycles and machines, leaving aside more detailed understanding of the relevant materials and problems or, vice versa, as an in-depth text on technological procedures, the results of which are taken for granted. Readers with a technical-scientific background would, perhaps, like to see information on innovative and original scientific developments or at least a collection of information that would be unavailable in a more compact publication. On the other hand, readers with a technical-manufacturing background would be more interested in a troubleshooting manual to help them deal with the day-to-day problems that affect production lines. The young technician a newcomer to the complexities of a production process about which he has learnt little or nothing during his academic studies would probably require an overview packed with everything he needs to know about what happens in the factory, from delivery of raw materials through to packaging of the finished product. To say that this first volume (and the forthcoming Volume II, which illustrates the production process), attempts to reconcile all these different needs is not as presumptuous as it might sound, although to state that it has actually succeeded in doing so certainly would be. What has been made, though, is a serious attempt to provide a succinct yet in-depth description of the production line without ignoring the descriptive and scientific foundations of the topic in hand: the result is this first volume which offers the reader an explanation of what happens during each individual stage of the production cycle, placing emphasis on the importance of correctly defining the nature of the raw materials, their significance in a ceramic body and the problems they can cause. An overview of possible products as a function of their technological characteristics is also provided. A brief introduction outlines the main subject areas, subsequently dealt with in more detail. Volume II, instead, aims to clarify the technological aspects of the individual stages of the production cycle: rather than providing a description of specific machines that would inevitably and quickly become obsolete, such clarification aims to explain the purpose and work cycle of each machine as regards the creation of the final product. A substantial part of the book is dedicated to explanations of the defects that can arise at each stage of the production cycle and there is also a clear overview of the correlation between finished tile defects and the stage of the production process in which such defect was generated. This first volume, as mentioned above, largely focuses on the manufacture of pressed tiles, yet the reader will also find facts and figures on other products and technologies where this aids understanding of reaction mechanisms. A detailed appendix of standards, tables and figures completes the volume.

9


10

Introduction

INTRODUCTION

When speaking of CERAMIC TECHNOLOGY it is important to introduce a few key definitions that will provide us with useful reference points later on. The first thing that needs to be defined is exactly what the word CERAMIC refers to. Just what is a ceramic material? While dozens of definitions are possible, it can effectively be described as any product, having a form, made up of non-metallic inorganic raw materials (whether mineral or artificial), which, from an incoherent powdery state, are transformed, via various processes, into a semi-finished item, which, through firing, becomes a solid object of partially crystalline and partially vitreous structure. Virtually all the transformations and/or mixings become permanent after firing. When speaking of inorganic raw materials it is necessary to bear in mind the level of abundance of the most common elements in the earths crust, as it will obviously be convenient and advantageous to manufacture ceramic items using the most widely available and economic raw materials. Fig. 1 (below) shows that these essentially include silicon and aluminium oxides

Fig. 1. The most common elements in the earths crust (from Kingery: Introduction to Ceramics - Wiley).

11


having different Fe, Ca, Mg, Na and K contents: these are, effectively, the 7 most common elements in nature and are always present in a ceramic body. Observing the Periodic Table (Tab. 1), PERIODIC TABLE INERT GASES

Atomic number Chemical symbol Atomic weight

it can be seen just how few ceramic-relevant elements there are (Al, Si, Ca, Mg, Fe, Ti, Na e K). Even when the main constituents of glazes (e.g. Pb, Zn, Sn, Zr, Cr, Ni, Cr, V, B etc.) are taken into account it is evident that the assortment of elements characterising a TRADITIONAL CERAMIC material is a restricted one. In keeping with the general, introductory nature of this chapter, it can be said that, where these materials are concerned, the field of study is limited to products normally made up of natural oxides such as: TILES. SANITARYWARE. TABLEWARE. BRICK. CERTAIN REFRACTORY ITEMS. Another approach is required where SPECIAL or ADVANCED ceramic materials generally made up of non-oxides or special oxides are concerned: bioceramic items, electrical and technical porcelains, electronic industry ceramics, catalysts, special refractory items etc. 12

Introduction

A better insight into the composition of a traditional ceramic material can be attained by taking a look at a standardised tile body, bearing in mind that with the appropriate modifications, similar observations can be made with regard to other traditional ceramic materials. The base body, then, is generally made up of: CLAYEY MATERIALS, which provide the plasticity needed to obtain a defined form. These include Al, Si and a proportion of Ca, Fe, Ti. FLUXING MATERIALS such as feldspars, nepheline etc., which, during firing, produce vitreous phases that act as particle-particle adhesives and promote solid-solid reactions; these contribute Na, K, Al and Si. OTHER MATERIALS such as talc, silica, pyrophyllite, CaCO3 etc. (the so-called INERT MATERIALS), used to obtain a certain type of performance: these largely contribute Ca, Mg and Si. ADDITIVES, largely employed to improve the rheology of aqueous solutions; these may be inorganic or organic and only limited amounts (< 1%) are introduced into the bodies. In summing up the main effects of each chemical component in a tile body, we can state that the presence of oxides confers: Al2O3 SiO2 Fe2O3 and TiO2 CaO and MgO K2O and Na2O

refractory characteristics and plasticity (where associated with the presence of clayey materials). structure, framework, including neo-formed phases. colour and, sometimes, fluxing properties. shrinkage control, via the formation of calcium and magnesium silicates. fluxes which form vitreous phases.

Despite this description of the ceramic characteristics that the chemical composition confers on the bodies, it is a well known fact that chemical analysis of a ceramic body or raw material is not in itself the most important aspect of product characterisation: it can easily be demonstrated that products with greatly differing performance and fields of use, such as low water-absorption floor tiles, sanitaryware items, bricks etc. all have very similar chemical composition. The quantity of Ca and Mg is significant as these will influence shrinkage during firing and subsequent water absorption. The higher the content of these elements the lower the degree of shrinkage, because calcium and magnesium silicates form; these increase in volume as temperature rises, thus opposing the shrinkage caused by the collapse of the silicate phases. Fig. 2b provides an example of the tertiary diagrams used to establish equilibrium between the various components. A good example of such behaviour is given in the subsequent tertiary diagrams (Figs. 3-8), which show compositions for product types such as majolica (low-shrinkage wall tiles), cottoforte (intermediate behaviour owing to high water absorption) and gres (high shrinkage and low water absorption). 13


Tab. 2a. Composition and characteristics of different body types.

14

Introduction

[Remember that, for a system in equilibrium, tertiary diagrams illustrate the composition of systems consisting of 3 components. They display the relative percentages accounted for by each component as shown in the diagram below (Fig. 2b): any point within the triangular area of the diagram is effectively an intersection of the lines running parallel to the sides of the triangle and the point at which these lines converge gives the percentage values of the components in the system. The percentage content of component A is read off on the left-hand side of the triangle, component B on its base and component C on its right-hand side. For example, in fig. 2b point P represents a system where component A accounts for 30% of the material (point E), B accounts for 20% (point F) and C the remaining 50% (point G)].

Fig. 2b. How to read off composition percentages on the axes of a tertiary diagram.

15


Fig. 3. SiO2/Al2O3/Fe2O3 tertiary diagram showing the compositional fields for majolica, cottoforte and red gres bodies (Vincenzini and Fiori, 1977, vol. 2).

G = red gres C = cottoforte M = majolica

Fig. 4. Fe2O3/Na2O + K2O/MgO + CaO tertiary diagram showing the compositional fields for majolica, cottoforte and red gres bodies (Vincenzini and Fiori, 1977, vol. 2).

16

Introduction

GBp = potassium white gres GBs = sodium white gres

GR = red gres

Fig. 5. SiO2/Al2O3/TiO2 + Fe2O3 + MgO + CaO + Na2O + K2O diagram showing the compositional fields for red gres and potassium and sodium white gres bodies (Fabbri and Fiori, 1983, vol. 1).

GBp = potassium white gres GBs = sodium white gres GR = red gres

Fig. 6. Al2O3/Na2O/K2 tertiary diagram showing the compositional fields for potassium and sodium white gres and red gres (Fabbri and Fiori, 1983, vol. 1).

17


GBp = potassium white gres GBs = sodium white gres

GR = red gres

Fig. 7. Al2O3/Na2O + K2O/TiO2 + Fe2O3 + MgO + CaO2 tertiary diagram showing the compositional fields for red gres and potassium and sodium white gres bodies (Fabbri and Fiori, 1983, vol. 1).

1 red gres clays 2 cottoforte clays 3 majolica clays

Fig. 8. SiO2/Al2O3 + TiO/Fe2O3 + MgO + CaO + Na2O + K2O tertiary diagram showing the compositional fields for majolica, cottoforte and red gres clays (Sandrolini and Palmonari, 1974, vol. 2).

18

Introduction

To provide an effective description of a raw material or a ceramic body and give information on its technological behaviour it is much more convenient to make use of mineralogical (or crystal) analysis and particle size analysis. To understand the usefulness of these techniques it is necessary to bear in mind how a clayey raw material acts within a body, giving it plasticity during the production process. In relatively simple terms, it can be said that a clay is made up of an association of tetrahedral SiO4 repetition units (T symbol) associated with octahedral units having an Al(OH)6 base (O symbol) as illustrated in Figs. 9-10.

Oxygen Silicon Aluminium, Magnesium etc.

Silicon Oxygen

Fig. 9. Phyllosilicate repetition units: SiO4-4 tetrahedrons and Al(OH)63- octahedrons.

Fig. 10. Two-dimensional projection of octahedral structural units.

These form two-dimensional particles, the distance between two identical structural repetition units varying as a function of the type of material, as illustrated in Fig. 11. The interlayer space between two particles, micelles or two strings of particles etc. may, by way of this repetition unit arrangement, vary from a minimum of 2.7 (in kaolinite) to a maximum of 8 (in chlorites) or more Angstroms [1Å = 10-8 cm], thus determining whether or not extraneous molecules or ions can infiltrate the structure. 19


(A) Na-montmorillonite (B) Ca-montmorillonite

(A) Vermiculite (B) Tri-octahedral chlorite

(A) Muscovite (B) Illite

Fig. 11. Examples of the different inter-lattice spacings that characterise clayey raw materials.

20

Introduction

Where water lies between the particles they are able to slide against each other more easily (i.e. thus conferring plasticity). Such spaces may also be occupied by fluxing ions such as Na+ and K+, which modify the technological properties of the raw material. For this reason it is highly important that the mineralogical nature of the clayey contents of a body be known so that the advantages and disadvantages of their employment can be identified: this information is obtained using X-ray diffractometry (XRD) techniques. In speaking of fluxing materials, mainly feldspars, an understanding of their mineralogical nature is once again important as the formation temperature of the vitreous phase and the viscosity of the formed glass correlate closely with the type of feldspar: sodium feldspars (mineralogically defined as albite) have low melting points and low molten material viscosity too, while potassium feldspars (microcline, orthoclase, sanidine etc.) have high viscosity, which can help in the event of sizing and sticking problems during firing. Furthermore, clear information as to the mineralogical nature of the individual raw materials of a body helps with the planning of special characteristics, such as the formation of eutectics (i.e. special compositions with lower melting points). Particle size distribution is another key factor: it is obvious that the purpose of forming and firing a semi-finished ceramic item is to obtain a product in which solid-solid reactions have been activated and completed as far as possible. In this regard results are greatly influenced by the contact surface area of the particles: a greater contact surface area is more conducive to the heat-induced passage from sintering to reaction to fusion (see Fig. 12).

Fig. 12. Schematic illustration of the sintering process.

While the clayey particles are themselves small, they need to be mixed with particles of other appropriately sized materials so as to plug as many gaps as possible and thus obtain maximum density: this is achieved via proper combination of several grain sizes. Therefore, proper control of body particle size distribution cannot be achieved 21


via simple sieve residue alone: checks should be run using appropriate instrumentation (X-ray or laser etc.) that takes into account the interaction principles between the individual particles (diffraction, scattering etc.). The production process Whatever the employed ceramic body raw material selection method, the subsequent production process will invariably be as follows: IN-QUARRY selection, mining and quality control. PREPARATION of raw materials for mixing. BODY preparation via GRINDING. FORMING. DRYING. Additional procedures to add AESTHETIC value to the product. FIRING. SORTING, PACKAGING and STORAGE. Each of these stages requires attention and must be planned and executed carefully: constant quality control is essential. There follows a stage-by-stage summary of the most common tile manufacturing processes: BODY PREPARATION CRUSHER

dry wet

Hammer mill Pendulum mill Alsing (ball) mill Continuous mill Fast agitating disperser

FORMING

Dry pressing Extrusion Casting Decoration with multiple press fillings (where applicable)

DRYING

by convection, slow or fast, by radiation

AESTHETIC PROCEDURES

generally glazing or special applications

FIRING

Traditional (slow) or rapid: for both Single firing (body+glaze) Double firing (glaze on already-fired biscuit)

22

Introduction

The following flow-charts (Tab. 3-4) summarise the most common production processes. raw material storage raw material storage

ê

ê

batching

batching

ê

ê

wet grinding

wet or dry grinding

í

ê

spray drying or wetting

î

continuous mill

ê

discontinuous mill

î

pressing

í spray drying

ê drying

ê

ê

pressing

biscuit firing

ê

ê

drying

glazing

ê

ê

glazing

glaze firing

ê

ê

firing

sorting

ê sorting

Tab. 3-4. Different ceramic tile production processes.

23


If every stage of the production process has been well planned, properly-made tiles will exit the plant: yet the aesthetic and performance requisites required of tiles in airport halls differ enormously from those intended for use in apartment bathrooms or on industrial flooring. Hence the concept of tile classification, summed up as follows. Tiles are generally classified on the basis of: INTERNATIONAL STANDARDS mainly based on the type of production process or water absorption of the fired product. and, even today COMMERCIAL TRADITIONS

that still use old names such as Gres, Majolica, Klinker etc.

The correct approach, of course, would be to take into consideration a complete set of FIRED TILE TECHNOLOGICAL CHARACTERISTICS, the key ones being: Intended use (floor/wall tile, indoor/outdoor). Water absorption, but also resistance to freezing/thawing cycles. Shrinkage. Bending strength. Resistance to abrasion and staining. Body colour. Slip resistance. Comprehensive knowledge of all these parameters will lead to proper classification of a ceramic tile, thus allowing it to be employed appropriately. This concludes our summary as to the nature of the subject of this volume. The following chapters feature more in-depth elucidation, indicating just what sort of analysis methods can and should be used and offering a systematic description of each raw material and its behaviour during the various stages of the production process.

24

Introduction

Chapter I IDENTIFICATION AND CHARACTERISATION OF CERAMIC RAW MATERIALS

Many properties of clays and other ceramic raw materials depend on the type and quantity of the various minerals that constitute them. The identification of such minerals is therefore of primary importance. Yet clear-cut identification is made difficult by the fact that ceramic raw materials rarely consist of pure, well crystallised minerals: instead, several minerals are generally present in appreciable quantities together with many other minor constituents. Given these circumstances, identification of the main phases can be difficult, especially where they are all rather similar. Sometimes, and this is often the case with clays, a mineral cannot be identified without prior purification and separation. Furthermore, a single clay may contain various minerals and is nearly always associated with significant quantities of quartz, calcareous materials, micas and others. Also, clayey minerals are made up of very small particles (dimensions as small as 100 Å, that is 10-6 cm, are frequent), making identification even more problematic. Moreover, clayey raw materials are often characterised by isomorphic substitution, created by the genesis conditions described in the following chapter. In general, then, analysis methods used for the study of these raw materials must allow for recognition of minerals of non-constant, often intermingled composition sometimes made up of extremely small particles. Since identification of a mineral depends on its key characteristics, which must necessarily always be the same independently of mine attitude and the surrounding environment, analytical methods which make use of the univocal properties of the individual raw material classes must be employed. Such properties can be summed up as follows: properties dependent on the chemical nature of the material properties dependent on the crystalline aspect of the mineral properties dependent on the atomic or ionic arrangement in the crystalline structure properties dependent on chemical or physical modifications in the mineral caused by external factors (e.g. enthalpy variations induced by heating or cooling). Other practical methods for identification or estimation in raw material mineral content exist: for example, when rheological properties such as plasticity, thixotropy etc. are particularly evident, this suggests the presence of certain clayey minerals. Similarly, magnetic properties can indicate the presence of ferromagnetic minerals etc. While these methods normally provide an overall picture as to the characteris25


tics of the predominant mineral in the mix, they do not identify its individual components. An accurate description of the analytical methods used in raw material mineral content recognition is outside the scope of this publication. Nevertheless, there follows an approximate list of the main techniques and the criteria on which they are based because, in the following description of the individual mineral types, reference will frequently be made to characteristic analytical data. Sampling Good analysis, whatever the type, where effected on a sample made up of a mix of base components, and especially where that mix is heterogeneous, will require proper preliminary sampling; this allows the necessarily small quantity of material used in analysis to represent what can amount to tons of raw materials stored in the body preparation department of a ceramic company. The analysed sample, then, must be representative of all the material in question, not just a portion of it. Beginning with the most general case (i.e. in-quarry sampling, for which there exist specially developed standardised procedures), samples must be taken from various points and different depths along the extraction front; if the material appears to be uniform, selection and conservation of a sample representing about 1% of the total sample may be effected (following mixing and quartering of the taken samples). Where the material is less uniform up to 5% of the taken sample may have to be selected in order to obtain a reasonably analysable sample. The same procedure applies to shiploads or truckloads of loose material. The selected preliminary sample is then homogenised further, then quartered again via subsequent formation of flattened heaps and the taking of opposite quarters, until a final coarse sample of some 10-12 Kg is obtained. This will be ground and quartered again until a suitable sample of about 3 Kg is obtained. The latter should have an average particle size distribution no greater than 0.5-2 mm and should not be over-ground so as to prevent oxidation of components sensitive to contact with air; for similar reasons, in-sample moisture content should not be completely eliminated during comminution and quartering, and should remain within the 4-12% range so as to prevent alteration or loss of soluble salts. Final drying and grinding will only be effected immediately prior to analysis as a function of the requirements of the analysis itself. The type of comminution can be critical as there is a risk of contamination from the grinding machines themselves especially where metallic or seriously altering the structure of certain minerals (e.g. those with a two-dimensional conformation).

26

Identification and characterisation of ceramic raw materials

Chemical analysis Determination of the type and quantity of each element in a sample is of great interest to ceramists. Yet in itself, such analysis often carries little significance. The presence or absence of certain elements in certain quantities may certainly help determine behaviour during firing but rarely provides information on the workability of the raw material in question (i.e. in grinding, spray drying, pressing, drying) or its physical behaviour. Nevertheless, properly performed elementary analysis is an exact science while many other techniques for the analysis and determination of technological properties are marred by error and uncertainty. Furthermore, the combination of structural data with constituent data allows for semi-quantitative identification of the minerals in the sample. A whole host of chemical analysis techniques exist. In particular, modern instrumental techniques provide fast, accurate results, the precision of which largely depends on proper selection and preparation of the sample, a must for the attainment of complete homogenisation. In chemically analysing any substance, results are normally expressed in oxides and, as specifically regards ceramic body materials, the main 8 are as follows: silicon dioxide (Silica SiO2), aluminum oxide (Alumina Al2O3), titanium dioxide TiO2, ferric oxide Fe2O3, calcium oxide CaO, magnesium oxide MgO, sodium monoxide Na2O and potassium oxide K2O. Oxides volatile at 1000 °C (carbon dioxide or carbon monoxide, sulphur oxides such as SO3 and SO2, together with water) are normally indicated as Loss of Ignition (L.O.I.). Accurate analysis will rarely see these components account for 100% of the sample as other elements generally accounting for less than 1% are always present in varying quantities. These include Barium, Strontium and other transition metals such as Copper, Chromium, Manganese and Boron, Lithium etc. This type of analysis, however, does not indicate how the various elements are combined and this can lead to errors in their technological evaluation: to understand this concept just consider the difference between a calcium oxide originating from feldspar rather than lime, or sulphur oxides that fail to take into account the presence of mineral sulphides (e.g. Pyrite). Before beginning a chemical analysis, then, it is necessary to select the sample carefully as per the above methods. Such samples will, after proper drying at temperatures that do not alter the volatile substance content, normally have a weight ranging from a few hundred milligrams up to 1-2 grams; subsequent grinding is effected using high-efficiency milling machines that do not contaminate the sample. Methods vary depending on the hardness of the sample, ranging from corundum or, even better, agate pestle-and-mortar tools to micro-mills with special hard alloy grinding media. With the sample weighed (and this is the point at which errors can creep into the entire analysis), a solubilisation method which allows complete homogenisation must be decided on: the sample is usually dissolved in chemical reagents to obtain a ho27


mogeneous liquid solution, or is solubilised in a molten state in a special glass with the resultant solid solution being analysed. Ceramic materials, unfortunately, being silicate, aluminate and oxide-based, are difficult to solubilise. There exists a vast argument-specific literature indicating hydrofluoric acid HF and other mineral acids, such as nitric acid, HNO3, hydrochloric acid HCl or sulphuric acid H2SO4 as appropriate hot liquid solubilisation agents requiring, as might be expected, the use of special containers. Classical wet-type chemical analysis involved complex, systematic treatment of samples to separate individual components prior to analysis true and proper, mainly effected using (lengthy and complex) gravimetric, colorimetric or complex metric methods (necessarily preceded by accurate preliminary calibration). These methods, while still valid today, have undoubtedly been superseded by the advent of more sophisticated instrumentation capable of extracting immediate analytical results from both solubilised and untreated samples alike. All instrumental information is ultimately influenced by the initial measurement of the mass of the sample and its proper preparation, which, whatever the method, must provide conditions that are as standardised and homogeneous as possible. The main aggression and acid solubilisation methods effected on ceramic raw materials are: Hot acid aggression in open containers using HCl, HNO3 and HClO4-based acidoxidant mixes: the need to break up the silicate matrix nearly always makes the use of HF indispensable too, thus making it impossible to use standard borosilicate glass. The use of open containers together with high-temperature solubilising acids aids the loss of volatile components. Hot acid aggression in closed containers (and thus at high pressure). These systems, now becoming more widespread owing to the rapidity of aggression and dissolution, usually employ Teflon containers and programmable microwave heating systems. Alkaline fusion and subsequent acid solubilisation, normally with HCl. There exists a vast range of alkaline fluxes, to be employed as a function of the desired fusion temperature and process efficacy: in all cases, of course, there is the addition, via the flux, of at least one cation (an aspect difficult to quantify in the unknown specimen). The most widely employed alkaline fluxes are NaKCO3, NaOH, LiBO2 and Li2B4O7, with various salts being added as detachment agents (Lithium halogenides or alkalines in general, mostly complexing agents etc. - see table on following page). Fusion may be manual or automatic, thus guaranteeing standardised conditions (always in platinum or similar containers). The obtained glass can undergo instrumental analysis directly or can be accurately solubilised to provide, following proper dilution, an analytic solution.

28


Melting temperatures of compounds used in fusion-type decomposition of materials. Compound

Compound

Compound

(1) After decomposition and transformation of bisulphate in pyrosulphate K2S2O7. (2) Normally introduced in the dihydrate form. (3) Like PbO after decomposition and outflow of carbon dioxide at 315 °C. (4) Like PbO after decomposition at 500 °C approx.

Whatever the employed instrumental technique, the preventive establishment of various calibration curves will be required via the utilisation of standard solutions or solids within the presumed analysis range for the new samples. The main methods used in quantitative ceramic material chemical analysis are based on interaction (of the fluorescent emittance, absorption or emission type) of the sample with electromagnetic radiation. The main methods are: X-RAY FLUORESCENCE (XRF), in which raw minerals or, preferably, minerals finely dispersed in an alkaline glass, are bombarded by high-frequency short wavelength radiation. The energy in the radiation induces fluorescent emission owing to the excitation of inner electrons in the orbitals of the elements; these sampleemitted electrons are collected in a special detector. The generated signal is associated with the position of the sample or the detector itself so as to identify the relative intensity of the signal that can be compared with a standard one. These methods allow for easy quantification of medium-high atomic weight elements as far as the lower Na - F limit. More recently, considerable efforts have 29


been made to extend sufficiently repeatable element assessment as far as Boron (Fig. 13). ABSORPTION SPECTROSCOPY (AAS-GFAAS). This technique exploits the energy absorption caused by the presence of atoms placed in the path of one or more monochromatic radiation beams generated by special lamps: in practice the solution to be analysed is injected into a flame, the temperature, geometry and composition of which ensures that the elements are present in atomic (not ionic) form and provides maximum interaction with the incident radiation. To obtain more accurate resolution it is sometimes possible to use fast induction heating as the atomisation energy source instead of a flame: this is effected in a tube made of graphite or other suitable material in an inert gas flow. Under specific conditions this technique can even be used for direct analysis of easy-volatilisation solid samples; it is also a technique that gives excellent analytical results for any metallic element and can provide superb resolution, in the order of, depending on the element, fractions of parts per million (Fig. 14). INDUCTIVELY COUPLED PLASMA-ATOMIC EMISSION SPECTROSCOPY (ICP-AES). Similar to the previous technique, but measures interaction with the incident radiation in terms of emission as opposed to absorption. In this case the sample solution is atomised by the combined action of a standard high temperature torch and application of a radio frequency source. With respect to absorption systems, this technique is advantageous in that it is possible to effect sequential analysis of each sample without having to modify the source. Detection limits for each element are generally poorer, yet results are actually better with some elements. In any case it is possible to switch from p.p.m. to p.p.b. fairly easily (Fig. 15). Other chemical analysis methods pertinent to the search for specific elements (e.g. FLAME PHOTOMETRY in tracking down alkaline elements) do, of course, exist, but a detailed account of these is beyond the scope of this book. It is, instead, worthwhile pointing out that specific chemical tests can be used to check for the presence of specific elements (carbon, sulphur, fluorine...) or anions (CO32- carbonates, SO42- sulphates): these employ analytical techniques that are generally simple and efficacious and provide extremely important information for evaluating the applicability of a raw material in a ceramic process.

30


Fig. 13a. Fig. 13-13a. Operating principle of X-ray fluorescence (XRF) chemical analysis.

Fig. 14. Operating principle of atomic absorption spectroscopy (AAS).

Optical Systems Generator

Torch (ICP)

Pump/ Nebuliser

SPECTROMERC

SPECTRO AS 300

SPECTROSONIC

Fig. 15. Operating principle of inductively coupled plasma-atomic emission spectroscopy (ICP).

31


Mineralogical (or crystallographic) analysis This type of analysis allows the technician to check for the presence of individual crystal phases in a sample and thus trace its mineral composition, the evaluation of which is of primary importance in defining the technological characteristics of a raw material or the contribution it will make to a body. A preliminary form of such mineralogical information is provided first by naked-eye and then microscopic inspection under reflected, polarised and transmitted light (on naked samples included in the resin): by combining the cited observations with other optical properties (e.g. refractive index) it is possible to obtain good discrimination and recognition results. In the mineralogical analysis of ceramic materials the entire science of optical microscopy should be viewed as a branch in its own respect consisting of numerous techniques (as illustrated by the abundant literature) that achieved maximum diffusion between the 1940s and 60s.The most widespread mineralogical analysis technique now in use is X-RAY DIFFRACTOMETRY (XRD). This can be effected on individual crystals or, more commonly, on powders (Fig. 16). The incident X-radiation falling on the sample, properly filtered so as to confer monochromaticity, interacts with the crystal lattice, giving rise to related diffraction images, via the equation Bragg nλ = 2dsinϑ, at the inter-lattice distance d of the crystal as a function of the diffraction angle. Since a wide range of X-ray sources can be used (the most common is provided by Ni-filtered Cu anticathodes emitting CuKα = 1.541 Å only) it is correct to list every interaction set of a particular crystal lattice with the list of the active lattice distances expressed in Angstroms (Å). Indexes and data bases are available for all crystalline substances (PDF, see the examples in tables 5 and 6). These are updated by an international controlling body. This reference material allows recognition of crystal types in natural and artificial samples, while recent software improvements now provide more modal management of diffractometry data, providing quantitative evaluations with regard to the presence of individual minerals. Of course, when working with powders it is essential that the sample be highly representative of the whole. There should be no preferential orientation, easily caused by bi-directionally developed crystals: this is why preparation and laying out of the sample are so important. Milling must be carried out as efficiently as possible, taking care not to alter the structural characteristics of the sample, especially where clayey, so as to aid homogenisation of all the phases and ensure casual orientation of all crystalline faces; where analysis with a spinning sample holder is excluded (such equipment is usually unavailable), one must be reasonably certain that a strongly non-oriented sample has been prepared, even in the concomitant presence of phases having different density and inappropriate crystallographic habit, such as clays and micaceous materials having micelles of a well-developed bi-directional form. X-ray diffractometry analysis is a fairly routine procedure. It is relatively quick 32


Fig. 16. X-ray diffractometry analysis equipment (left) and the operating principle on which it is based (top).

and simple and, on sufficiently simple matrices, leads to fairly easy interpretation of the gathered data, allowing moreover through combination with the verified presence of certain crystal phases, attribution of a chemical formula and combination with quantitative chemical analysis so-called rational analysis of a raw material or body (i.e. approximation of its composition expressed in standard minerals). It will therefore be possible to evaluate a ceramic body not only in terms of its oxide-based chemical composition, but also its mineralogical composition, expressed in Quartz, Kaolinite, Illite, Calcite, Dolomite, Albite (Sodium feldspar), Microcline (Potassium feldspar) etc. (see Fig. 17 for example). Thermal analyses As the heading suggests, there is more than one way of analysing samples via the parameter variations that occur as a function of an increase (or decrease) in temperature. Variations in weight, temperature, heat evolved or absorbed, dimensions, gaseous substances emitted etc. can all be registered as a function of a particular temperature gradient. Each of these techniques has a corresponding type of analysis, as listed below: TGA (ThermoGravimetricAnalysis), 33


Tab. 5. PDF chart containing crystallographic data for quartz (SiO2).

Tab. 6. PDF chart containing crystallographic data for kaolinite.

34


Porous single firing, coloured body dry grinding, pressing

Fig. 17. Example of body chart with rational XRF + XRD analysis.

DTA (DifferentialThermalAnalysis), DSC (DifferentialScanningCalorimetry), TMA (ThermoMechanicalAnalysis) or DIL (Dilatometry), EGA (EvolvedGasAnalysis). These types of analysis are not only an obvious method for observing and predicting the behaviour of an individual raw material or a body during drying, firing or cooling: they are also an excellent aid in determining the mineralogical composition of a material, as the observed effects closely correlate with the crystalline structure and the phase transformations of various minerals. Since the presence of minerals and salts such as dolomite, carbonates, sulphates, sulphides and fluorides etc. (Figs. 18-22) can be recognised, such analyses can also help identify chemical parameters. The instrumentation used for thermal measurements essentially consists of a measuring head (which houses the appropriate physical form of the sample and transforms variations in such sample into amplifiable, manageable electrical signals) and a heating system. The latter generally uses electrical elements which must 35


Fig. 18. Comparison between the most significant (appropriately schematised) DTA curves representing clayey minerals: A = kaolinite; B = metahalloysite; C = Na-montmorillonite; D = Ca-montmorillonite; E = vermiculite; F = sepiolite; G = polygorskite.

Fig. 19. Differential Thermal Analysis (by G. Peco 1952). C = kaolinite; Cr = cristobalite; H = halloysite; Q = quartz; I = illite; Ca = limestone; M = montmorillonite. TG, DTG, DTA and ETA analysis on Hirshau kaolin (by E. Kaisersberger and C. Urso).

36


Fig. 20. Examples of dilatometric curves. Unfired samples undergo irreversible transformation (curve a) and fired ones, in cooling, essentially return along the curve followed during heating (curve b).

be extremely stable, homogeneous and programmable so as to ensure perfect repeatability of measurement. It is often possible to combine several such analyses with just one instrument (e.g. TGA + DTA, TGA + DSC, DTA + EGA or TGA +DTA + EGA), allowing the user to obtain a wealth of information from just one measurement. To provide the reader with a short summary of the operating principle behind some of these instruments, there follows a description of those most commonly used in the ceramic industry, beginning with the dilatometer and simultaneous TGA + DTA analysis. 37


Fig. 21. Dilatation-Shrinkage (DS) and Differential Thermal Analysis (DTA) of secondary minerals (by G. Peco 1970) Ca = limestone; Q = quartz; Cr = cristobalite; F = feldspar.

Fig. 22. Dilatation-Shrinkage (DS) and Differential Thermal Analysis (DTA) of standard clayey minerals (by G. Peco 1970). C = kaolinite; H = halloysite; M = montmorillonite; Mi = mica.

38


Dilatometric analysis measures the expansion and shrinkage of an appropriately pre-formed sample, usually parallelepiped or cylinder-shaped, and approximately 10 (∅) x 35-45 mm in size (unfired, pre-fired or fired). Simultaneously, elongation of the measuring system and support must be compensated for: the latter is usually made of quartz, for measurements of up to 950-1000 °C, or alumina (up to 1350 °C) owing to their limited expansion. The heat will obviously cause all the materials, especially the clayey ones, to dilate and shrink as a function of that which occurs in their interior. More specifically, the loss of volatile components (especially water) induces a collapse of the crystalline structure of the clays which is detected as intense shrinkage at specific temperatures. It is therefore possible, by observing the contraction (or expansion) characteristics over temperatures up to or even beyond the normal firing range, to distinguish between free or combined waters in the lattice. It is also possible to identify critical temperatures where structural alterations are occurring and the speed of such changes. Transitions such as α to β quartz and α to β cristobalite and other reactions both reversible and irreversible can be noted and their potential influence on the heating processes during manufacture considered. The dilatometric analysis of standard materials illustrated in figures 20-21-22, is significantly representative of this possibility. There is also another and highly important use for dilatometric measurements effected on ceramic raw materials and bodies: the measurement of a tendency of a material to expand or shrink within a certain temperature interval (i.e. its linear thermal expansion coefficient). This is expressed by the following formula: αT1-T2 = ∆L / (L0 × ∆T) where α is the linear thermal expansion coefficient between T1 and T2, L0 is the initial length of the sample and ∆L is the difference between LT1 and LT2. It is, however, consolidated practice in a factory to use, instead, the cubic coefficient. This is obtained by multiplying the linear coefficient obtained from dilatometric measurements by 3, on the arbitrary assumption that the propensity towards dimensional variation is the same in all directions. This data is extremely important in evaluating compatibility between different materials during heating and, more importantly, during cooling, where any tension that has built up as a result of dilatometric incompatibility becomes permanent. Firing of a glaze that has been applied to a ceramic body is a classic example (Fig. 23). In the light of the above it is clear that there exists the opportunity or, rather, the need, to effect a series of fine analytical checks on raw materials used in the ceramic production process: most companies do not have the relevant instruments and thus checks are generally outsourced to laboratories or external bodies with the equipment for accurate examination.

39


Fig. 23. Opposing effects of dilatometric incompatibility between glaze and body in the cooling part of the cycle: case A αglaze > αbiscuit - case B αglaze < αbiscuit.

Fig. 24. Dilatometric apparatus.

40


However, in-company organisation of production checks will allow work to be carried out with continuity and homogeneity and is thus of great importance. The greater the accuracy and frequency of such checks, the greater the chances of promptly correcting errors or improper settings; the more punctual and orderly the recording, interpretation and filing of such results, the easier it will be to recommence and optimise the manufacture of a given product line at some later date. For these reasons, manufacturers would be advised to carry out the following checks. Plastic, semi-plastic and non-plastic raw materials Determination of the characteristics of incoming material via measurement of: Moisture content Size and appearance Residue Carbonates Reducing agents Moisture content expressed in % A sample of wet material is weighed (Pu), then dried to constant weight in a laboratory drier at 110 °C for about 24 h. Its dry weight (Ps) is then measured. The difference between the two weights divided by the weight of the wet sample multiplied by 100 gives the moisture content of the sample: Pu - Ps u% = _______ x 100 Pu Appearance of incoming material Form and dimensions (piece size) of the incoming material should be noted, as should colour and all other key characteristics. Sieve residues One hundred grams of dry sample is weighed and treated in a laboratory agitator for about 30 minutes with 300 ml of de-ionised water to which 1g of sodium tripolyphosphate or sodium polyacrylate is added so that the particles are completely broken up and dispersed. The slip is sieved at 1,000, 2,500 and 10,000 mesh/cm2 (180, 125 and 60 µm screens).

41


Fig. 25. Apparatus for determination of the various particle size distribution bands.

The residue obtained in the various sieves is dried to constant weight and then weighed. The resulting values directly provide residue percentages for the various mesh sizes. With non-plastic raw materials it may be interesting to carry out dry measurement by sieving about 200 g of dried material through a sieve column of appropriate mesh sizes [e.g. 11(2,000) - 35(1,000) - 140(500) - 590(250) - 1,600(150) - 4,700(90) and 9,500(63) mesh/cm2 (mm)]. Carbonate content expressed in % CaCO3 Measurement is carried out using a calcimeter (e.g. Pizzarelli or Dietrich-Fruhling). One gram of sample material dried at 110 °C is placed in the container. Diluted hydrochloric acid (1:1) is introduced into the container via a tap. If the material contains carbonates of any kind CO2 will be released, which, by exerting a pressure proportional to the quantity of carbonates in the sample, lowers the water column in the graduated cylinder: from here, the percentage of carbonates in CaCO3 form contained in the raw material can be read off directly or calculated. Once the released volume of CO2 and the absolute temperature are known, the following calculation can be made: % CO2 = 53621 × VCO2 (litres) / °K

42


Fig. 26. Apparatus for the determination of carbonate minerals (calcite and dolomite).

Assuming that CO2 has been generated by CaCO3 only: % CaCO3 =% CO2 . 2.273. Determination of total reducing substances Effected via RedOx back-titration. Approximately 2 g of powder < 200 µm are transferred into a 400-500 ml beaker to which 10 ml of a K2Cr2O7 1 N solution (49 g in one litre of distilled water) and 10 ml of concentrated H2SO4 are added. The container is then hand-shaken for about one minute, then left on a slow magnetic stirrer for 30 minutes (the colour should remain yellow-orange: if it is green another 10 ml of dichromate is added). Then add 10 ml of phosphoric acid 85% to complex the iron which is present and dilute with 200-250 ml of distilled water; add approximately 2 ml of indicator solution (0.16 g of diphenylaminosulphonate of Barium in 100 ml of distilled water), and titrate with FeSO4 0.5 N (140 g in 200 ml in distilled water + 40 ml of H2SO4 concentrate, made up to a litre and stored in a dark glass container) under slow agitation. Becoming bluer and bluer, it will change to light green at the equivalence point. The obtained value should be corrected in line with the effective reducing oxide ratio of the reagents, obtained by effecting the same titration without the sample. Formula: ml of reduced K2Cr2O7 × 0.69 (stoichiometric factor) ___________________________________________ = total reducers % sample weight (g) 43


Example: Blank control test: 10.0 ml of K2Cr2O7 added, 21.5 ml of FeSO4 necessary for end point without clay: correction factor = 0.465 (10/21.5). Measurement test: 2.125 g of sample + 10 ml K2Cr2O7, 19.25 ml of FeSO4 necessary for end point: 10 - (19.25 x 0.465) × 0.69 /2.125 = 0.34% total reducers. This measurement essentially detects the organic substances in the sample, but is also influenced by the presence of any reducing substance, especially mineral sulphides such as Pyrite, Marcasite etc.

44


Chapter II ROCKS

The rock cycle An understanding of rocks may be achieved by classifying them according to how they are formed, giving us: Igneous rocks (intrusive and extrusive). Sedimentary rocks. Metamorphic rocks. The term rock refers to a natural aggregate of one or more minerals. The term mineral refers to a homogeneous natural solid with a well defined chemical composition and its own crystalline structure. The interconnected processes that lead to rock formation constitute what is known as the rock cycle. The earth is essentially made up of a series of concentric layers - the Crust, the Mantle and the Core.

Cross section of the Earth Crust Mantle Molten outer core Solid inner core

Fig. 27. Simplified cross-section of the Earths interior.

45


A part of the mantle (the external part) is molten as a result of the high temperatures and pressures at such depths; if this part of the mantle shifts to areas where temperature and pressure are lower (i.e. towards the earths crust) it may solidify, thus giving rise to intrusive rock. Where the molten mass (MAGMA) reaches the surface, it cools much more rapidly, giving rise to extrusive rocks. Intrusive rocks may subsequently be brought to the surface by forces of tectonic origin. Degradation processes effected by both atmosphere and hydrosphere continually alter the surface of the earths crust: for example, erosion may dismantle existing igneous rock of both intrusive and extrusive origin. Such surface degradation gives rise to sedimentary rocks. Aqueous solutions can lead to precipitation of salts or supply materials to organisms. allowing them to form shells or skeletons. Granules produced by the erosion of pre-existing rock, salts separated from solutions by the activity of organisms and salts which precipitate directly from the solutions themselves give rise to an accumulation of sediments. The latter are compressed and permeated by other layers forming above them and thus undergo that transformation (DIAGENESIS) which leads to the formation of sedimentary rocks. These continue to be buried by later deposits and are consequently shifted to greater and greater depths. However, tectonic forces may bring them back to the surface where they are again exposed to degradation. Alternatively, movements in the earths crust may bury rocks (sedimentary, extrusive, intrusive) even deeper; beyond certain depths, the changes in pressure and temperature cause profound modifications. This is where metamorphic rocks are formed. Like the others, metamorphic rocks too can be brought to the surface and subject to degradation, thus resulting in new layers of sedimentary rock. Alternatively, rocks may be carried to extreme depths (or close to very high temperature zones) where they melt and give rise to new magma of a composition different from that of the original magma.

Volcano Atmosphere

Sea

Plutons

Sedimentary Rocks

Magma

Metamorphic Rocks

Fig. 28. The magma-rock-magma cycle.

46

Rocks

The mantle-fed magma can arrive at the surface rapidly via volcanoes (2) or consolidate within the earths crust (1). In this case tectonic forces (3) may brings plutons (igneous intrusions) to the surface where the atmosphere and hydrosphere cause degradation (4). The eroded materials are carried away by rivers and streams into depository basins (5). This is where accumulation begins (6), transforming the sediments into sedimentary rock (7). These may be returned to the surface (8) where the destruction-accumulation pattern begins anew, or may sink deeper (9) where they are transformed into metamorphic rocks. Similarly, intrusive rocks (1) can be transformed into metamorphic rocks (9) by variations in temperature and pressure. If pressure and temperature conditions are particularly intense the metamorphic rocks will fuse and give rise to new magma (11). The black arrows indicate tectonic forces. Magma Magma is an essentially silicate, high temperature solution in a partially or totally molten state. It also contains considerable quantities of volatile elements (i.e. gases), which directly influence both the physical behaviour of the magma and its composition. The most important of these are: water vapour, carbon dioxide (CO2), carbon monoxide (CO), hydrogen, ammonia (NH3), hydrochloric acid (HCl), hydrogen sulphide (H2S), sulphur dioxide (SO2) and sulphuric trioxide (SO3); these gases are part primary and part secondary (i.e. originally present in the magma or introduced by contact between the magma and the overlying rock). Cooling~consolidation of the magma forms magmatic (or igneous) rock. The key characteristic of these rocks is that they are made up of crystals of varying size and form, visible to the naked eye or with the aid of magnifying glasses or microscopes. Rocks of magmatic origin are of a varying chemical and mineralogical structure in which the main components are quartz, feldspars (orthoclase and plagioclase), micas (muscovite and biotite), amphiboles, pyroxenes and olivines. Clays and rocks Clays, materials of prime importance to the ceramic industry, are extremely variable rocks made up of different minerals, generally of a silicate nature, sedimented in the form of platelets, fibres, belts etc., and are, for the most part, associated with other minerals of detrital origin that often play a key role. Variety is vast, since the combinations of initial rock types, morphological and climatic conditions, types of erosion and alteration, transport and sedimentation conditions, state and duration of attitude are virtually incalculable. To render the concept of genesis clearer, consider a rise or any zone exposed to atmospheric agents: erosion is caused by the combined action of rain, wind and the freeze-thaw cycle, resulting in material that accumulates in the form of soil or is washed far away by surface waters, giving rise to sedimentary basins. 47


Subsequently, through orogenetic or tectonic movements or simple accumulation of successive layers, the accumulated materials will, over geological time spans, form a hollow and consequently be subject to enormous pressures and profound morphological modifications, being turned to stone. Then, that stone undergoes diagenesis (that set of phenomena that modifies a sediment, giving rise to greater coherency). At a more advanced stage of transformation there is cationic substitution on the crystal lattices and further increases of pressure and temperature may lead to further formation of new rock (metamorphism). Finally, further orogenetic or tectonic movements may bring these completely modified rocks to the surface. In the processes of alteration and erosion of rock, then, tectonics play a key role, causing a series of movements/displacements of rock formations and stratigraphic units. Tangential forces cause corrugation while vertical movements result in the formation of new hollows in which accumulation of sediments can once again begin. Primary rocks Since characterisation and understanding of the clayey formations found in natural deposits within the earths crust are of enormous interest, it is worthwhile providing an introduction to the primary rocks that give rise to them. In the field of petrography there are three main rock categories: Consolidated rocks of magmatic origin Also known as igneous rocks, these derive from solidification of volcanic magma, a process that occurs: a) deep underground: plutonic or intrusive b) on the surface: volcanic or effusive. These two magmatic rock types respectively give rise to basalts and granites, which rarely come together in hybrid form. Sedimentary rocks Produced by the consolidation of sediments, solid particles suspended in fluids (usually water) that transport them considerable distances from the site of erosion. These particles settle from the fluid owing to their greater density. The complete formation process, then, involves the alteration phase of the original materials followed by erosion, transport and, finally, depositing, the final result of which is greatly influenced by the size of the particles and the pressures/temperatures exerted on the sedimentary layer by subsequent geological events in the zone, any intervening chemical reaction etc. Among the sedimentary rocks, the most common minerals are clays, quartz and calcite; it is important to distinguish: a) Detrital rocks (argillites, pelites, conglomerates) 48

Rocks

b) Organogenic rocks (limestone, marble, dolomite) c) Precipitation-derived rocks (evaporites, chalk). It is also important to point out that further classification is often effected according to dimensional criteria. Such criteria give rise to different classifications of well defined terminology such as those illustrated in the table below: ROCKS

CONSTITUENTS

mm

Coherent

Incoherent

RUDITES

GRAVELS

Mass Blocks Pebbles Granules

> 300 64 4 2

ARENITES

SANDS

Very coarse Coarse Medium Fine Very fine

1 0.500 (= 500 µm) 0.250 (250) 0.125 (125) 0.063 (63)

LUTITES

SILTS

Coarse Medium Fine Very fine Clay particles

0.0315 (31.5) 0.0156 (15.6) 0.0078 (7.8) 0.0039 (3.9) < 4 µm

Metamorphic rocks These are formed by chemical-physical solid state reactions caused by shifts in the earths crust, such as, for example, in the sinking of surface rock masses. In general, the crystals of rocks formed in this way point in the direction of the tensions and loads acting on them (e.g. schist); the temperature increase associated with the geothermal gradient originating from increased proximity with the earths mantle leads to dehydration of the original minerals (e.g. formation of potassium feldspars from micas). In following the origin and history of a clayey formation it is extremely important to have an understanding of the type of rock making up the erosion basin and 49


the sedimentation basin so as to better understand the mechanisms that form clayey rocks. At present the earths crust, on average, is thought to be as follows: % by weight (oxides) SiO2 Al2O3 FeO + Fe2O3 CaO Na2O K2O MgO TiO2 Others

% by weight (elements) 60 15.5 7 5 4 3 3.5 1 1

O Si Al Fe Ca Na K Mg Ti Others

46.7 27.7 8.1 5.1 3.6 2.7 2.6 2.1 0.5 0.9

% by volume (elements) 92 0.8 0.8 0.7 1.6 1.5 2.1 0.6 0.1

The minerals representative of this composition are not very numerous, and are those that generally give rise to clays.

50

Rocks

Main minerals in the earth’s crust CALCIUM-SODIUM FELDSPARS (Plagioclases) Albite Oligoclase Andesine Labradorite Bytownite Anorthite

0 - 10% An 10 - 30% An 30 - 50% An 50 - 70% An 70 - 90% An 90 -100% An

Na[AlSi3O8]

Ca[Al2Si2O8]

POTASSIUM FELDSPARS

22%

Orthoclase, Microcline, Sanidine

K[AlSi3O8]

QUARTZ

SiO2

PYROXINES Diopside Augite

42%

18% 9%

(Ca,Mg)[Si2 O6] (Ca,Mg,Fe,Ti,Al)2 [(Si,Al)2 O6]

AMPHIBOLES

6%

Hornblende, Biotite, Olivine MUSCOVITE

KAl2[AlSi3O10](OH,F)2

FELDSPATHOIDS Leucite Nepheline

3% <1%

K[AlSi2O6] KNa3[AlSiO4]4

Clayey formations may, then, vary greatly in terms of composition, degree of consolidation etc. It is thus necessary to classify them by grouping them together according to: Cohesion and water content - thus differentiating plastic clays from lithificated ones.

51


Composition - in such a way as to differentiate pure clays (made up of just one clayey mineral) from carbonatic clays (known as marls) and sandy clays (containing quartz, SiO2) etc. If the mineralogical composition of the clays and their stratigraphic distribution etc. are variable it is said that they belong to the group of the clastic or hallogenetic rocks. If the clays are found at the (sedimentation) location where they were formed they are known as residual, while they are defined as allochthonous if they have been transported and excavated in some other area. In any case, the problem of clayey rock classification is extremely complex, and is beyond the scope of this work. Yet examination of individual clayey minerals, and the characteristics that they have in an industrial context is, as might be expected, of extreme interest. Structure of clayey minerals Having provided a brief summary of clayey formations in relation to the other rocks from which they are derived, we shall now take a look at just what they consist of so as to obtain a better understanding of their genesis. Of particular interest is structure, as it is this which determines their behaviour and differentiates them, especially as regards their capacity to combine with water a key factor in providing plasticity. Here, an introduction to simplified silicate classification will be useful (i.e. the ways in which repetitive tetrahedral SiO4 units are organised). Links are made between vertices only, with oxygen atoms forming bridges between silicon atoms, thus giving rise to: Nesosilicates, made up of independent tetrahedrons.

Sorosilicates, with two, three, four or six tetrahedrons, possibly in a closed loop configuration.

52

Rocks

Inosilicates, having chains of tetrahedrons linked by cations.

Phyllosilicates, in which the union of three vertices of each tetrahedron with another forms layers.

Tectosilicates, where the tetrahedrons unite to form a 3-dimensional lattice.

As mentioned above, clayey minerals are generally alumino silicate hydrates of varying complexity, consisting of lamellas that can slide against each other (phyllosilicates). The molecular structure of such minerals consists of the combination of tetrahedral repetition units (coordination 4), with a silicon atom surrounded by 4 oxygen atoms, and octahedral repetition units (coordination 6) with an aluminum atom sometimes substituted by magnesium or iron surrounded by 6 oxygen atoms. Among the most common elements in the earths crust, oxygen is undoubtedly the most widespread and is one of the largest too: 53


ION

IONIC RADIUS (Å)

O 2Si 4+ Al 3+ Fe 3+ Ti 2+ Ca 2+ Mg 2+ Na + K+

1.3 0.4 0.6 0.7 0.6 1.0 0.8 1.0 1.3

Bearing in mind the ionic radius of oxygen and silicon (the second most widespread element) it can be seen how this has the bulk to be at the centre of a compact, regular polyhedron of minimum potential energy, with a coordination number of 4 (i.e. the tetrahedron [SiO4]4-). The octahedral configuration with a coordination number of 6 is given by the regular arrangement of the oxygen ions or, more commonly, hydroxyls (OH-), which have the same radius around an Al3+ ion of a size suitable for containment at the centre of such a structure. Both silicon-based tetrahedrons and aluminium-based octahedrons can combine together to form repetition units via apex-apex, edge-edge or face-face bonds, giving rise to different structural repetition units of various shape. The combination of tetrahedral and octahedral structural units, arranged in regularly overlain layers with well-defined inter-layer distances, is also possible: those distances are extremely important in determining the macroscopic characteristics of individual clayey minerals. The simplest configuration is the two-layer one, with alternation of units made up of tetrahedral linked 2-by-2 by an apex (Si2O5)2- and octahedral layers (Figures 29 a and b show grouping of the kaolinite and montmorillonite).

(A) Halloysite (B) Kaolinite

Fig. 29a. Example of possible structural arrangements in clays.

54

Rocks

(A) Na-montmorillonite (B) Ca-montmorillonite

Fig. 29b. Possible structural arrangements of clays.

This can be described by the rough formula (OH)3Al2(OH)Si2O5, or, more appropriately, Al2O32SiO22H2O, which corresponds better with the chemical analysis, expressed in oxides. This molecule is electrically neutral (6 positive charges from the aluminum + 8 positive silicon charges = 14 positive charges, saturated by the 4 negative charges of the OH- groups and 10 negative charges from the remaining 5 oxygens = 14 negative charges) and consists of particles around 1 µ thick, although the finest may be smaller than 0.005 µ (i.e. the equivalent of 7 elementary sheets), thus forming ionic crystals of varying size. Clayey minerals of greater complexity may have octahedral layers interspersed between tetrahedral layers, in 3-layer arrangements as is seen with talc and pyrophyllite; this structure is also seen in Illite (mica group), where the triple tetrahedral/octahedral/tetrahedral layer is associated with another octahedral layer etc. (Figs. 30-31). (A) Vermiculite (B) Tri-octahedral Chlorite

Fig. 30. Further examples of phyllosilicate conformations.

55


(A) Muscovite (B) Illite

Fig. 31. Further phyllosilicate conformation examples.

This way of organising different types of clayey sheet constitutes the most useful distinctive criteria in classifying clayey mineral groups. The fact that there is, between the sheets, a space large enough to allow the introduction of one or more water molecules (inter-sheet or inter-layer water) gives rise to a further classification, this time a technological one, as the presence of such water greatly influences the capacity of these various layers to slide against each other and thus ultimately determine the plasticity of the material. It is thus possible to formulate a simplified classification of the main clayey materials: Tetrahedrons/Octahedrons

Mineral group

Lattice space (Å)

Non-crystalline

ALLOPHANES

1/1

KAOLINITE (Nacrite, Dickite)

7.2

1/1

HALLOYSITE

2/1

PYROPHYLLITE TALC

2/1

MICAS (ILLITE)

2/1

SMECTITES (Montmorillonite etc.) VERMICULITE

2/1/1

CHLORITE

7.2 - 10.1 10 variable 10 - 21 10 - 14 14.1

Within such groups, the nature and position of the substitutive cations constitute a further classification criteria.

56

Rocks

For example, substitution within an octahedral cavity is particularly significant, as in the passage from kaolinite Al2Si2O5(OH)4 to a mineral of the serpentine group Mg3Si2O5(OH)4. These substitutions also determine the ionic exchange capacity of these materials that will be discussed further on. Indeed, the electrical balance must be maintained, and this means that ions originating from the environment surrounding the material can compensate for any charge deficit. In this sense the different cations behave in a selective manner, effecting an orderly substitution as follows Na+ < K+ < Ca++ < Mg++ which means that the calcium ion can replace the sodium ion more easily than the reverse. This compensatory exchange effect may vary as a function of the ionic charge of the solutions. An enormous variety of exchanges can take place, even including in-tetrahedron substitution of silicon with aluminum; thus, as a function of the ionic charge, compensation processes are activated, mainly through the participation of K+ and Na+ ions that position themselves, via electrostatic attraction, on the basal faces of the micelles. In the case of Illite, for example, substitution of aluminum with silicon may take place in the octahedron and this weakens the total ionic charge of the clayey micelle in which there was a charge imbalance of a sign opposite to that at tetrahedral level: the K+ cations, then, are fixed less stably on the surfaces, but tend to position themselves laterally, largely where there are discontinuities or fractures, and this alters the characteristics of this clayey material even further. The process occurs, in the same way, and especially as regards substitutions in the octahedral layers, in smectite-type minerals. In these cases there is a general decrease in ion-sheet and sheet-sheet binding forces, thus leading to a variable lattice distance that consequently allows insertion of whole water molecules, with all the resulting consequences. Similar phenomena occur in the chlorites, but the inter-layer water is often replaced by Al(OH)3, Fe(OH)3 and, above all, Mg(OH)2. Cationic exchange capacity (CEC), easily established using routine analytical techniques, is an intensive measurement which provides a realistic image of just how much these phenomena are present in a clayey mineral. It gives a measure of the ease of cationic exchange in the structure of the mineral, via the introduction of Na+ ions. Observing that: Clayey mineral

CEC (meq/100g)

Kaolinite Illite Montmorillonite

5 - 15 20 - 50 50 - 150

57


it is evident how isomorph substitutions are more marked in complex structures having several associated layers of elementary tetrahedral-octahedral elements. Characteristics of the main clayey minerals From the above it can be seen that it is possible to effect a more or less complete characterisation of clays in general via the description of a few clayey minerals (kaolinite, illite, montmorillonite and chlorites), although it should be born in mind that polyphyllitic mixed layers often occur (e.g. Bravaisite, consistent in a mix of Illite and Montmorillonite). We shall now go on to examine the main characteristics of the individual classes of clayey raw materials.

58

Rocks

Chapter III RAW MATERIALS FOR CERAMIC BODIES

Kaolinite Origin of the term Derives from the Chinese words kau-ling, meaning high hill, indicating the province of Jianxgi, the place where this raw material was first extracted for ceramic manufacturing purposes. Mineralogical structure The base element consists of a tetrahedral sheet in which each tetrahedron consists of four oxygen atoms (each with two negative charges), positioned at its apexes, and a silicon atom (with four positive charges) at its centre. This is associated with an octahedral sheet in which each octahedron consists of six OH- hydroxyl units (each carrying a negative charge) located at its apexes and a cation (usually trivalent but also bivalent) at its centre. This base element is illustrated in Fig. 32.

Structure of kaolinite according to Gruner

Elementary kaolinite cell

Fig. 32.

59


The tetrahedral unit thus has four non-saturated negative charges; the octahedral unit is negative too with three (if the cation is trivalent) or four (if bivalent) non-saturated charges. In the kaolinitic base element the two sheets bind by substituting an OH hydroxyl on each octahedron with the sharing of an oxygen atom (at the tetrahedron apex) which had a free charge. This anion is thus shared by a tetrahedron and two octahedrons in that each OH hydroxyl unit is balanced by the aluminum cations of two adjoining octahedrons. With this coupling the kaolinite structure is electrically neutral. This structure can be repeated (vertically) a great many times while maintaining the same distance between equal, adjoining levels. It is this structural rigidity that explains the non-expandability of the preserved lattice despite the weakness of the forces on which maintenance of the structure depends (Van der Waals and hydrogen links). Chemical composition Al2(X2O5)(OH)4 Where x is the element in the tetrahedral position (Si4+ and, as vicariant, Al3+ and Fe3+, which lead the structure to be deficient by one positive charge). Theoretical composition SiO2 = 46.54%; Al2O3 = 39.50%; H2O= 13.96% Genesis and characteristics of deposits containing kaolinite These deposits originate from the transformation of crystalline rock feldspars which leads to simultaneous removal of the alkaline elements (sodium and potassium). In those rare cases that give rise to the formation of nearly-pure kaolin (usually from feldspathic veins) it is milky white, has low plasticity, a greasy surface and leaves a persistent whitish film on the skin when touched. Usually, though, kaolinitic deposits are mixtures of kaolinite, quartz (resistant to alteration and transformation phenomena) and feldspars (relics of an incomplete aggression and transformation action); ferrous oxides (largely responsible for any colouring of the material) and micas also have to be considered. Kaolin deposits are generally primary in origin. This means that they exist where the source rocks once existed and therefore have not been transported; secondary deposits, instead, are those that have undergone removal and re-deposition in an aqueous (usually lacustrine) environment. These processes concentrate and select the finest particles, while, however, contaminating them with materials of organic vegetable origin (leaves, branches, roots, carbonaceous elements). It is important to remember the difference between deposits of kaolin (of poor plasticity) and kaolinitic clays. The latter, of secondary origin, are characterised by high plasticity (ball-clays); this is a consequence of natural particle size selection 60

Raw materials for ceramic bodies

and the addition of plasticizing materials (vegetable and carbonaceous parts) that sometimes blacken the raw material. Location of main deposits and extraction techniques The most important kaolin and kaolinitic clay deposits are found in England (Devon and Cornwall), Germany (Westerwald and Bavaria in the districts of Wiesau-Tirschenreuth and HirschauSchnaittenbach), France (Brittany), Czechoslovakia (Karlovy Vary) and the USA (Georgia and South Carolina). The oldest industrial method for the extraction of kaolinite from the source rock involves the use of high-pressure water jets (monitors). These are directed against the walls of the deposit so that the kaolinitic part is removed, suspended in water. That suspension is then classified by sedimentation and cyclones. This technique was invented in Great Britain and is still used on British extraction sites where suitable. As far as ball-clays are concerned, modern extraction methods have virtually eliminated underground mining in favour of open-cast methods with terraces and ditches. In Devon (England) the overburden is removed by bulldozers and scrapers, with extraction being performed on banks varying in thickness from 3 to 6 m using hydraulic excavators, draglines and bucket-wheel excavators. Numerous beds of lignite may result in up to 50% of the removed material being discarded as waste. Working conditions in the pits, together with the impermeable nature of the raw material in question, can lead to flooding, thus requiring the use of high-capacity pumps to keep the extraction zones operative. Extraction is highly selective everywhere: removal is preceded by systematic core sampling and laboratory tests which provide the required pre-mining information. The techniques adopted involve terracing methods that employ continuousoperation hydraulic wheel dredges that feed a series of conveyor belts (covered to protect them from rain). These carry the clay to the storage areas. Alternatively, hydraulic excavators are used to load trucks to take out the clay. Mineralogical analysis of kaolinitic raw materials; standard behaviour and identification Identification via X-ray diffractometry (Fig. 33) essentially involves observation of the basal plane diffractions. The specific lattice distances are as follows: (plane 001) 7.18Å; (plane 020) 4.47Å; (002) 3.57 Å; (003) 2.38 Å; (004) 1.78 Å. The diffraction pattern for CuKa radiation gives peaks in the following angular positions: 12.51 (001), 20.90 (020), 25.0 (002), 51.11 (004). The presence of certain chlorites can mask the presence of kaolinite owing to coincidence of basal reflections at 12.50° (corresponding to 7.18 Å). In this case any splitting that occurs in the reflection on the 002 plane must be checked for, given that there exists a small difference between the lattice distance of 61


disordered kaolinite

Fig. 33. Example of a diffraction pattern for highly disordered kaolinite.

kaolinite (3.47Å) and chlorite (3.54Å) to which positioning of the peaks at 24.90 and 25.20 correspond respectively. The DTA (Differential Thermal Analysis) diagram for kaolinite (Fig. 34) shows a net endothermic peak at 580-600 °C, associated with the transformation of kaolinite in meta-kaolin with crystallisation water loss; at around 980 °C there is, instead, a marked exothermic peak that is probably connected with the crystallisation of mullite or alumina (Fig. 35). Thermogravimetric (TG) analysis shows a sharp weight loss at around 520~580 °C (i.e. when the crystallisation water is lost); no variations, instead, are observed towards 980 °C, in that this particular exothermic reaction is caused by the liberation of bond energy and not the removal of material. Up to 450 °C, dilatometry (Figs. 36-37-38) is characterised by weak dilatation (because of substantial alumina content). There then comes (beyond 550 °C) an initial shrinkage phase associated with the elimination of the OH hydroxyls (this extends towards 880 °C) followed by a second, more marked stage of contraction associated with the start of sintering reactions.

62


Fig. 34. DTA curves for several minerals in the kaolinite group.

Fig. 35. Mullite crystals formed by heating of kaolinite in a siliceous matrix (magnification x 37,000) (by Kingery: Introduction to Ceramics - Wiley).

63

Linear expansion (%)


Temperature °C

Fig. 36. Dilatometric curve for a high-purity kaolin sample, showing two abrupt contractions corresponding to the dehydroxylation phase and the recrystallization reaction.

Fig. 37. Dilatometric curves for kaolins of increasing quartz content (from curve a to curve c); the effect of the quartz gradually prevails over that of the kaolinite.

64


Fig. 38. Dilation-Shrinkage of kaolinite-quartz mixes (by G. Peco 1970). (0, 12, 24, 36, 48% of quartz).

The table below gives standard mineralogical analysis for some common kaolinitic raw materials: BALL CLAY Kaolinitic Minerals Micaceous Minerals Montmorillonite Minerals Quartz Feldspars Others

1 50 21 22 1

2 48 19 28 1

3 76 11 7 3

5 27 25 4 37 5 2

6 36 32 28 4

CHINA CLAY Kaolinitic Minerals Micaceous Minerals Montmorillonite Minerals Quartz Feldspars Others

7 85 14

8 80 16 4

9 79 15 1 5 1

11 92

12 87 2

3 ~ 2 65

2 8 1


Technological characterisation of pressed kaolinitic materials It is, of course, necessary to distinguish between ball-clay, china-clay and kaolins. Ball clays are characterised by: fairly good (> 0.3%) post-pressing expansion excellent unfired bending strength (green > 12 and dried > 30Kg/cm2) at 1100 °C shrinkage varies from 2 to 4% while apparent porosity is 10~15% again at 1100 °C, bending strength may vary (from 100 to 200 Kg/cm2) as a function of impurities, which, to varying degrees, aid sintering again at 1100 °C, the coefficient of cubical expansion may vary between 150 and 200 x10-7.

China clay, instead, has the following characteristics: greater post-pressing expansion poor bending strength of both green and dried product possible post-drying expansion at 1100 °C moderate dimensional contraction and high water absorption values (15-20%) are observed; bending strength (< 150 Kg/cm2) and the coefficient of cubical expansion (120-180.10-7) are low, once again depending on the quartz content. On the whole the material is highly refractory. Finally, the kaolins are characterised by: considerable post-pressing expansion water absorption (at 1100 °C) between 20 and 25% very low modulus of rupture in bending (80-150 Kg/cm2) very low coefficient of cubical expansion, generally between 100 and 150.10-7.

Ceramic uses of raw materials not subject to purification or treatment This largely concerns ball-clays which are generally employed in their natural form, at most mixing them with various other types in order to supply users with blends of stable ceramic performance. Ball-clays are employed in vast range of ceramic products (sanitaryware, tiles, pottery, ornaments etc.) where they act as plasticizing binders. The percentages in which they are employed vary from product to product, and also within the different compositions used to obtain the same type of finished article. However, as a rough guide, the following quantities are generally employed: Tiles (wall) Tiles (floor) Porcelain tile Pottery (earthenware) Sanitaryware

66

25-50 25-75 30-40 20-35 22-26


In the field of frits and glazes kaolinitic materials, purified using wet techniques, are used as components in mixes to be fritted (melted) and as a plasticizing additive during the grinding of frits. Kaolins of lesser purity are employed in the manufacture of alumina refractories where they act as alumina carriers. For some such products untreated kaolins contaminated by the presence of iron may sometimes be used. However, only a small percentage of the worlds total kaolin output is actually employed in the ceramic industry: almost 90%, in fact, is channelled into the paper industry where it is used as a filler or surface coating. It is also used in minor quantities as: a filler for tyres and rubber a filler for paints a pharmaceutical support (pills) a cosmetic and toothpaste support a powdery insecticide support a fertiliser support. Illite Origin of the term From the name of the American state of Illinois where the mineral was first identified. Mineralogical structure Similar to that of the micas. The basic structural unit consists of two tetrahedral sheets sandwiching an octahedral one (of trioctahedral form), as illustrated in Fig. 39. It is virtually the same as the montmorillonite structure except that some silicon cations are always substituted by aluminum to which potassium is added, thus maintaining the electrical balance of the structure. However, the potassium cation is attached securely and is not easily replaced as happens with montmorillonite. The structural differences between illite and mica lie in lesser substitution of silicon with aluminum and, therefore, less potassium aggregation. Furthermore, illite may contain interlayered water molecules. In addition to the trioctahedral illites, others of dioctahedral form (a sheet of tetrahedrons with a sheet of octahedrons) exist. Alteration causes the latter types to turn into montmorillonite while the former are transformed into vermiculite. The SiO2/Al2O3 ratio (taking into account substitutions in the octahedral sheet) varies from 2 to 4; it is frequently 3. Glauconite (ferroan illite), an illite of particular interest, is generated by alteration of the biotite.

67


Fig. 39. Crystallographic structure of illite.

Chemical composition Illite is represented by the following formula: Al2 (Si3AlO10) K (OH)2 - Kx Yet it is not completely definable; aluminum substitutes the silicon less frequently and, as a consequence, less potassium is required. Genesis and characteristics of deposits Illite deposits are generally linked to the laying down of sediments. Of course, deposits never consist exclusively of this mineral: it is generally found in association with chlorite, montmorillonite, kaolinite, quartz, calcite, dolomite, etc. Location of main deposits and extraction techniques Historically, the ceramic industrys main source of supply has been the illite-rich clayey deposits located in the Apennine hills of the Italian region of Emilia. Here lie the extensive grey-blue clay beds dating back to the Pliocene, the socalled Complesso Emiliano made up of the marls of Montepiano (red and green clays that go by the name of red-beds) and the Antognola series (marls and greyblue clays). The chemical make-up of such clayey rocks is illustrated in the tertiary diagrams (Figs. 40-43) on the following pages.

68


Illite-chlorite clays Illite-kaolinite materials German English French

Fig. 40. Al2O3/Na2O + K2O/TiO2 + Fe2O3 + MgO + CaO tertiary diagram showing compositional fields for red gres clays and the imported kaolinitic materials used in white gres (Fabbri & Fiori, 1985/7).

Illite-chlorite clays

Illite-kaolinite materials German English French

Fig. 41. Na2O + K2O/Fe2O3 + TiO2 /MgO + CaO tertiary diagram showing compositional fields for red gres clays and the imported kaolinitic materials used in white gres (Fabbri & Fiori, 1985/7).

69


Illite-chlorite clays Illite-kaolinite materials German English French

Fig. 42. SiO2 /AlO3/TiO2 + Fe2O3 + MgO + CaO + Na2O + K2O tertiary diagram showing compositional fields for red gres clays and the imported kaolinitic materials used in white gres (Fabbri & Fiori, 1985/7).

nepheline syenite

Feldspars Feldspathic rocks Quartz-feldspathic sands Quartz sands

Fig. 43. SiO2/Al2O3/TiO2 + Fe2O3 + MgO + CaO + Na2O + K2O tertiary diagram illustrating the compositional fields for complementary raw materials used in the production of gres (Fabbri & Fiori, 1985/7).

70


Yet Italy is not Europes only source of illite: of particular renown are the Hungarian deposits of white illite associated with montmorillonite, the plastic Miocene clay of Czechoslovakia, and the vast Ukrainian deposits of kaolinitic illite that producers of porcelain tiles (also known as vitrified stoneware or referred to by its Italian name gres porcellanato) have recently begun using. In the United States (Illinois), the illite deposits for the production of expanded clay have been in use for some time. Extraction techniques in hilly areas consist of successive ripping and scraping operations on terraces (usually more than 20 m wide and some tens of metres long). These continue until deposits are exhausted. Alternatively, on slopes, high-power bulldozers are used. In semi-flat areas where the extraction zone provides more room for manoeuvre, scrapers are preferred to bulldozers. This vehicle, as the name implies, scrapes wide strips of terrain, funnelling the material inwards towards the machine: this system provides excellent homogenisation of the raw material, a process aided by the vehicle loading system. Ripping and scraping has several points in its favour. Firstly, it allows work to be carried out over a vast surface area, with large-scale transportation of material resulting in low unit costs. Secondly, there is the advantage of flexibility: unsuitable extraction zones can be isolated and the same machines can be used for both handling and treatment of the extracted clay (crumbling, homogenisation, drying). This last sequence takes place in the quarry itself: a 15-20 cm thick layer of the material is spread out and left in the sun for a full day, occasionally being churned over to ensure proper uniformity of exposure to the suns rays. Mineralogical analysis of illite raw materials; standard behaviour and identification Identification via X-ray diffractometry essentially involves observation of basal plane diffractions. Characteristic reflections appear at 8.80, 17.80 and 26.70° 2ϑ, corresponding to lattice distances of 10, 4.9 and 3.33 Å respectively. The illite-relevant DTA diagram (Figs. 44-45) shows a first endothermic peak at around 140 °C (removal of zeolite water) followed by a second peak towards 600 °C (hydroxyl elimination). Towards 900 °C formation of alkaline silica-aluminates takes place; where calcium is present, calcium aluminates, calcium silicates (wollastonite) and calcium ferrites are also formed. These neo-formations are indicated by an exothermic peak. Thermogravimetric (TG) analysis highlights moderate weight losses, largely concentrated where removal of zeolite water and hydroxyls reaches a maximum. Dilatometry (Fig. 46) produces a steady gradient up to and beyond 600 °C, sometimes preceded by a hint of contraction at around 120-140 °C; from 600 °C to over 800 °C comes the typical flattening out of the curve (i.e. no dimensional variation) followed by an almost vertical contraction as the dilatometer leverage system induces plastic deformation of the sample. 71


Temperature °C

Fig. 44. The DTA curve for illite is characterised by weak endothermic effects at temperatures which, in other clayey minerals, cause much more intense reactions.

Temperature °C

Fig. 45. A typical DTA curve for muscovite and dioctahedral micas in general.

Standard semi-quantitative mineralogical analyses of the three Emilia Apennine clay types (taken by way of example of such clayey rocks) are provided below: 1. 2. 3.

Illite xxx xxx xxx

Chlorite xx xx xx

Montmor. Kaol. xxx x xxx x x

Quartz x x xx

Feldsp. x

Carb. xx x tr.

Mica tr. tr. tr.

As can be seen, the mineralogical picture is a fairly constant one and, on the whole the only significant differences concern carbonate content which swings from a maximum of 20-23% in Pliocene and Pleistocene grey-blue clays to 1372


Fig. 46. Dilatation-Shrinkage of illite-quartz mixes (by G. Peco 1970). (0, 12, 24, 36, 48% of quartz).

16% in those of the Antognola series and just 0-6% in the red-beds. With regard to red-beds it should be pointed out that these sometimes have a very high carbonate content (sometimes higher than 16%!). A final observation concerns natural plasticity (i.e. that of the unground raw material); clays of the Antognola series and especially red-beds have undergone diagenesis processes which have partially or completely inhibited this typical property of clayey materials. Technological characterisation of pressed illite materials Tiles formed from predominantly illite raw materials have the following characteristics: normal post-pressing expansion good unfired bending strength (green and dried) around 1100 °C shrinkage varies from 5 to 8% with water absorption values in the 0-5% range. Nevertheless, in the (frequent) event that these clays are accompanied by a substantial carbonate content (from 15 to 23%), dimensional variation switches from +1 to 1% while porosity falls between 22 and 15% 73


the bending strength of fired products is markedly influenced by the presence or absence of carbonates: where they are present the material is significantly weaker (from 100 to 200 Kg/cm2) while where they are absent strengths of 200 Kg/ cm2 are exceeded with ease the coefficient of cubical expansion (at 1100 °C) generally falls within the 180225.10-7 range. On the whole tiles obtained from these raw materials are readily vitrified. Carbonate-free tiles achieve complete vitrification at 1050-1080 °C, within very tight palliers (ranges); carbonitic ones are slightly more refractory and become vitreous beyond 1100 °C, collapsing suddenly with a total absence of pallier (a characteristic of carbonates). Where clays are rich in carbonates fired product colouring varies from salmon pink to yellowy and greeny-yellow (at vitrification); other clays, instead, become bright red at lower temperatures, switching to dark brown when liquid-state reactions gain the upper hand. Uses in different industries White firing illite clays are employed in certain industries as an alternative to kaolinite clays (e.g. as powdered supports or fillers). Some red firing types are employed in the production of light expanded materials used by the construction industry: in this case they must contain substances that make the (previously pelletized) material expand during firing, or low-cost additives having this characteristic must be used. Curiously, illite is also used to produce cat litter; oddly, a renowned Italian product directory (Villavecchia) cites only this use and completely forgets to mention that clay is also employed in ceramics. Montmorillonite (smectite) Origin of the term From Montmorillon in France where this clayey mineral was first identified. The term bentonite indicates rock largely composed of montmorillonite, characterised by enormous swelling when it comes into contact with water. The name comes from Fort Benton in Wyoming, USA, the site of important deposits. Smectite is a synonym for montmorillonite. Mineralogical structure The (trioctahedral) base structure is made up of two tetrahedral sheets enclosing an octahedral one (see Fig. 47). Specifically, this is the structure of pyrophyllite. Some members of the montmorillonite family derive from substitution of silicon with aluminum in the tetrahedrons and substitution of magnesium and bivalent iron in place of aluminum in the octahedrons. 74


The mineral crystals of this group are bound by particularly weak forces. There are no hydrogen bonds (as in kaolinite) in that there is no possibility of contact between the octahedral and tetrahedral layers belonging to different base structures, because the former are prisoners inside the latter. The only links are the Van der Waals weak forces: consequently, the insertion of water molecules is extremely easy, resulting in the crystalline structure expanding to nearly six times its original volume. The greater reactivity of the montmorillonite family (with respect to, for example, the kaolinite family) derives from the suitability of the internal faces that mark the boundary of each base structure, in that these elements are easily separable. The property known as isomorphic substitution (i.e. the capacity for replacement of a cation with another of a different charge) is also much evident. If, for example, a trivalent (aluminum) ion replaces the (tetravalent) silicon of the tetrahedrons, electrical equilibrium is only achieved following absorption of an external (mono or bivalent) cation; in this second case the number will be equal to half the negative charges released by the isomorphic substitution. The peculiar behaviour of montmorillonite is partly explained by the extremely small size of its particles. Then there is the significant diversity between the sodium and calcic montmo-

Fig. 47. Elementary montmorillonite cell.

75


rillonites: the former are much more plastic with a much higher ionic exchange capacity. It is also fairly common to encounter mixed structures created by the association of montmorillonite and illite (interlayers). These are formed owing to the similarity of the two different mineral sheets. Chemical composition X2Y4O10 (OH)2 X = trivalent aluminum cation and bivalent magnesium substitute (octahedral sheet) plus mono or bivalent cations that aggregate to the structure to balance nonequilibrated electrical substitutions. Y = tetravalent silicon cation and trivalent aluminum substitute (tetrahedral sheet) plus mono or bivalent cations that aggregate to the structure to balance nonequilibrated electrical substitutions. Theoretical composition SiO2 = 66.7%; Al2O3 = 28.3%; H2O = 5% Genesis and characteristics of deposits Many montmorillonite clays were formed by alteration of volcanic rocks (pyroclastic rock and volcanic breccia) such as: rhyolites, trachytes, dacites, andesites, basalts and liparites. The bentonite of Wyoming derives from in situ alteration of volcanic ash that was accumulated in a Cretaceous formation. Sedimentary genesis is equally common. Sedimentary deposits of montmorillonites usually consist of layers or lenses several metres thick; those of hydrothermal origin may have somewhat irregular forms that complicate extraction. Mineralogical analysis of montmorillonite raw materials; standard behaviour and identification Identification via X-ray diffractometry largely involves observation of basal plane diffractions (Fig. 48). It is not uncommon for the extreme fineness of the particles to cause difficulties during diffractometry testing; where conditions permit analysis to be carried out, reflection is detected at 6-6.9° 2ϑ (corresponding to a lattice distance of 12.5 Å), which shifts towards a lower angle in the event of glycolation of the sample (4.8-5° 2ϑ) and collapses at 8.8° after being heated to 550 °C. Smectite, however, is often in interlaminated form associated with illite. In this event the resulting reflection band is greatly widened around 7.2°, expands again at 4.8-5.2° and collapses, after heating to approximately 550 °C, at about 8.8°. The DTA diagram for montmorillonite shows two endothermic peaks at 180 °C and between 450 °C and 650 °C respectively. The first of these peaks is linked to removal of interlayer water and the second to removal of the OH- hydroxyls (see Figs. 49-50-51). 76


1

-

1 1

1

1

-

Fig. 48. Diffractometry analysis of a montmorillonite (Q = quartz, M = montmorillonite).

Thermogravimetric (TG) analysis shows, in correlation with DTA endothermic peaks, two weight losses at 180 °C and 450/650 °C respectively. After an initially short tract of limited expansion, the dilatometric curve shows

Temperature °C

Fig. 49. DTA curve for montmorillonite; the low temperature dehydration peak (100-250 °C) is the most intense and characteristic and its form varies greatly as a function of the interlayer cation: a) Camontmorillonite; b) Na-montmorillonite. The other dioctahedral smectites (beidellite and montronite) have curves similar to those of montmorillonite.

77


Temperature °C

Fig. 50. Typical DTA curve for trioctahedral smectite minerals such as saponite.

Fig. 51. Some montmorillonite and vermiculite DTA curves.

78


a first contraction towards 180 °C followed by further expansion (which reaches its maximum towards 573 °C should the raw material also contain free quartz). Towards 900 °C further contraction is induced by sample sintering reactions and softening. Location of main deposits and extraction techniques In Europe, the most important extraction sites are in Great Britain (Surrey, Somerset and Bedfordshire), France (Limousin), Germany (Bavaria), Hungary (Istenmezb and in the Tokay hills), Italy (Sardinia and the island of Ponza), Czechoslovakia, the ex-USSR (Caucasus and Kazakhistan), Greece (islands of Milos and Mikonos) and in the formerly Yugoslavian countries. The United States has the well-known Cretaceous deposits of Wyoming (bentonite) and others situated in Arizona, Oklahoma, Texas and Nevada. Other North American deposits are found in Canada (Alberta and Manitoba) and Mexico (in the regions of Puebla and Monterrey). Substantial bentonite deposits are also to be found in South America (Argentina), Asia (India and Japan) and South Africa. Quarries opened on predominantly montmorillonite clayey deposits are generally of the open-cast type. These are thus earth-moving operations employing traditional methods. Nevertheless, workers at these sites often have to deal with the materials extremely high plasticity and water take up. Having to handle a material with a water content of 15 or 25% is not uncommon. Under such conditions the enormous plasticity often puts bulldozers out of action as huge mattresses of material tend to build up on their tracks. Hence it is often preferable to use hydraulic excavators or dredgers. The latter also allow for more selective extraction and the accurate mining of wide strips of limited thickness. Technological characterisation of pressed montmorillonite materials Montmorillonite deposits are generally characterised by the substantial presence of highly plastic colloidal particles. Tiles made of raw materials having a significant portion of minerals from the montmorillonite family have the following characteristics: generally low post-pressing expansion (0.5%) very high unfired bending strength (green and dried) intense drying shrinkage (> 1%) on fired products obtained at temperatures as low as 1020 °C dimensional variation oscillates from 5 to 10% (shrinkage) while the corresponding porosity may range from practically zero to more than 5%, depending on the degree of postsintering expansion. These values vary greatly as a function of alkaline ion content in the inter-lattice spaces again at 1020 °C, bending strength varies greatly (from 150 to 250 Kg/cm2) depending on the degree of expansion.

79


Ceramic uses While the use of clayey raw materials containing montmorillonite in layers mixed with illite is quite common (especially in the manufacture of tiles, bricks and pottery), utilisation of almost exclusively montmorillonite materials is extremely limited owing to the characteristics listed in the preceding paragraph. It should be pointed out that montmorillonite has a markedly negative effect on the rheological properties of slips. Bentonite, when hydrated, swells enormously and takes on the aspect of a colloidal gel. In ceramics, pure montmorillonites are, in fact, only employed as plasticizing additives, and even in this case they are never used in proportions of more than 5%. Uses in other industries Bentonite is, first and foremost, a key constituent in the sludge used in oil drilling operations. This sludge carries the lithic material removed during drilling to the surface. The use of a fluid that is more viscous than water is a great aid to such drilling operations, as long as it can be pumped. Drillers generally prefer to use sodium montmorillonite: the clay must be of high purity and free from any abrasive (quartz) particles. Montmorillonite raw materials are also used in de-colouring; here, the employed material, known as fullers earth, predominantly consists of montmorillonite, although there are plenty of such deposits where the predominant mineralogical type is attapulgite or halloysite. Another interesting industrial application of montmorillonite is in the clarifying of alcoholic beverages (wine, beer, liquors etc.). This operation involves the coagulation and removal of the colloidal impurities in suspension. Montmorillonite (the sodium form is preferred in this case) is added directly to the beverages and (after agitation) then separated using filter presses. Chlorite Origin of the term Derived from the Greek cloros (meaning green), this word was introduced by Werner to define this mineral on account of its predominant colour, stemming from a high reduced iron content. Mineralogical structure These are lamellar minerals having an aspect similar to that of mica. The classic chlorite structure consists of an alternating association of a micaceous-like sheet (where an Al3+ or Mg2+ octahedral plane is enclosed by two hexagonal planes of silica tetrahedrons) with a brucite one (characterised by a flat hexagonal lattice of silica tetrahedrons with a central cation) as illustrated in Fig. 52.

80


Most chlorite minerals have a trioctahedral structure, yet dioctahedral forms have also been identified. Cases of degraded chlorites showing lattice expansion have also been observed: such expansion is caused by a capacity to adsorb zeolite water as a result of weathering of the brucite sheet. The existence of chlorite-vermiculite and illite-chlorite interlaminates should also be noted. Chemical composition The mineralogical structure is not original, but derives, rather, from those of the two associated micaceous and brucite sheets. Composition of the micaceous sheets is as follows: X8Y6O10 (OH)4 where: X is due to Si4+ or Al3+ Y is due to Mg3+ or Fe3+

Fig. 52. Elementary chlorite cell.

81


The brucite sheets, are instead, composed as follows: X6 (OH)12 where: X is made up of Mg2+, Al3+ or Fe2+. When, in the micaceous structure, an Si4+ replaces an Al3+, the resultant electrical charge difference is balanced by an excess charge in the brucite sheet generated by the replacement of Al3+ with Mg2+. The theoretic formula of true chlorite (or leptochlorite) is as follows: Mg3 (Mg3-x Alx) (Si4-x Alx) O10 (OH)8 Oscillation of the x value (which can vary from 1 to 2) and the nature of the substituents lead to the formation of different varieties, such as bavalite (ferrous chlorite), clinocore, penninite, prochlorite, etc. Genesis and characteristics of deposits Deposits containing chlorites largely originate form sedimentations of the marine variety. Minerals in the chlorite family may derive directly from the source rock (detrital minerals) in that they are quite stable or may have been formed at the expense of silicates rich in Fe and Mg, such as biotite and amphiboles, in the presence of a weak washing away process. Deposits consisting mainly of chlorites have yet to be found. Location of main deposits and extraction techniques Some of the richest chlorite mineral deposits are to be found in the Apennine mountains of Emilia (Italy). The clayey part of such deposits, in fact, mostly consists of illite and, in decreasing order, chlorites. This is the so-called Complesso Emiliano, made up of: the Marls of Montepiano (red and green clays that go by the name of red-beds) that some researchers believe to be transgressive on previous formations and then dispersed by local orogeny phenomena the Antognola series (marls and grey-blue clays). Mining techniques in hilly areas see the application of ripping and scraping techniques on terraces which continues until the material is exhausted (such terraces are generally more than 20 m wide and some tens of metres long). On slopes high-power bulldozers are generally used. In semi-flat areas where the extraction zone is of considerable size, scrapers are preferred to bulldozers. This vehicle, as the name implies, scrapes long, wide strips of terrain, funnelling the material inwards for collection. 82


This system results in excellent homogenisation of the raw material, a process, in fact, aided by the vehicle loading system. Ripping and scraping has several other useful characteristics: it allows work to be carried out over a vast surface area, with large-scale transportation of materials resulting in low unit costs flexibility: unsuitable extraction zones can be isolated the same machines can be used for both handling and treatment of the extracted clay (crumbling, homogenisation, drying). This last sequence takes place in a clearing within the quarry grounds; a 15-20 cm thick layer of the material is spread out and left in the sun for a full day, occasionally being turned to ensure proper uniformity of exposure to the suns rays. Mineralogical composition of deposits containing chlorites Identification via X-ray diffractometry (Fig. 53) essentially involves observation of basal plane diffractions. The characteristic angular reflections (for Cu Kα radiation) are detected at 6.2°, 12.5° and 25.1° 2ϑ, corresponding to lattice distances of 14.3, 7.1 and 3.54 Å. The 12.5° diffraction corresponds to that of the 001 kaolinite plane which can cause errors in interpretation. To prevent this inconvenience the analysis is repeated by preheating the sample to 550-600 °C. In this case the detected basal reflections are cancelled while those at 6.2° are reinforced. DTA analysis (Fig. 54) of different chlorite types gives widely varying results; a first endothermic peak is generally noted towards 400 °C when the hydroxyls of the brucite sheet are removed. A second peak is seen between 600 and 800 °C: this corresponds to removal of the hydroxyls from the micaceous sheet. As would be expected, thermogravimetric (TG) analysis shows two corresponding weight losses. Dilatometry analysis of the unfired product fails to indicate the low temperature weight loss caused by removal of the zeolite water (unless there is substantial alteration of the structure).

83


No. 0

d(Å) 12.5

1 2 3 4 5

2.59 2.56 2.45 2.39 2.26

hkl 001 202 201 203 202 204

10

15

Fig. 53. X-ray diffraction patterns for two different, randomly-oriented chlorite polytypes: a) and b) illustrate how the two samples have significant, yet differentiated peaks with respect to the 001 basal plane and illustrate evident variations as regards the diffraction planes marked 1, 2, 3, 4 and 5.

Temperature °C

Fig. 54. DTA curves for the chlorites group show evident endothermic and exothermic peaks (curve a); as iron content increases (curve b) there is a gradual lowering of the temperature at which the exothermic reaction takes place.

84


Standard semi-quantitative mineralogical analyses of the three Emilian Apennine clay types are provided below:

1. 2. 3.

Illite xxx xxx xxx

Chlorite Montmor. xx xx x xx x

Kaol. xxx xxx x

Quartz x x xx

Feldsp. x

Carb. xx x tr.

Mica tr. tr. tr.

1. blue-grey Pliocene-Pleistocene clays 2. clays of the Antognola series 3. red-beds Observation of typical chlorite behaviour is, as might be imagined, significantly complicated by the constant presence of other clayey minerals. Technological characterisation of pressed samples prepared with chlorite raw materials In evaluating the behaviour of clays having a substantial chlorite mineral content we shall once again refer to the raw materials of the Emilia Apennines, although this is done with some reserve since these minerals are primarily made up of illite. Characteristics are as follows: normal post-pressing expansion good unfired bending strength (green and dry) at around 1020 °C shrinkage is between 2-6% with water absorption values of 13-5%. Nevertheless, in the (frequent) event that these clays are accompanied by substantial carbonate content (from 15 to 23%) shrinkage varies from +1 to 1%, while porosity lies between 22 and 15% the presence of carbonates also has a strong influence on the bending strength of the fired product; in the first case decidedly lower values (from 110 to 160 Kg/cm2) are recorded while in the second they can easily exceed 200 Kg/cm2 linear expansion coefficients for fired materials (at 1020 °C) generally settle between 25 and 75.10-7 °C-1. In practice, this figure is expressed in cubed form, generally for temperatures between 20 and 450 °C. On the whole the tiles obtained from these raw materials are quite fusible. Those without carbonates vitrify completely at 1050-1080 °C within a very tight range. Carbonate ones are slightly more refractory and become vitreous beyond 1100 °C, shrinking suddenly (a characteristic of carbonates). Post-firing colour ranges from salmon pink to yellowy to greenish-yellow (on vitrification) where clays are carbonate-rich; others have much deeper hues of red (at lower temperatures) and turn dark brown once liquid-state reactions gain the upper hand.

85


Ceramic uses Once again, Italian clayey raw materials of a predominantly illite-chlorite nature are considered. They are mainly applied in the ceramic industry as follows: grey-blue Pliocene-Pleistocene clays in high-porosity coloured bodies (majolica) for glazed wall tiling grey clays of the Antognola series for medium-porosity coloured bodies (cottoforte) for glazed floor and wall tiles; unglazed Tuscan cotto type bodies red-beds for frost-proof or low-porosity (single firing) glazed/unglazed (red gres) for residential, industrial and outdoor flooring. Pliocene grey-blue clays (i.e. those with a higher degree of natural plasticity) are widely used in the manufacture of rustic pottery (majolica), ornaments and in the production of large extruded bricks (lug bricks and ceiling bricks). Talc Origin of the term Derives from the Arabic word talq. Mineralogical structure Mineral of mica-like structure with an elementary sheet made up of two hexagonal planes of silica tetrahedrons containing an Mg2+ octahedral plane (Fig. 55). The basic structural element is neutral and highly stable. Talc, because of its structure, may be considered as belonging to the family of clayey minerals. It has a characteristic lamellar shale (talc schist) structure and is greasy to the touch. There exists another variety of more massive structure known as steatite. Chemical composition Talc is a magnesium acid metasilicate having the following formula: Mg3Si4O10 (OH)2 As no vicariant substitution with cationic elements of insufficient charge is possible, there is no introduction of alkaline cations. Nevertheless, talc rocks usually contain various impurities which affect their properties significantly. Talc is generally about 50% silica, more than 30% magnesium and a little less than 10% alumina.

86


Fig. 55. Elementary talc cell.

Genesis and characteristics of deposits Talc is of metamorphic or hydrothermal origin. In the latter case it is often associated with dolomite and magnesite. This association provides the key to correct interpretation of the genesis of this mineral, brought about by reactions between acid hydrothermal solutions (rich in silica) and the two magnesium carbonates. Sometimes, instead, it is associated with ultrabasic rocks (such as olivine, amphiboles and augite), its origin lying (as with chlorite and serpentine) in their transformation via metamorphic forces. Deposits are often stromatolith in nature. In the case of the most important Italian deposit (at Fontane in Val Germanasca, Piedmont) there is just one large strip over 2 km long, ranging in depth from 1 to 15 m: it lies sandwiched between gneiss, mica schists and limestones (at the base) and mica schists and chloro schists (above). The same band, however, also contains contaminants made up of dolomia, amphiboles and strips of wall rock. Location of main deposits and extraction techniques Deposits are found in many countries, the key ones being Germany (GbpfersgrUn and the Fichtelgeibirge hills), Italy (Piedmont, Lombardy and Sardinia), Austria, Spain (Region of Barcelona), France (the Saint Barthèlemy massif), Greece, Finland, former USSR countries (Ural mountains), USA (New York, California, Georgia, Montana, Texas, Virginia, etc.), Canada (Ontario), India, Korea, China (Szekiang and Shantung) and South Africa. Because of the above-cited stromatolith nature mine shafts need to be sunk to avoid removal of enormous quantities of overburden. The tout-venant is put through a sorting and grinding process that brings it within the particle size distribution range required by users.

87


Standard behaviour and identification X-ray identification is based on observation of basal plane diffractions (Fig. 56). The lattice distances 9.36 Å, 4.68 Å, 3.11 Å, 2.63 Å and 2.48 Å have corresponding (Cu Kα radiation ) peaks at the following angular positions: 9.44°, 18.86°, 28.62°, 34.0° and 34.5 2ϑ. Differential thermal analysis (DTA) of a pure talc shows just one endothermic peak between 900 and 1000 °C, which marks the point at which the OH hydroxyls are removed. In certain cases, the presence of chlorite and the loss of zeolite water are highlighted. The corresponding TG analysis indicates marked weight loss. Dilatometry too is often influenced by the presence of other minerals, especially chlorite, which causes a net increase in dilatation between 600 and 800 °C. 4

3.11

9.36

4

4.68

18.86

28.62

9.44

4.57

4 2.63

4

34.0

34.5

2.48

4

4 4

Fig. 56. Diffractometry (XRD) analysis of talc (t = talc).

Utilization and technological characterisation Talcs may contain a high percentage of ferrous hydroxides and can be used to produce tiles. These minerals are usually characterised by post-pressing expansion, and poor green/dried bending strength. In firing (at 1020 °C) products show high absorption (greater than 25%), slight dimensional expansion and are very fragile. The coefficient of cubical expansion for fired materials (at 1020 °C) is very high (270.10-7 °C-1). During firing, talc (especially where it has a high degree of purity and a low iron content) behaves more or less like an inert material up to about 1050 °C while at higher temperatures it begins to participate in reactions, acting as a flux. 88


Temperaturs difference

Mass %

Temperature °C

Shrinkage

Fig. 57. Thermogravimetric (TG) analysis and DTA of talc.

Temperature °C

Fig. 58. Dilatometry analysis of talc.

Given this behaviour, even small in-body percentages (2-5%) can make a useful contribution by forming, with the alkali in the feldspars, eutectic mixes with particularly low melting points. Such effects are often desirable in the manufacture of compact materials with extremely low water absorption (e.g. porcelain tiles). Ceramic uses The introduction of talc into ceramic bodies allows manufacturers to control thermal expansion. This dilatometric aspect is especially important in the manufacture of tiles. 89


At lower temperatures (i.e. before the development of massive liquid-state reactions) it is characterised by a high expansion coefficient which aids glaze-body match and, therefore, acts as a carbonate substitute. At higher temperatures, instead, talcs have the opposite effect and are introduced to improve thermal shock resistance in ceramic products that sinter at these temperatures. Talc is introduced for specific reasons in a number of ceramic materials: Porous wall and floor tile bodies: in some cases talc may be used in quantities as high as 40%. Talc-based formulations are commonly used in the USA owing to the generally low cost of the material there. Not only do they aid dilatometric match between glaze and body, they also significantly reduce moisture expansion of the fired product caused by absorption of humidity and, therefore, the risk of delayed crazing. Talc-rich bodies are particularly suitable for rapid firing in that, while maintaining optimum general characteristics, they emit only minimum gas. Generalised utilisation, though, is complicated by the difficulty of properly mixing small or very small quantities of talc into a body and, above all, by costs. Faenza pottery: these are calcareous or calcareous-magnesium compositions in which a part of the carbonates (from 10 to 15%) are replaced by talc. As a result the unfired body is more workable and there are fewer breakages as a result of thermal shock. Gres and vitrified single firing products: talc, when added in small percentages (generally no more than 5%) lowers firing temperatures and increases cycle speeds. White body compositions tend to bleach (perhaps because magnesium ferrite formations kidnap the iron). In this case, of course, talc of considerable purity must be employed. Vitreous-china sanitaryware: added in small quantities (2-3%), talc improves thermal shock resistance. Refractories: in the case of semi-silicate refractories, using talc (in quantities ranging from 2 to 10%) improves resistance to thermal shock (but also causes a reduction in refractory) and increases the bending strength of the unfired product. The use of talc is practically indispensable in the production of platelets and support items which are often subject to extreme thermal gradients, especially where fast firing cycles or shuttle kilns are used. Their utilisation is limited by the formation of eutectics, which can markedly lower their refractory characteristics. Electrical insulators: talc is the main raw material as these items are (after firing) made up of cordierite (5 SiO2 . 2 Al2O3 . 2 MgO) and clinoenstatite (SiO2 . MgO). Steatite is generally preferred for these products. Table pottery: confers translucency and makes the pieces tougher. Bricks: small additions of talc aid extrusion by acting as a lubricant. Increased bending strength of both unfired and fired items is observed, as is a reduction in firing loss.

90


Talc is also used in the production of certain magnesium glazes, while in sanitaryware manufacture it is sprinkled in the mould cavities to aid release (especially in the manufacture of large items). Many of the above-cited uses require high-purity talc, except where used in coloured tile body mixes. Uses in other industries Used as a filler in the paper, plastics and paint industries. Where of especially high purity it is often used as an alternative to kaolin. One of its better-known uses is as a support for deodorant powders in the cosmetics industry.

91


“NON-PLASTIC” MATERIALS Silica Origin of the term From the Latin word Silex, meaning hard stone. Mineralogical structure The base structure of silica consists of a tetrahedron with oxygen anions at its apexes and a centrally-positioned silicon atom. A complete tetrahedron, then, has four negative ions (Figs. 59-60-61). Anhydrous silicon has three crystalline forms: quartz, trydimite and cristobalite and a glassy amorphous (i.e. non-crystalline) one. Each main type has varieties that are stable in particular temperature ranges.

Fig. 59. Structural arrangement of silica tetrahedrons in quartz (British Ceramic Society).

Fig. 60. Crystalline structure (A) and vitreous structure (B) (according to W.H. Zachariasen and B.E. Warren).

92


Fig. 61. Trydimite (a) and cristobalite (b) structures (Worral, Clays and Ceramic Raw Materials Elsevier).

The most important transformations take place at the following temperatures: → Trydimite 870 °C Quartz Quartz α → Quartz β 573° Trydimite → Cristobalite 1470° → Trydimite β 117/163° Trydimite α Cristobalite α → Cristobalite β 220/ 270°

Fig. 62. Equilibrium diagram for silica, including the metastable portions (from Kingery, Introduction to Ceramics - Wiley).

93


Chemical formula SiO2 Genesis and characteristics of deposits Firstly, a distinction should be made between primary deposits (quartzites) and those of volcanic origin (perlites). Primary deposits appear under stromatolith form in pegmatites or schists or in zones of contact with magnetic bodies (halos). These veins vary greatly in size (from just a few centimetres to tens of metres). Of course, only the most substantial ones are of interest as only these will make extraction economically worthwhile. The purest deposits are whitish in colour with a typically translucent appearance. The presence of impurities, though, leads to clusters of varying colour (from pinkish to grey, from yellowish to brown). The term sands refers to clastic, loose sediments made up of granules between 2 and 0.25 mm in size. These, then, are relatively coarse, non-cemented sedimentary deposits homologous of sandstones. The latter are rocks with an identical granulometric structure but are distinguished by intense cementation which give them a lithic appearance and consistency. Sandy deposits may be of alluvial, lacustrine, marine or even wind-driven origin. This is, of course, a reference to the depositary environment of the particles, as in the first three cases sedimentation will naturally be preceded by fast-flowing water transport. Only a river or torrent in flood, in fact, is able to reach the speeds needed to sustain and carry the coarse particles of which sand is made. Deposition in lakes and seas always takes place close to the shore, that is, in the vicinity of estuaries where the water slows down as it disperses into the lake/sea. In seabed sedimentation, remoulding is frequent: this phenomenon involves redistribution of sediments by strong currents or tides. Fluvial depositions, instead, take place where there is a fall-off in water speed: the classic case is that of the river delta, where the river is also affected by the braking action of the tides or where the river reaches a plain and begins to meander. Smaller-scale accumulation is observed on the inner banks of river bends where water travels much more slowly than on the outer section of the bend (because of centrifugal force) where, instead, erosion takes place. Different sedimentation conditions will, of course, determine differing particlesize distribution of materials and the presence/absence of finer (clays) or coarser (gravel) contaminants. During transport, sedimentation and compaction, sandy deposits undergo a maturing process largely induced by chemical-physical aggression. The latter cause alteration, transformation and dissolution of the less stable mineralogical elements and modify the texture of the particles (dimension, form, etc.). Generally speaking, secondary silica deposits contain more non-quartzose impurities than primaries: such impurities often include micas, feldspars, oxides and ferrous hydroxides, clayey materials and vegetable substances.

94


Note, though, that the term quartzites also refers to high-purity sedimentary quartz deposits with an extremely high degree of cementation and to metamorphic deposits characterised by recrystallization of the original granules. A last, more specific type of silica deposit concerns the diatomites. These are sedimentary rocks mainly consisting of the siliceous shells of diatoms (tiny marine algae), and are characterised by high porosity and friability. Such deposits, where in fresh water, are known as fossil flours while sedimentary deposits in marine environments are known as tripoli. Location of main deposits and extraction techniques Stromatolith deposits renowned for their purity are mined in Germany (Mountains of Taunus Gebirge), Brazil (within Paleozoic formations), ex-Yugoslavia (Strumica, Macedonia) and the USA (Connecticut). Important European deposits of poorly cemented siliceous sediments are found in France (Fontainebleau), Belgium (Region of Namur) and Germany. Well-known, abundant, well-cemented siliceous deposits are found in many States within the USA (mainly along the Appalachian chain and in the South Eastern States), while some of the best-known European sites are located in England (Sheffield), Germany (again in the Taunus mountains, in Turingia and in Silesia), France (Brittany, Central Massif and Normandy). Extraction techniques vary depending on the inherent characteristics of the deposits. Stromatolith deposits can only be extracted with the aid of explosives: tunnelling may sometimes be necessary too. In the most fortunate cases extraction is effected via a series of controlled explosions across a wide extraction front. The tout-venant is then transported to processing plants where separation of siliceous and sterile fragments is effected. Sorting is often carried out manually on conveyor belts: more recently, optical recognition systems have been experimented. Explosives are also employed on metamorphic and sedimentary deposits with a high degree of cementation. Pre-mining techniques (a series of small explosive charges which shake the rock and fracture it internally) are sometimes employed, thus paving the way for ripping and scraping operations. Where the terrain contains poorly cemented sedimentary deposits, the material can be extracted simply by using earth-moving machinery such as bulldozers, scrapers and hydraulic excavators. Standard behaviour and identification Mineralogical identification via X-ray diffractometry essentially involves observation of basal plane diffractions. The specific diffraction pattern (for Cu Kα radiation) shows peaks in the following angular positions: 26.63 (corresponding to a lattice distance of 3.34 Å) and 20.84 (corresponding to 4.26 Å) (see Fig. 63). Differential thermal (DTA) and thermogravimetric (TGA) analysis of an anhydrous silica show behaviour influenced only by thermal effects connected with the restructuring that corresponds to phase transitions (Fig. 64). 95


3.34

1

4.26

1

20.84

26.63

2.45 36.54

1

2.28

2.23

2.12

1 1 39.46

45.7

50.1

1 42.45

1 1.98

1.81

1

Fig. 63. Diffractometry (XRD) analysis of quartz (q = quartz).

heating

cooling

Temperature °C Fig. 64. Endothermic effect (during heating) and exothermic effect (during cooling) in a quartz sample; the thermal inertia of the system provokes a delay in the appearance of maximum intensity, both in cooling and heating.

Dilatometry (Fig. 65) shows, during heating, a sharp increase at 573 °C. This corresponds to the switch from quartz α to quartz β , which, as is known, is accompanied by a sharp increase in volume. 96


The same phenomenon occurs even more markedly just beyond 220 °C, when cristobalite (rarely present) passes from the α to the β form. If, during cooling, not all the free silica has been bound by the neo-formation of more complexly-structured minerals (e.g. wollastonite or mullite), the passage from quartz α to quartz β is signalled by marked dimensional contraction. This is a particularly critical moment in ceramic manufacturing as excessively fast cooling can cause certain types of damage to the tile. Industrial treatment of the tout-venant Sands are generally purified and classified so as to obtain products having characteristics suitable for their intended use. Such processing plants differ according to the type of deposit and market requirements. The key stages in the treatment of sand are: washing with elimination of clayey components, attrition (rubbing the particles to detach the contaminant surface film), magnetic separation, grinding, flotation (chemical reactions with anionic or cationic reagents) and leaching (where acids attack the contaminating elements).

Fig. 65. Dilatometric behaviour in various structural forms of silica (from: Worral, Clays and Ceramic Raw Materials - Elsevier).

97


With quartzites the main (and most costly) task is grinding of the rocky fragments; this is followed by granulometric classification. Magnetic separation systems can be used to remove some impurities. However, the efficacy of these methods is frequently undermined by electrostatic charging of the dust: as a result, the use of powerful magnetic fields may cause the loss of a substantial portion of output. Companies often mine crystalline deposits containing both quartz and feldspars. In these cases treatment becomes a somewhat more complex affair, being designed to separate the two mineralogical components as efficaciously as possible. Note that some users of silica such as the glass industry demand materials of highly specific particle size distribution. Glass makers, in fact, require that all particles fall within the 0.1 - 0.8 mm range. Ceramic uses Single fired tiles, whether white or red body, use only small quantities of sand (maximum 10-15%). With coloured bodies even low-purity sands (e.g. those with a substantial iron content) can be used without any particular problems. Often, compositions for white single firing products do not include the direct introduction of siliceous sands: the same effect is attained by employing natural mixes of feldspar, clay and quartz known as partially kaolinized feldspars. Some porcelain tile formulas also make use of minimum percentages of siliceous materials. The addition of silica to vitreous-china and porcelain bodies serves the purpose of balancing the presence of SiO2 and alumina, thus providing the stoichiometric ratios, which, in firing, lead to the formation of mullite. In the tile industry the main aim is to open up the mixes to aid degassing of any impurities (organic substances) in the raw materials and removal of water during the drying process: a further positive effect is reduced contraction and deformation (during firing). Another important aspect concerns particle size distribution and the (crystalline or amorphous) state of the quartz particles. Ceramic companies tend to use already ground powders so as to reduce body grinding times. While natural silica deposits already consisting of very fine particles do exist (pyroclastic and diatomite deposits), these materials, unfortunately, have extremely negative effects on the rheological properties of ceramic slips and cannot be used. Manufacturers of vitreous-china and porcelain sanitaryware require especially pure siliceous materials. Proportions used are generally in the order of 20-25%. Earthenware pottery requires about 10-15%, while glazed white body wall tiles employ quantities in the order of 10-20%. Silica is also used in the production of acid refractories. Siliceous bricks are manufactured starting with quartzites: these are ground to the required particle size distribution, with just small percentages of binder being added. 98


Semi-silicates require the introduction of about 10% alumina; fireclays are 3040% alumina, the remaining part consisting of quartz. Diatomites are used for insulating bricks. Silica is also used in the preparation of frits where it is introduced in quantities of 20-40%; for the most part it is air classified, although certain formulations require (sometimes significant) quantities in fine sand form. Uses in other fields The main consumer of silica is the glass industry. Required raw material specifications depend on the intended final product. For instance, for standard (green) bottles sands containing even l% of Fe2O3 may be used. For semi-whites tolerance drops to 0.2%. Sheet glass will not withstand percentages above 0.3% while white glasses have a tolerance of less than 0.013%. Finally, makers of the most sophisticated (optical) glass require sands having a maximum Fe2O3 content of 0.008%, while SiO2 must account for 99.5%. A very low TiO2 content (less than 0.05%) is also highly important and chromium and cobalt must be present in the order of just a few parts per million. Silica is also extensively employed in the metallurgical industry (in special alloys). Perlite, thanks to its superb thermo-acoustic insulation properties, is used as a lightweight aggregate in the preparation of construction blocks. Finally, diatomites are used in the production of highly efficient filter beds. Feldspathic minerals Origin of the term The etymology of the terms defining the most common predominantly feldspathic rocks is as follows: feldspars: from the German Feldspat (field spar) where the term spar refers to coarse crystals that break up according to their elementary forms nepheline: from the Greek neféle (cloud); refers to the decomposition that occurs when treated with acid, forming a jelly-like cloud pegmatite: from the Greek pegma, meaning concretion aplite: from the Greek haplois (simple), a reference to its mineralogical composition felsite: from the German felsbildend meaning rock-former. Mineralogical structure The basic feldspar structure consists of a ring of four tetrahedral units (Fig. 66); potassic and sodium feldspars have three silicon tetrahedrons and one aluminum one while in calcic feldspars half the four tetrahedral units are silicon-based and half aluminium-based.

99


Potassium feldspar can appear in two crystalline forms: orthoclase (monoclinic) and microcline (triclinic). Nepheline crystallises into a hexagonal system and has a structure highly similar to that of trydimite (a silica allotrope stable between 870 and 1470 °C), from which it is only distinguished by the substitution of an Si4+ with an Al3+: that substitution is accompanied by the introduction of Na+ and, to a much lesser extent, K+, which ensures the electrical neutrality of the structure.

a

b

K Fig. 66. Elementary feldspar cell.

The crystals appear as flattened hexagonal prisms and are stable at low temperatures. Carnegieite is an allotrope form of nepheline, stable at high temperatures; it derives, by way of a process analogous to that illustrated above, from cristobalite (another allotrope form of silica). Note that nepheline is found together with alkaline feldspars in rocks called nepheline syenites, characterised by a silica content deficiency. Chemical composition The general chemical formula for feldspars is as follows: X Y4 O8 where: X is generally made up of Na+, K+ or Ca2+ Y almost always represents Al3+ and Si4+, but is sometimes partially substituted by Fe3+. 100


The three main feldspars have the following formulas: orthoclase (microcline or sanidine): K (AlSi3O8) albite Na (AlSi3O8) anorthite Ca (Al2Si2O8). Note, however, that solid solutions between these three feldspars are frequent; in particular, albite and anorthite form, at high temperature, a continuous series of crystalline solutions that remain intact even after cooling (plagioclases). These go by the following names (the albite/anorthite ratio is indicated in brackets): oligoclase (7/1) andesine (2/1) labradorite (1/2) bytownite (1/7). Orthoclase often contains, in its solid solution form, substantial percentages of albite. Na and K feldspars theoretically consist of:

Na feldspar K feldspar

% %

SiO2

Al 2O3

Na2O

K2O

68.7 64.8

19.5 18.3

11.8

16.9

The general formula for nepheline is, instead: X4 (Al4Si4O16) where X mostly consists of Na, with K accounting for no more than 1/3. Theoretical nepheline is made up of: 41.5% SiO2, 35.2% Al2O3 , 17.5% Na2O and 5.8% K2O. Genesis and characteristics of deposits Feldspar, aplite, nepheline, pegmatite, felsite etc. deposits are of igneous origin (i.e. originating from the solidification of magma fluid). There are, however, partially feldspathic deposits of sedimentary origin (sands and sandstones). In plutonic deposits (i.e. those consolidated deep underground) feldspathic rocks take a stromatolith form and are generally associated with quartz and muscovite. Genesis of these veins is thought to be connected with magmatic segregations (rich in alkalis) that stem from a particular chemical reaction or fusion and the recrystallization induced by metamorphic phenomena. For example, in the case of pegmatites (usually associated with granitoid intrusions) the development of large crystals (feldspars and mica) has led experts to hypothesize low-viscosity solutions that have saturated fractures in pre-existent rocks. 101


Aplites, characterised by much more minute crystallisation are, instead, thought to have been formed by solutions having higher viscosity, probably as a result of the removal of a good part of the volatile substances, which would otherwise have kept the solution more fluid and aided the development of phenocrysts. For example, the bulk of English feldspathic materials used to come from the region of St. Austell (Cornwall) where a granite rock (kaolinized to varying degrees) known as Cornish Stone emerges. The colour of the rock ranges from white to red/purple to tan. The partial kaolinization is thought to be the result of weathering that took place between the Cretaceous and the Tertiary periods. The genesis of feldspars found at Pinzolo (Trento, Italy) seems, instead, to be linked to stromatolith segregation from magma. The aplite of Campiglia Marittima (Grosseto, Italy) constitutes an example of (quartz-feldspathic) acid magmatic differentiation that has crystallised in the absence of volatile substances. Nephelines (or, rather, nepheline syenites) are plutonic rocks formed by the cooling of silica-poor fluids. Felsites are eruptive rocks characterised by a micro or crypto-crystalline base body with or without phenocrysts. Sanidine (a variety deformed by the orthoclase) is, instead, found in recent volcanic rocks (i.e. magma consolidated on the surface) but rarely appears in concentrations sufficient to make it economically interesting. Feldspathic sandstones and feldspathic sand deposits are mined for their high alkali content. Both are the result of weathering and subsequent accumulation of acidic source rock (granite, pegmatite and feldspathic metamorphites) in an aqueous environment (river deltas, lakes or shoreline marine areas). In such deposits the separation of feldspars from other minerals (usually quartz and mica) has already occurred naturally; unfortunately, though, transport has resulted in the addition of contaminants; a richness of clayey minerals or ferrous hydroxides and the presence of strata where heavy minerals are concentrated is common. Location of main deposits; extraction and concentration techniques Feldspathic and pegmatite veins are mined all over the world. Some of the most important are found in Scandinavia, the countries of the former USSR (Karelia, Ukraine and the Kola peninsula), Turkey, Great Britain (Cornwall and the Isle of Man), Italy (Pinzolo, Dervio and Vibo Valenzia), Germany (Oberfranken, Oberpfalz and Hagenarf-Weidhaus), France (Perpignan), the ex-Yugoslavian nations (Macedonia), USA (North Carolina, Georgia, Connecticut and California), Greece (Macedonia), Canada (Ontario and Quebec), Mexico, Japan, India, South Africa and Australia. Aplite deposits have been found in the USA (Virginia and North Carolina), Japan, Italy (Campiglia Marittima) and Germany (Weidhaus and Lesslohe). The most important nepheline syenite deposits are located in the ex-USSR nations (Karelia, Siberia and the Kola peninsula), Canada (Quebec, Ontario and British Columbia), Norway, Finland (Island of St. Jerndy), Brazil (Minas Gerais and San Paolo) and the USA (Arkansas). Given their compactness, the extraction of feldspathic and feldspathoid rocks 102


generally requires the use of explosives. In the event of partial kaolinization explosives may be limited to a pre-mining role, being used to shatter the deposit before machinery is moved in to extract it. However, machinery can sometimes be used to extract the material from such deposits directly. The tout-venant is sorted manually (e.g. as it passes along on a conveyor belt), although more recently, automatic sorting systems using optical recognition technology have been employed. After crushing, it is then sub-divided by particle size. Magnetic separators are sometimes used: these easily retain the metallic particles that stem from wear on the machinery itself, but have little effect on weakly magnetic materials such as biotite, a common contaminant. In this case, getting results means applying a more intense magnetic field, yet in doing so there is a real risk of attracting useful rocky particles that were electrically charged during grinding. Such contaminants can sometimes render the deposits unusable; a well-known case regards the above-cited Cornish Stone, where deeper and deeper extraction has transferred operations into less kaolinized areas that are decidedly richer in fluorine minerals. As is known, fluorine is a source of severe environmental pollution, so the rock cannot be used unless it is put through floatation treatment. This is so costly that the material has effectively been priced out of the market, explaining why mining in this still-important deposit has been brought to a halt. Concentration of alkali content via floatation is a widespread practice in the USA. This technique allows the quartz to be separated from the feldspar using various chemical reagents, creating special foams that remove, in suspension, one of the two minerals. With sedimentary deposits (feldspathic sands and sandstones) the most common treatment consists of washing (which removes the mainly clayey fine particles) followed by drying in a rotary kiln. Where the deposit is poorly cemented earth-moving machinery (bulldozers and scrapers) can be used to extract it without any need for grinding. If not, explosives are used in the mine, then crushers break up the fragments. Mineralogical identification of Feldspars Positioning of the characteristic diffractometry (Cu Kα radiation) peaks and lattice distances for some feldspathic minerals are illustrated in the following table: Albite: Angular position of peaks = 27.86 - 23.52 - 13.84 - 24.16 - 22.04° 2ϑ Lattice distances = 3.20 - 3.78 - 6.39 - 3.68 - 4.03 Å Orthoclase: Angular position of peaks = 27.70 - 29.94 - 26.82 - 23.54 - 21.14° 2ϑ Lattice distances = 3.21 - 2.98 - 3.32 - 3.77 - 4.20 Å 103


Nepheline Angular position of peaks = 29.46 - 22.96 - 27.08 - 21.08 - 30.68° 2ϑ Lattice distances = 3.03 - 3.87 - 3.29 - 4.21 - 2.91 Å Leucite: Angular position of peaks = 27.24 - 25.86 - 16.43 - 30.60 - 31.48° 2ϑ Lattice distances = 3.27 - 3.44 - 5.39 - 2.92 - 2.84 Å For albite, DTA readings give endothermic peaks between 820 and 900 °C, indicating its allotropic transformation. Analysis of oligoclase and labradorite gives analogous peaks between 780 and 820 °C. With nepheline, thermal activity is observed towards 1250 °C as a result of the minerals allotropic transformation. Since there are no matter-removing reactions, TG analysis provides us with no information at all. Dilatometry shows an absence of variation up to 1050 -1100 °C, after which the start of contraction which reaches its peak after 1100 °C is observed. Ceramic uses Feldspathic materials are used extensively wherever manufacturers intend to achieve a high degree of vitrification: for example, in porcelain tiles and light coloured low porosity products 25 to 55% is used. In vitreous-china sanitaryware and porcelain feldspathic materials are introduced in quantities ranging from 20 to 30% and from 17 to 37% respectively. Such percentages, of course, not only change from composition to composition but also as a function of the alkali content of the feldspathic material being added. Producers select potassium or sodium feldspar according to the desired specifications of the final product: bear in mind that potassium feldspar is a less powerful flux than sodium feldspar but provides manufacturers with a wider vitrification range (Figs. 67-68).

Height

Potassium materials are sometimes added to compositions intended to produce medium-high porosity tiles; a classic example of this is feldspathic earthenware (presently in decline).

Fig. 67. Fusibility test sequence for the Alavus feldspar as revealed by a Leitz heating microscope (softening point 1240 °C, melting point 1400 °C).

104


Fig. 68. Softening range of feldspathic fluxers identified by heating microscope (A softening point, B melting point).

Feldspar is sometimes added to lower the expansion coefficient when other raw materials tend to produce a biscuit with an excessively high coefficient that can, after glazing, take on the convexity caused by excessive glaze compression. Large quantities of (sodium or potassium) feldspar are also used in the production of frits: here, percentages vary from 20 to 40%. Nepheline can be used instead of feldspar, however, because of its expense (especially where consumers are far from its source) its application is limited to circumstances where its high fluxing properties are essential. Applications in other industries More than 50% of the worlds feldspathic materials are absorbed by the glass industry. However, this industry tends to use only materials with a high alkali and very low iron content (depending on the type of glass being produced). Therefore, floated feldspars, especially nepheline, are employed extensively. The latter contribute massive quantities of sodium and alumina, allowing manufacturers to produce a less viscous glass that is easier to shape. Pyrophyllite Origin of the term Derives from the association of the Greek words pyros (fire) and phyllon (leaf), probably owing to its tendency to peel when heated. 105


Mineralogical structure A clayey mineral with a mica-like base structure consisting of two hexagonal layers of silica tetrahedrons containing an octahedral layer of Al3+ (see Fig. 69). Like talc, the external surfaces of the particles are covered with the oxygen atoms of the SiO4 sheet, while the hydroxyls are always in the interior, protected by the two tetrahedral layers. This condition gives structural neutrality, considerable stability and chemical inertia. Chemical composition Al2Si4O10(OH)2 No vicariant substitution or alkaline/alkaline-earth cation introduction takes place. Genesis and characteristics of deposits Pyrophyllite originates from metamorphic or hydrothermal processes. It is often associated with sericite (a mineral which, in turn, derives from alteration of biotite). A soft rock (hardness scale 1) of rather high specific weight (2.8), its smooth, patinated surface makes it similar in appearance to talc. Location of main deposits and extraction techniques Relatively pure pyrophyllite beds are uncommon. Nevertheless, substantial deposits are mined in Japan, the countries of the former USSR (the pyrophyllitic schists of the Urals), the USA (between the tuffs and breccias of North Carolina), Brazil and South Africa. The latter go by the name of stumatites and can be worked to obtain objects of well-defined shape.

O OH AI 8.13Å

SI

Fig. 69. Elementary pyrophyllite cell.

106


The friability of the material allows it to be extracted using heavy-duty earthmoving machinery (bulldozers, hydraulic excavators etc.). In South African deposits extraction is effected manually and with extreme care as the aim is to obtain sizeable, intact fragments. Standard identification behaviour Diffractometry (see Fig. 70) identifies the mineral via observation of basal plane diffractions. The lattice distances of 9.19 Å, 4.59 Å and 3.04 Å correspond (where Cu Ka radiation is employed) to peaks at 9.61°, 19.31° and 29.1° 2ϑ respectively. Differential Thermal Analysis (DTA, see Fig. 71) shows an endothermic peak between 640 and 850 °C (with a maximum at 690/780 °C) owing to loss of the OH hydroxyls. At the same temperatures a steep gradient is also observed on the Thermogravimetric (TG) chart. In dilatometry analysis (Fig. 72) the mineral is characterised by a highly intense expansion phase which begins at the allotropic transformation temperature of quartz and continues at an unchanged gradient up towards 750° and then less intensely until 850 °C. Inversion of this tendency is observed beyond 950 °C. Technological characterisation of pressed samples prepared with pyrophyllitic raw materials This is a highly refractory material that recrystallizes in mullite (Al6Si2O13 rough formula or 3Al2O3 2SiO2 oxide formula) and silica at around 1200 °C and melts at around 1630 °C. Tiles made with pyrophyllatic raw materials have the following characteristics: normal post-pressing expansion poor unfired bending strength (especially the dried product) intense drying expansion.

3.34

1 0

9.19

3.03

0 0

-

12.35

9.61 8.86

7.16

4.98

+

17.7

9.97

4.59 19.31

4.25

20.85

26.6

29.1

2.45

2.55

1+ 35.12

1

36.5

2.12

1 42.4

45.4

1.99

-

1 + 0

Fig. 70. Diffractometry analysis (XRD) of pyrophyllite (P = pyrophyllite, Q = quartz, M = muscovite, K = kaolinite).

107


Fig. 71. Differential thermo-analysis (DTA) of pyrophyllite.

Fig. 72. Dilatometric analysis of pyrophyllite.

on products fired at 1020 °C expansion is intense (3.3%) with water absorption touching 27%. At 1200 °C no dimensional variations are observed and water absorption is about 19% at 1020 °C bending strength is insubstantial; only at 1150 °C is a strength of 120 Kg/cm2 attained. Ceramic uses The highly refractory nature of pyrophyllite limits its range of applications. There are few examples of it being used in the manufacture of white body wall tiles 108


or glazed floor tiles (as in Brazil, etc.). In such cases, it is used in extremely low percentages. Generally speaking, pyrophyllite lowers the expansion coefficient and limits firing shrinkage: it might, therefore, be used in the production of bodies that vitrify at high temperature (floor tiles) yet show reduced shrinkage where fired at low temperatures (wall tiles). It is also used to produce tableware, as it gives the fired product a glossy appearance. Its main use, however, is in the manufacture of refractories, insulation ceramics and refractory crucibles. Applications in other industries An alternative to kaolin, it can be used as a rubber and plastics filler and a support for insecticides and paints. In a category of its own, South African stumatite is used in the small-scale manufacture of ornaments owing to its excellent pliability. The shaped pieces are fired at 1300 °C without any shrinkage, thus yielding considerably hygroscopic manufactured items. Wollastonite Origin of the term Named in honour of the English mineralogical chemist W.H. Wollaston. Mineralogical structure Wollastonite has a sorosilicate structure where two silicon tetrahedrons connect to form a pair sharing an oxygen ion (Fig. 73). Usually of fibrous appearance (the crystals are needle-shaped or, sometimes, tabular), this mineral has a specific weight of 2.9, scores 4.5-5 on the Mohs hardness scale and is white in colour.

Fig. 73. Elementary wollastonite cell.

109


The natural form, stable at ambient temperature, changes to allotropic pseudowollastonite at 1125 °C). Chemical composition A calcium metasilicate with the following formula: Ca SiO3 Theoretical wollastonite is as follows: 48.3% CaO + 51.7% SiO2 Genesis and characteristics of deposits Although a fairly common mineral, economically viable extraction sites are few and far between, usually being generated by contact between limestone and igneous rock (contact metamorphism). Wollastonite is usually found together with calcite, garnet (alkaline-earth aluminosilicate), quartz, feldspars, granates, diopside, tremolite etc. A closely-knit association between wollastonite and other minerals is often observed, giving rise to intergrowth phenomena that makes (economically worthwhile) separation of the mineral virtually impossible. In the best-known deposit (Willsboro in New York State, USA) the lithologic situation is characterised by bands of Iskarn (a contact-formed metamorphic rock of hybrid mineralogy) mainly consisting of lime, magnesium and silica that lie against anorthosites (feldspathic rock). The wollastonite is extracted from the Iskarn via a crushing, classification and floatation process (to concentrate the effectively usable mineral) and high-intensity magnetic separation (to remove the garnet). In the Lappeenranta deposit (in Finland) the ore beds lie in the contact zone linking granite and carbonate rocks (limestone and dolomite). Veins of amphibolite and pegmatite run though these rocks: granates are completely absent. The treatment process consists a grinding sequence followed by floatation-type separation of the wollastonite from the calcite and quartz and elimination of iron via magnetic separation. Natural deposits of the pseudo-wollastonite formed by de-vitrification of glasses is, though, rare. Location of main deposits The main extraction areas are in the USA (New York and California), Finland (Lappeenranta), Mexico, Kenya, in the ex-USSR nations, India (Rajastan), South Africa, Sudan, Spain, Japan, New Zealand and Yugoslavia (Serbia). Synthetic wollastonite Made from limestone and siliceous fireclay (a kaolinitic clay of disordered structure).

110


Mineralogical composition of wollastonite rocks and standard behaviour X-ray diffractometry (Fig. 74) identifies the mineral through observation of basal plane diffractions. Where Cu Kα radiation is used the lattice distances 3.83 Å, 3.51 Å, 3.31 Å, 3.08 Å and 2.97 Å correspond to the following angular positions of the peaks: 23.2, 25.3, 26.9, 23.5 and 30.0, 2ϑ. Wollastonite is a highly stable mineral that shows no material loss or allotropic transformation below 1000 °C: consequently, no information can be obtained from DTA or TGA investigation. Where, of course, there are fairly consistent proportions of other minerals their specific thermal reactions can be observed. Dilatometric testing of unfired items shows a progressive, linear reaction of shallow gradient; the curve flattens out before 1000 °C and the first signs of the contraction phase are observed at 1020 °C. Technological characterisation of bodies containing predominantly wollastonite-type raw materials Tiles formed with (rough) wollastonite and an addition of 10% water have the following characteristics: normal post-pressing expansion minimum unfired/dried bending strength no drying shrinkage minimum firing shrinkage at 1020 °C (0.3%) very low fired product bending strength until 1060 °C where it exceeds 110 Kg/ cm2.

7

1 7

3.83

3.31

Ceramic uses Mainly used in the manufacture of floor and wall tiles.

3.51

7

29.5

3.08

2.97 30.0

32.9

7

Fig. 74. Diffractometry analysis (XRD) of wollastonite (Q = quartz, W = wollastonite).

111

23.2

25.3

26.9

2.72

2.55

7

35.1

2.30

7

2.34 38.4

39.1

2.16 2.18

77

41.69 41.3

47.4

1.91

7

7 7

7


Used extensively in the USA and Scandinavia where this mineral is available cheaply. In single firing products the addition of wollastonite improves the bending strength of the bodies, reduces firing shrinkage and speeds up kiln cycle times due to the smoother thermal expansion characteristics of the body. In porcelain tiles it is used as a bleaching agent. On unfired items wollastonite acts (like sands) as a lean raw material while during firing (beyond 980-1050 °C) it acts as a flux. Cooling can be faster as the beta to alpha quartz inversion at around 570 °C is much reduced. The introduction of this mineral in ceramic compositions also gives dimensional uniformity and aids glaze-body match. Further characteristics of ceramic bodies based on wollastonite include: low thermal expansion, lustre, a smooth surface and minimum swelling. Wollastonite is also an important component in frits (being used in quantities ranging from 5 to 20%), particularly in those used to produce sanitaryware and fine ceramics where it provides an improved melting range and increases lustre. It is also introduced, albeit in very limited percentages, into certain refractory compositions (e.g. saggers) where it is thought to improve thermal shock and impact resistance. Applications in other industries Some 50-60% of the worlds wollastonite output is consumed by the ceramic industry (bodies and frits), the remaining part being employed as a filler in the plastics and glass industry. Carbonates Origin of the term Calcite: from the Latin calx used by the Romans to indicate the CaO obtained by roasting lime. Dolomite: from the surname of the French chemist Dolomien who first distinguished it from limestone. Aragonite: from the Spanish region of Aragona where the mineral was first identified. Nature and chemical composition The main carbonate mineral deposits are made up of: Limestones: CaCO3 Dolomites: CaMg (CO3)2 Magnesites: Mg CO3

112


Crystalline structure Different carbonate minerals crystallise into the following systems: Limestones: Calcite rhombohedral Aragonite orthorhombic Dolomite: rhombohedral Magnesite: rhombohedral Aragonite is irreversibly transformed into calcite at around 500 °C. Genesis and characteristics of deposits Limestone Originates from (marine environment) chemical-type deposition of over-saturated saline solutions and accumulation of (carbonate) minerals from dead marine organisms. Genesis is aided by warm water, a high ambient temperature, ventilation and low water turnover (closed seas). This sedimentary phase is followed by diagenesis during which the original constituent fragments disappear and are replaced by recrystallized calcite. Calcareous deposits often include impurities (clay, sand) that sedimented together with the chemical precipitation and the organogenic accumulation. Concentrations of impurities are often observed in the form of lateral and vertical strips, and such contaminants (phosphates and sulphides) can sometimes render the deposit worthless. Dolomites Similar to that of limestone; however, during diagenesis about half the Ca2+ cations are replaced by Mg2+ (present in the saturating marine solutions) giving rise to regular stratification between the two carbonates. The increased pressure that accompanies accumulation aids substitution in that magnesium is smaller than calcium and is embedded more easily. Magnesites Genesis is identical to that of dolomite, the two minerals often being found together. It can be said that magnesite represents the final stage in the substitution of calcium with magnesium. Both a crystalline and cryptocristalline type can be distinguished. In the latter, genesis is thought to have its origins in the reactions between carbonate-rich water and magnesium-rich silicate rock. Location of main deposits and extraction techniques Limestone Found just about everywhere. In Europe genesis mainly dates back to the Devonian, Triassic, Jurassic, Cretaceous and Tertiary eras and has resulted in the forma-

113


tion of important parts of some of the continents mountain chains (Alps, Apennines, Carpathians etc.). Some of the more recent deposits, such as those concentrated in North Africa (Tuf) and the Caribbean (caliza arrecifal) have not undergone diagenesis and are found in a semi-coherent state. Although calcareous rocks are not particularly hard their extraction does require the use of explosives; the above-cited semi-coherent deposits can be extracted with the aid of earth-moving machinery. Dolomites Dolomite deposits (which take their name from the famous Alpine chain, itself named after the French scientist Dolomieu) are also common, their origins dating back to the Silurian, Devonian, Triassic and Jurassic eras. Like limestone, the extraction of dolomites requires the use of explosives. Magnesites The main big-crystal deposits are located in Austria, ex-USSR, Czechoslovakia, Spain, Brazil, China, the USA, Canada and Australia. Cryptocrystalline deposits are of stromatolith form and are contained in serpentines and ultrabasic rocks. Key deposits of this type are found in Greece (Macedonia), the ex-Yugoslavian nations (Balkan region), Austria (Steinmarck), Turkey (Eskisehir), India, the USA, countries formerly part of the USSR and Canada. Standard behaviour and identification Limestones Diffractometry analysis (Cu Ka radiation) reveals peaks at the following angles: 29.3, 31.3, 35.9, 39.34, 43.08, 47.42, 2ϑ corresponding to lattice distances of 3.04, 2.85, 2.50, 2.29, 2.09, and 1.91 Å (see Fig. 75). Differential thermal analysis (DTA) (Fig. 76) shows a marked endothermic peak at 920 °C that corresponds with dissociation of the mineral and removal of the carbon dioxide (CaCO3 → CaO+CO2). At the same temperature TG analysis reveals a steep gradient owing to the weight loss caused by removal of the CO2. Dolomites Diffractometric analysis (using Cu Ka radiation ) highlights the characteristic peaks corresponding to lattice distances of 2.88, 2.67, 2.53, 2.40, 2.19 and 1.80 relative to the following angular values: 30.97, 33.51, 35.34, 37.39, 41.14, 50.57. 2ϑ. DTA clearly shows how, as in the case of dolomite, dissociation takes place in two separate stages, at 780 and 920 °C respectively. Thermogravimetric (TG) analysis also highlights the removal of the carbon dioxide at those temperatures. The mineral behaves as if it were a mix of calcium carbonates and magnesium. 114


2.88

$ 3.04

#

# 2.50

2.29

#

29.33

31.3

35.9

39.34

43.08

49.42 47.42

# 2.85

# 30.97

2.67 33.51

2.40

2.53 35.34

# 2.09

1.87 1.91

##

$$$ 37.39

41.16

$ $

2.19

$

44.93

51.1 50.57

1.78 1.80

$$

Fig. 75. X-ray diffractometry analysis of dolomite (left.) and calcite (right). (D = dolomite, C = calcite).

Temperature (°C) Fig. 76. Differential thermal analysis curves for calcite (curve a) and dolomite (curve b) with in-air decomposition.

115


Magnesite Differential thermal analysis (DTA) reveals an endothermic peak connected with decomposition of the mineral and removal of the CO2 at a temperature ranging from 400 to 650 °C. Correspondingly, thermogravimetric (TG) analysis reveals a steep gradient in that zone. Influence on pressed materials Differences in technological behaviour between a carbonate-free clay and a clay with carbonates (calcite and dolomite) are particularly marked during firing. Dissociation and the release of carbon dioxide result in increased porosity and body expansion. To illustrate this the firing data for two clays from the Emilia Apennines in Italy is given below: one mix (B) is virtually carbonate-free while the other (A) has a carbonate content of about 15% (both bodies have been dry-ground). A 1095° 1055° 1010° 960° 910° CaCO3 tot.%

B

Shrink.

Water. Abs.

Shrink.

4.8 1.2 0.9 0.9 0.9

7.3 16.6 17.8 18.0 18.0

7.3 7.8 4.1 2.2 1.4

15.0

Water. Abs. 0 1.3 8.7 11.4 13.8 2.0

Clay A is a classic cottoforte composition obtained from in-quarry mixing of a highly carbonate clay (about 20%) with one that is virtually carbonate-free (the latter is type B). It is important to note that the presence of carbonates tends to narrow the vitrification range, hence it is essential that compositions for low-absorption (> 2%) bodies do not contain more than 3-4% carbonate. This level of carbonate content may be useful in removing organic substances contained in the clays, thus reducing black core and swelling defects. Ceramic uses Limestones and dolomites are essential raw materials in the manufacture of lowporosity, low-shrinkage ceramic items such as porous wall tiles and glazed single fired tiles intended for both wall and floor (monoporosa). Their inclusion (limited, though, to 20/22 and 14/16% respectively) allows bodies of suitable absorption values (especially for glazing purposes) and almost zero shrinkage (or with a maximum shrinkage of 0.5~0.6%) to be produced. This last characteristic has, in the past, allowed high-stack firing technology to be used in situations that would not be feasible if shrinkage were too high. 116


In single firing wall tiles limestone and dolomite are used (as a natural calcareous clay corrective) in CaCO3 percentages of up to 12-15%, giving the tiles, thanks to the formation of calcium silicates and magnesium, low firing shrinkage (< 1%). Their presence, however, complicates firing (especially pre-heating) of the glazes, as heat-induced dissociation of the carbonates causes CO2 emissions which must pass through the glaze without hindrance so as not to alter it and cause pin hole defects (called de-gassing effects). In this sense it is also possible to make modifications to porcelain tile bodies by adding carbonates of pre-fixed particle size distribution to the slip or the powder, thus obtaining a new body suitable for the production of quality wall tiles. The presence of carbonates, then, generally leads to an increase in the expansion coefficient, thus improving compatibility with the majority of commercially available glazes. Consequently, carbonates are used extensively in the manufacture of majolica and earthenware. Dolomite and magnesite are also used in the manufacture of basic refractory items. Carbonate-rich clays are usually used in the manufacture of extruded products. At this point it is important to point out the importance of the so-called coquinas. These are calcareous fragments sometimes larger than 1 cm. They derive from deposits of chemical origin, or sometimes originate from the shell fragments of longdead marine animals (bivalves, gastropods, brachiopods, etc.). In both cases, whenever raw materials containing these elements are extruded or dry-ground, problems can arise. With tiles, dimpled white-yellow spots appear on the body, leading to pin-holing of the glaze. In the case of bricks the coquina fragments, partially dehydrated after firing, can re-expand as a result of re-hydration or re-absorption of atmospheric CO2: if this happens after the product has been laid flaking and breakage will result, with easily imaginable effects on the appearance of walls. Applications in other industries The bulk of limestone output is absorbed by cement and lime manufacturers. For these products, a minimum magnesium content is a must. Recent years have seen high-purity limestones steadily replace kaolin as paper fillers (largely for economic reasons). Magnesite, instead, is used as a filler in the manufacture of insulating materials and as a support for fertilisers and chemical products.

117


ACCESSORY MINERALS PRESENT IN CERAMIC RAW MATERIALS Halloysite Origin of the term The name halloysite was given to a Belgian rock first studied by Omalius dHalloy. Mineralogical structure A mineral from the kaolinite family, it associates a tetrahedral sheet with an octahedral one that has the quality of being hydrous (dioctahedral) (Fig. 77). The intervening layer of water molecules gives the sheets a certain mobility. They thus tend to roll up into cylinders, as examination under an electron microscope clearly shows. This characteristic behaviour can be explained by examination of the differential tensions between the tetrahedral sheet and the octahedral one.

Fig. 77. Elementary kaolinite, halloysite and muscovite cells.

118


However, the attraction between the sheets is evidently higher than the differential tension: if the contrary were true the lattice itself would be destroyed. The cylinders can, however, unroll and flatten out, turning into metalloysite, characterised by lower water content and a lattice distance of 7.4 Å (against the 10.1 Å of halloysite). Metalloysite loses its entire water layer at around 200 °C and irreversibly changes into a kaolinite-like structure. In passing from kaolinite to metalloysite to halloysite a progressive increase in hydration is observed, accompanied by an increased degree of disorder of the lattice. It is interesting to note how crystal size in the various kaolinite-family minerals depends on the degree of stability (and disorder) of the structure. Note that the smaller particles belong to halloysite while the larger ones are seen in kaolinite. Chemical composition Al2 (X2 O5) (OH)4 2H2O X = element in a tetrahedral position (to all intents and purpose this is Si4+ only, in that substitutions with vicariants such as Al3+ or Fe3+ are almost never observed: as is known, the latter result in the structure lacking a positive charge). Genesis and characterisation of deposits Halloysite is a constituent of numerous clayey deposits. It appears as a compact, whitish mass (where free from chromophore contaminants) or may be yellow or pink. Of shiny appearance, like kaolinite, it is greasy to the touch. Nevertheless, it is not uncommon for it to be found in conditions of high plasticity. Halloysite is generally generated from volcanic source rocks (ranging from the acid to the neutral) and volcanoclasts (pyroclastic rock and volcanic breccia). Broadly speaking, the transformation of these source rocks is similar to that which leads to the formation of kaolinite. As in the latter, removal of alkalis (sodium and potassium) occurs, a process that is completed in full only in certain cases. Persistence of substantial quartz residues is not uncommon; associations with kaolinite and the presence of cristobalite and alunite (hydrous potassic sulpho-aluminate) are often observed too. Halloysite deposits may be primary (i.e. found at the location of the pre-existent source rock) or secondary (i.e. where particles have been transported and re-deposited in an aqueous environment). The genesis of halloysite deposits can often be traced to hydrothermal alteration. Location of main deposits and extraction techniques While quite common, halloysite is rarely the most important mineral in a deposit. 119


Some American States (Utah, Nevada, Idaho, North Carolina and Georgia) have particularly extensive extraction sites. The most important European deposits are found in France (Dordoigne) while minor deposits exist in Czechoslovakia and ex-Yugoslavia. Quarries are generally of the open-cast type and employ standard earth-moving machinery (bulldozers, scrapers, hydraulic excavators, etc.). Mineralogical analysis of halloysite raw materials. Standard behaviour and identification Identification via X-ray diffractometry exclusively involves observation of basal plane diffractions (see Fig. 78). Correspondence charts showing lattice distances (d) and positioning of the peaks (using Cu Kα radiation) for halloysite and metalloysite are given below: HALLOYSITE d 2ϑ

= =

7.20 Å 12.20°

4.42 Å 20.10°

3.56 Å 24.95°

4.42 Å 20.10°

3.57 Å 24.90°

METALLOYSITE d 2ϑ

= =

7.10 Å 12.40°

Fig. 78. Diffractogram pattern for halloysite, characterised by an intense reflection at about 4.4 Å, often stronger than the basal ones at 7.2-7.4 Å and about 3.6 Å, which are broad and asymmetric towards lower angles.

120


The differential thermal analysis (DTA) diagram for halloysite reveals a characteristic endothermic peak just before 200 °C associated with the loss of interlayer water; at higher temperatures behaviour is similar to that of kaolinite. A new endothermic peak starts at 450 °C as a consequence of removal of the OH- hydroxyls while at around 980 °C, in connection with the crystallisation of mullite or alumina, a distinct exothermic peak is observed. Correspondingly, thermogravimetric (TG) analysis reveals a sharp weight loss just before 200 °C and towards 450 °C, connected with removal of the inter-layer water and the OH hydroxyls respectively. No variation, of course, is noted at 980 °C as this exothermic reaction does not involve the removal of material, just the freeing of bond energy. Dilatometry analysis shows initial, weak dilatation balanced by evident shrinkage just beyond 150 °C: this is caused by the removal of interlayer water. Once this phase is complete shrinkage continues, albeit on a much shallower gradient, until just beyond 450 °C when a new, marked contraction occurs (loss of the OH- hydroxyls); that contraction persists towards 880 °C at which point sintering reactions take hold and steeper-gradient shrinkage is observed, with consequent softening of the sample. Technological characterisation of pressed halloysite materials Tiles made of a predominantly halloysite raw material would have the following characteristics: normal post-pressing expansion good unfired bending strength (green and dried) considerable drying shrinkage fired at 1100 °C, dimensional contraction is already considerable (3-8%) with an apparent porosity of 20-25% at 1100 °C bending strength is generally in the 100-150 Kg/cm2 range the coefficient of cubical expansion (for samples fired at 1100 °C) is incredibly low (even lower than 100.10-7 °C-1). Note that intense drying shrinkage makes removal of the water a real problem and that this operation often leads to the fracture of a substantial number of pieces. Ceramic uses In practice, halloysite is rarely used in the manufacture of tiles whatever their colour (usually a coarse halloysite is used in combination with other raw materials); it is, in fact, best avoided as there are considerable complications at the drying stage. It is sometimes employed in the manufacture of bricks (especially in the USA), fire-bricks and refractories (bear in mind that this is, in any case, a raw material of high alumina content). The use of halloysite in these products is justified when the plasticity of the mix needs to be increased. 121


Finally, note that this material has rather peculiar rheological properties: consequently, its use in products involving wet grinding of raw materials leads to poor deflocculation of the body. On the whole, then, utilisation of a halloysite raw material must be well contained (within 5-20%): such percentages will allow for fairly generalised use (pottery, porous and vitrified tiles). Applications in other industries In some fields it can be used as an alternative to kaolin; in particular, halloysite can be used as a filler in the manufacture of paper, tyres and rubber. Certain Fullers earths, well-known for their de-colouring power, have substantial percentages of halloysite. However, their main constituent is always montmorillonite. Attapulgite, Sepiolite, Vermiculite Origin of the term Attapulgite: from Attapulgus (Georgia - USA), where the first identified sample was found. Sepiolite: owing to its similarity with the cuttlefish (sepia) bone. Also known as Meerschaunm, a German word meaning sea foam. Vermiculite: derives from the long, worm-like filaments generated by rapid heating. Mineralogical structure Attapulgite: this is a silicon inosilicate. The tetrahedron apexes alternately point in opposing directions; this distribution is coordinated between the two chains in the sense that apexes pointing towards the other chain will find, on the latter, a specular situation. Octahedrally coordinated Al3+ and Mg2+ cations lie between the apexes of the tetrahedrons of these two chains. The structure is completed by the introduction of OH- hydroxyls at the extremity of the octahedrally-arranged cations while the interstices are filled with water molecules. Sepiolite: again an inosilicate, very similar in structure to attapulgite, yet differs in terms of the size of the platelets and the minimum possibility of cation substitution. Sepiolite platelets are 50% larger than attapulgite ones. Vermiculite: a phyllosilicate, the structure of which consists of overlaying micalike silicon units; on the whole the construction has a well-defined negative charge (as observed in chlorite minerals) balanced by the (hydrous) Mg ions inserted between the silicate layers. The magnesium may be substituted by other cations. Dioctahedral and trioctahedral forms also exist. This mineral has the characteristic, when heated rapidly, of swelling up to 50 times its initial size and giving rise to worm-like elements. 122


Chemical composition Attapulgite: (Mg, Al)2 Si4O10(OH) . 4H2O Sepiolite: Mg4Si6O15(OH)2 . 6H2O where: X is usually made up of Mg2+ Vermiculite: Mg3-x (Al,Fe3+ , Fe2+ etc.) x (Si, Al, Fe3+)4O10(OH)2 Mg0.33 . 4H2O The substitutions that lead to a deficit of valence are balanced by the insertion of alkaline-earth cations. Genesis and characteristics of deposits Attapulgite and sepiolite originate from hydrothermal processes that have been followed by transport and re-depositing. The raw material appears as a fibrous, lamellar or spongy mass of considerable size. Sepiolite not only alternates with limestone layers, but is also a filling in limestone cavities. Colour ranges from white to light pink (where dry), changing to dark pink or grey-brown when wet. Water-free fragments tend to float on water until saturation causes them to sink. The genesis of vermiculite, instead, can be traced to ultrabasic intrusions rich in biotite and pyroxines and, sometimes, to dolomite rocks that have undergone hydrothermal or weathering-induced transformations. This explains the presence of biotite relics in vermiculite deposits. The mineral generally ranges in colour from off-white to green and then brown and tends to separate along fracture lines. Location of main deposits On the whole deposits of these clayey minerals are few and far between. Sometimes attapulgite and sepiolite deposits constitute the upper part of refractory (kaolinitic) clay deposits as is the case at Sezanne (France). Another attapulgite deposit, again in France, is found at Argenteuil. The main European sepiolite deposit is located in the Spanish province of Vallecas where it consists of greyish layers several metres thick within a Tertiary evaporation basin. The deposit at Synia (Tanzania) is particularly appreciated on account of its lightness and the closed porosity of the extracted pieces. As for vermiculite, commercially-mined deposits exist throughout the USA (Montana, Wyoming, Colorado, Nevada, Arizona, North and South Carolina, Georgia and Texas) and in South Africa (Transvaal). The mineral vermiculite is found in these deposits in percentages ranging from 30 to 60%. Standard behaviour and identification Attapulgite The curve produced by differential thermal analysis (DTA) is rather complex and varies considerably as a function of the specific composition of the mineral 123


being examined. Nevertheless, above 100 °C, there is a distinct endothermic peak associated with the removal of zeolite water. This is followed by a second peak (towards 600 °C) associated with evaporation of the OH- hydroxyls. Since these are effectively weight losses they are also show up in TG analysis as the customary ondiagram gradient increases. Sepiolite The DTA curve shows a first endothermic peak at around 150 °C (loss of interlayer water) and a second at 800 °C (removal of the OH- hydroxyls), while towards 850 °C a net endothermic peak caused by neo-crystallisation of enstatite (SiO2 .MgO) and cristobalite (SiO2) is observed. Vermiculite Differential thermal analysis gives a marked endothermic peak at around 150 °C with a lower endothermic effect at 180 °C; both refer to the removal of water molecules. At 550 °C, and more markedly at 850 °C, endothermic peaks associated with hydroxyl removal are clearly observed (Fig. 81-82). All the reactions involve removal of material and consequently show up on the TG curve too. Ceramic uses Use in the ceramic industry is extremely limited. Attapulgite and sepiolite (minerals rich in magnesium) can be employed in the production of magnesium glazes. The attapulgite of Argenteuil has been used in the manufacture of Sèvres porcelain. Sepiolite is introduced into compositions for porous tiles, insulation porcelains, pottery and sanitaryware. Thanks to the noted porosity of its structure it may also be used in the production of insulating refractory bricks.

Fig. 79. A typical X-ray diffractometry trace for vermiculite.

124


sepiolite quartz potassium feldspar

Fig. 80. Example of a diffractogram for a sepiolite sample containing quartz and potassic feldspar impurities.

Temperature °C Fig. 81. DTA curves for vermiculite. The overall configuration of the endothermic effect at 100-250 °C is modelled mainly by the type of interlayer cation: A = magnesium; B = calcium, C = sodium.

125


Fig. 82. DTA analysis of Palygorskites and Sepiolites.

Applications in other industries Attapulgite and sepiolite are mainly employed as sludge in oil drilling operations. Attapulgite is also used to de-colour oils, as a support for fertilizers, in pharmaceutical products, cosmetics and as a paper and paint filler. Owing to the extreme lightness conferred by its high porosity, Synia sepiolite is, oddly enough, used to manufacture tobacco pipes and various other items, their surfaces being water-proofed by lining the pieces with Canada balsam. Vermiculite, thanks to its enormous cation exchange capacity, is, instead, used to fix radioactive elements in certain types of waste. It is also used in the manufacture of thermal insulators. Micas These mainly concern another two phyllosilicates, as well as the already-seen principle components of an illite-chlorite type, having the following compositions: biotite: 6SiO2 Al2O3 . 6(Mg,Fe) O. K2 O. 2H2O muscovite: 6SiO2 3Al2O3. K2O. 2H2O Decomposition of these via a hydration process gives rise to sericite: 6SiO2 3Al2O3 . K2 O . 4H2 O 126


The structure of these minerals is similar to that of illite; they consist, in fact, of two layers of silicon tetrahedrons enclosing a layer of aluminum octahedrons. Numerous substitutions come into play, creating electrical imbalances that are adjusted by the introduction of new ions. Biotite (or ferroan mica) has a brown-blackish, shiny appearance and is a prime constituent of granites, rocks related to feldspars: consequently, it is often encountered as a contaminant in mined feldspar deposits. The presence of biotite, within certain limits, correlates with that of anorthite (calcium feldspar). During alteration of the source rocks (weathering, contact metamorphism etc.) it can easily alter into sericite, and is thus found only rarely in sedimentary rocks. Muscovite (or potassic mica) appears in the form of metallic-like, hexagonally-shaped, shiny, flexible platelets that flake easily. These platelets can become enormous (even over 1 metre), as seen in Indian deposits. Like biotite, it is present in certain types of granite, alteration transforming it into illite; nevertheless, it is frequently found in sedimentary rocks (clays and sands), sometimes in significant percentages (even over 10%). Sericite appears as long flat greenish fragments or needles. In certain sedimentary rocks its lithoid, non-carbonate nature (as in certain sandstones) constitutes, together with the ferrous hydroxides, the cement that binds the various components. Diffractometry analysis of biotite (using Cu Kα radiation) highlights the characteristic lattice distances (10.1 Å, 3.37 Å, 2.66 Å, 2.45 Å, 2.18 Å) with corresponding peaks at the following angles: 8.70, 26.42, 33.66 and 41.40° 2ϑ. A similar analysis of muscovite gives peaks at 26.80, 8.86, 34, 88, 45, 50 and 29.82° corresponding to lattice distances of 3.32, 9.95, 2.57, 1.99 and 2.99 Å. DTA of biotite yields a completely flat line up to 1000 °C, while the same test on muscovite shows an endothermic peak between 800 and 950 °C corresponding to removal of the OH- hydroxyls (Fig. 83). Biotite constitutes a significant problem in certain feldspar deposits, where only high-intensity magnetic separators can remove it. However, it is not uncommon for the separators to attract feldspar and quartz particles that were electrically charged during grinding. In such cases a choice usually has to be made between somewhat partial separation of the biotite (resulting in commercialisation of a product that, after firing, varies in colour from grey to blackish and is blemished by black dots) or a consistent loss of useful material which the magnetic separators inevitably remove together with the mica. When present in substantial percentages, muscovite too can cause serious difficulties. A well-known example of this problem is Cornish Stone, extraction of which ceased because the deposit was steadily becoming more and more muscovite-rich. While muscovite can be split from the other components by winnowing and floatation, such processes are very expensive, thus pricing the material out of the market, which automatically selects products unaffected by this problem that do not necessitate separation systems. 127


Fig. 83. DTA analysis of micas.

As mentioned above, muscovite is often encountered in sedimentary rocks; for example, it is found in various clays of the Emilia Apennines (Italy), but in amounts too small to cause any production problems. The presence of mica in porous bodies (fired at temperatures of approximately 1000/1020 °C) is easily seen with the naked eye. Mica is rich in potassium and, on vitrified bodies, acts as a fluxer and contributes to vitrification of the material. Note that the lamellar structure of muscovite makes it difficult to grind, a characteristic that is even more marked when wet grinding technology is employed. This may lead to clogging as the slip is passed through the sieves. It is thus advisable to use sieves that eliminate the trapped material efficiently. Mineral hydroxides Mainly the hydroxides of: aluminum: gibbsite (also known as hydrargillite) represented by the formula Al (OH)3, diaspore [AlO(OH)] and boehmite (Al2O3.H2O) iron: goethite (FeO. OH). Aluminum hydroxides are the main constituents of bauxites. 128


DTA and TGA analysis of the various minerals give the following dehydration temperatures: gibbsite: from 240 to 380 °C diaspore: from 410 to 570 °C bohemite: from 450 to 580 °C goethite: starting from 250 °C. Characterised by significant colloidal content, hydroxides can cause problems during drying and preheating in that, containing high percentages of hydroxyls, they can cause dimensional settling of the pieces. Fig. 84 shows key diffractometric indices for several oxides and hydroxides. DTA results for a predominantly diaspore rock are also given, showing net endothermic peaks at 500 °C. (Figs. 84-85) Soluble salts This term refers to all those carbonates, sulphates and alkaline/alkaline-earth chlorides soluble in water. This capacity varies enormously from mineral to mineral: values are sometimes so low (in the order of just a few parts per thousand) that they are, to all intents and purposes, insoluble. Table n. 7 gives in-water solubility for these minerals. The most important soluble salts are: Ca, Mg, Na, K and Al sulphates Na-Al, K-Al double sulphates Ca, Na and K carbonates Na and K chlorides. Among these, calcium carbonate (its definition as a soluble salt considered by some to be improper) is present in all the clays used to produce porous bodies. Also worth remembering is ferric sulphate, generated by oxidation of the sulphides (pyrite and marcasite) as per the following reaction: FeS + H2O + 7/2 O2 → FeSO4 + H2 SO4 The geological history of the earth shows that the formation of sedimentary ceramic raw material deposits is closely linked to that of soluble salt minerals. The presence of these minerals is caused by precipitation conditions in sea water at the same time as fine particles from suspensions of continental origin are deposited. Hence the presence of such mineral contaminants in clayey deposits of sedimentary origin. Nevertheless it should not be forgotten that the (mostly lacustrine) continental deposits also derive from the break-up of pre-existing rocks that may, in turn, have been of sedimentary origin, thus allowing transfer of soluble salts from one generation of rocks to the next. Soluble salts show up in a variety of ways. 129


Fig. 84. List of the main aluminum, manganese, and titanium oxide and/or ferrous hydroxide reflections in decreasing interplanar spacing order.

Fig. 85. DTA analysis of mineral hydroxides.

130


Water solubility (in g/l) CaCO3 0.013 CaCl2 595.0 Ca(OH)2 1.850 0.320 CaMg(CO3)2 CaSO4 2.090 CaSO4 . 2H2O 2.410 BaCO3 0.020 BaCl2 310.0 BaSO4 0.002 0.106 MgCO3 MgCl2 542.5 MgSO4 260.0 Mg(OH)2 0.009 Na2CO3 71.0 NaCl 357.0 Na2SO4 47.6 K2CO3 1120.0 KCl 347.0 K2SO4 68.5 Tab. 7. In-water solubility of certain salts.

Distribution may be widespread but so microscopic as to be undetectable with the naked eye except in periods of drought when evaporation of impregnation water leaves them on the surfaces of clayey deposits, resulting in a characteristic whitish film. Sometimes, consistent concentrations are observed in depressions in the ground underlying the clayey formations: this is where rainwater the salt vehicle gathers. Circumstances sometimes cause macroscopic particles to form. It is not uncommon for gypsum crystals even a few millimetres wide to be found, in their characteristic lenticular, spearhead or prismatic form. This grouping also includes the coquinas rounded, lightly-coloured carbonate elements that are the bane of extruded item manufacturers. Here, we are in the field of (apparently) general or casual distribution that involves one or more veins of a formation. Sometimes, instead, deposits are concentrated in layers just a few millimetres thick which follow sedimentation planes exactly and separate two adjoining veins. However, concentrations along fractures in the deposit are far more likely. These stem from the dissolution and recrystallization caused by impregnation waters but are probably also the consequence of diagenesis phenomena, at least as far as the oldest clayey formations are concerned (those deposits have never undergone metamorphosis in that they would have been transformed into schists or shale). 131


They assume many different forms, ranging from thinly-spread films of varying consistency and hardness to protruding crystallisations, soft and fragile or compact and tough. As usual, differential thermal analysis (DTA) provides useful information as to the exo and endothermic reactions that take place during heating. Gypsum (dihydrate calcium sulphate) shows a first endothermic peak at 140 °C (transformation into a hemi-hydrate), a second, smaller one at 150 °C (transformation in anhydride) and a tiny exothermic one at 350 °C indicating the change from soluble anhydride to insoluble anhydride. At 1125 °C the anhydride changes from type α to type β (small endothermic peak) and dissociation of CaO from SO3 begins. Pyrolysis of the chlorides, instead, is complete by 700 °C; the magnesium sulphate begins to disassociate towards 900 °C, while in the case of sodium sulphate such reactions may persist beyond 1000 °C. It thus follows that neo-genesis of alkaline sulphates and alkaline-earths is almost completely annulled in ceramic bodies that sinter at temperatures above 1000 °C. Finally, in the case of ferric sulphate, pyrolysis begins towards 170 °C, while in magnesium sulphate it is only seen above 400 °C. As regards its influence on slip rheology it can be stated that the presence of dissociated sodium and potassium soluble salts aids deflocculation while calcium, magnesium and even iron ones cause flocculation. Thixotropy too depends on the surface properties of the colloidal particles of clay and is significantly influenced by the electrolytes that act on alterations between the particles. It is known that the quantity of alkaline chlorides is directly proportional to the tendency towards formation of kaolinitic clay suspension gels and inversely proportional in the case of slips consisting of montmorillonites. The presence of soluble salts can thus produce opposing effects which depend on the clayey minerals present. The firing behaviour of the main soluble salts is characterised by pyrolysis, with consequent elimination of the gaseous phase and dehydration phenomena. For example, the loss of the gypsum water molecules occurs at low temperature, overlapping water loss in expandable-lattice clayey minerals, thus contributing to total shrinkage of the piece at the drying and preheating stage. The consequences of the above-cited phenomena increase in significance as firing cycle times shorten. Certain reactions, almost insignificant in traditional tunnel kiln cycles, cause huge problems in the rapid and ultra-rapid cycles of single firing items, difficulties that can often only be resolved by eliminating certain raw materials. Where rapid cycles are employed a further complication is posed by the resistance of the glaze, as it covers a body that has not been stabilised by a first firing. Pyrolysis reactions involving the removal of gas when the glaze has high fluidity, or the employment of unsuitable glazes, give rise to small craters on the face of the tile that damage the finished item irreparably. 132


Single firing glazes will only let gas bubbles pass through where fusibility is extremely high or extremely low; in the former case low-temperature fusibility allows re-closure of craters opened up by bubbles breaking the surface. In the latter, the glaze only fuses at high temperature and thus preserves an earth-like porosity well into the firing process, allowing gas bubbles to pass through without hindrance. Some authors claim that soluble salts cause black core problems, yet the source of such defects is actually the presence of non-mineral carbon and iron. The main effect of soluble salts on unglazed fired pieces is the appearance of efflorescence. This term refers to deposits that settle on the surface of the biscuit (sometimes on the dried product too) by way of the mineral salt precipitation that occurs under saturation conditions. This phenomenon, generally seen on bricks, but also on both pressed and extruded tiles, only occurs when the ceramic body is porous enough to allow, within its interior, migration of salt solutions. On reaching the surface, these give rise to rapid water evaporation. The raw material must, of course, have a relatively high content of such salts (and alkaline oxides and lime) or they must form during firing by way of pyrolysis of the pyrite or the presence of sulphur in the kiln fuel. The phenomenon is perceptible as soon as fired products contain more than 0.5%. Efflorescence is, in order of frequency (on dried and fired products), caused by sodium sulphates, calcium, potassium and magnesium, the carbonates of these metals and chlorides and alkaline nitrates. On fired items sulpho-aluminates and alkaline carbonates are also observed. The formation of efflorescence is greatly influenced by factors affecting the movement of saline solutions inside the fired bodies (dimensions and distribution of pores) and the way in which water evaporation occurs: bear in mind that, under certain circumstances, this takes place not on the surface but in the interior of the ceramic body. Where salts migrate to the surface of pressed tiles (generally seen on double firing wall tiles) there may be an accumulation of alkali-containing salts in those zones subject to greater heating during drying (i.e. tile edges). The extreme consequence of this phenomenon, also caused by re-wetting of the biscuit during glazing, is the alteration of the chemical composition of the glazes at the edges of the tile via increased contribution of fusible salts; it makes this part of the tile, already subject to greater interaction with the heat of the kiln, even more sensitive to temperature variations and, on average, more fusible. This can lead to the side of the tile having an overly-fused, clearer-coloured appearance generated by air bubbles in the glaze. A similar phenomenon generated by salt migration is seen on porcelain tiles. This defect is generated by salts building up at the tile support points during drying, thus giving rise to zones with a greater post-firing vitreous phase content of shinier appearance. In both cases the problem can be resolved in addition to attempting to elimi133


nate salt-carrying raw materials by making drastic changes to the drying cycle, accelerating it to the limits of piece bending strength so as to aid high-speed water evaporation and thus prevent migration and accumulation of salts on certain parts of the tile. Vegetable substances and reducers Vegetable substances (some examples are given in Fig. 86) found in clayey raw materials can be divided into two main groups: non-carbonized and carbonized materials. The former include roots, wood fragments, leaves and humic acids; these elements are largely present in recent deposits or those connected with the pedolith layer, or occur simply as a result of incomplete clearing of vegetation prior to extraction.

Fig. 86. Organic substances found in clays.

134


These vegetable substances are normally eliminated during the firing process, sometimes with the aid of oxidising substances specially added to the body: they start combusting at around 250 °C and stop at approximately 450 °C. The latter group consisting of various types of carbon and generally found in older geological formations is considerably more stable. Normally immune to treatment with oxidants, these materials will, during firing, burn away in the 300 to 600 °C range. The presence of vegetable substances causes various difficulties in the production of both single and double firing items. In fast single firing especially where cycles are less than an hour, iron minerals are present (red body) and manufacturers aim to produce low-absorption bodies centrally-placed, black ellipsoidal stains (black core) can form and, in the most serious cases, there is intense swelling. These result from incomplete combustion of the organic substances, giving rise to graphitic carbon which blackens the ceramic body, while the CO2 from the burnt part, unable to pass through the piece, causes swelling. Furthermore, where inorganic salts are present, Na2SO4 salts also form: these contribute to the formation of vitreous phases, blackened or darkened by the presence of chromophore cations (e.g. Fe, Cu, Cr etc.). Where mineral sulphides (e.g. iron or copper sulphides) are present complex reactions take place, giving rise to combinations between the silica and sulphide decomposition products (Fig. 87) and resulting in a black, highly fusible glaze.

Fig. 87. DTA analysis of mineral sulphides.

135


FeS2 + O2 → 4FeS + 9O2 → 2FeSO4 + ½O2 →

FeS + SO2 2Fe2O3 + 4SO3 Fe2O3 + 2SO3

[350 - 450 °C] [500 - 800] [560 - 775]

The actual appearance of such phenomena will also depend on other factors such as powder grain size and moisture content, pressing force, firing curves and in-kiln atmosphere. Negative effects can be reduced by: introducing lean (non-plastic) raw materials into the composition (sands, lapilli, chamotte) adding oxidants to the composition (MnO2, various nitrates) increasing the particle size of components and powders reducing the moisture content of the powders (but taking care not to unduly lower the dry bending strength) lowering pressing pressure applying a more appropriate firing curve that makes the 250-600 °C interval last as long as possible establishing an in-kiln atmosphere that is as oxidising as possible. Sulphur and sulphides (alunite) Certain feldspathic or clayey materials sometimes contain sulphurs which vary in colour from whitish to yellow. These are generally raw materials of an origin associated with volcanic areas. Materials of this type found in Lazio (central Italy) are a classic example. Sulphur gives rise to a number of problems: during firing it turns into sulphur dioxide and corrodes the internal structure of the kiln: sulphur dioxide also gives rise to problems of an environmental nature both inside and outside the plant. Many of the clays used to manufacture tiles (e.g. those from the Emilia Apennines in Italy) contain ferrous sulphides (pyrite and marcasite). Pyrite causes formation of the classic black dot, that is, on-biscuit mini-craters with a dark core, which, in turn, lead to specks in the glaze. The mechanism that gives rise to this is thought to be linked to the random presence of pyrite crystals close to the surface of the bodies. At a temperature of around 400 °C an oxidising process transforms the sulphur in oxide with removal of the SO2: the chemical process is accompanied by an increase in volume and causes flaking of the small fragment covering the crystal, thus leading to mini-cratering. Alunite is a hydrous potassium sulpho-aluminate defined by the following formula: K2O(Al2 O3)3 (SO3)4. 6H2O This mineral is generated under hydrothermal conditions that often lead to the kaolinization of feldspathic veins. Thats why it is usually found together with kaolin, as seen in the famous kaolinitic deposit of Djebel Debar (Algeria). Observation 136


of a DTA chart for this material shows a first endothermic peak at 550 °C (water loss) while a second (at 850 °C) indicates the removal of a good part of the sulphur dioxide that accompanies the formation of alumina and potassium sulphate. The exothermic peak visible just after 700 °C has yet to be fully understood. The freeing of sulphur trioxide involves problems cited previously. Vitreous materials Found in clayey or sandy deposits when, during genesis, volcanic materials are introduced at the same time as sedimentation of the water-suspended particles. When small, such products may be transported (by wind) a long way from the point of origin (i.e. from the volcanoes). The characteristic vitreous state of these materials is unstable and tends to evolve towards the crystalline; however, it is a transformation that requires an extremely long time to develop and reach completion. This explains why vitreous particles of this type are still found in sediments that are not particularly recent (as with certain red-beds of the Emilia Apennines). The presence of these materials gives rise to melting dark dots on the biscuit surface; with certain glaze types, this problem can become more marked and lead to production problems.

137


138


Chapter IV RAW MATERIALS FOR FRITS AND GLAZES

It is common to apply a surface coating to tiles which, when fired, produces a vitreous layer that is hygienic, easily cleanable and provides aesthetic qualities. Bearing this in mind, let us now take a look at the most common techniques and materials used in such technology. Firstly, it is important to define just what is meant by the term frit (a name for a major component in most glazes. A frit is obtained by melting a mixture of materials in either continuous (basin) furnaces or intermittent (rotary) furnaces. Glazes contain combinations of one or more frits with other additives to obtain various effects to be described later. The pre-melting or fritting of materials serves to produce an amorphous, homogeneous compound, free from any non-dissolved raw material residues or bubbles, which, on subsequent grinding will become even more homogeneous. Fritting aims to: 1) make certain components (boron derivatives, alkaline salts, lead derivatives) insoluble: these would otherwise dissolve during the grinding stage if used in their raw state. 2) completely eliminate all organic impurities, thus removing all volatile components via reactions that would otherwise occur during firing of the finished product. 3) disperse certain impurities (of ferrous and/or metallic nature) evenly within the mass: even where such impurities are present in very low percentages they can still cause local defects. Melting ensures that these contaminant particles are dispersed throughout the molten mass and incorporated in the composition, thus eliminating defects such as iron, copper specking etc. 4) aid reactions at high temperatures (i.e. 1400 °C or more), forming vitreous compounds that would otherwise develop at standard body firing temperatures. Raw materials for melting of frits In selecting raw materials for the production of ceramic frits the following criteria must be taken into consideration: consistency of chemical composition over time consistency of particle size distribution over time low Fe and Cr content absence of difficult-to-melt minerals (kyanite and sillimanite) that, remaining in the frit as unmolten elements, compromise its quality 139


last but not least, materials should be economically viable and readily available.

The most important and thus the most commonly used materials are: quartzes and sands sodium feldspars potassium feldspars boric acid pentahydrate borax colemanite (calcium borate) anhydrous borax (sodium borate) ulexite (sodium and calcium borate) calcium carbonate zirconium silicate zinc oxide barium carbonate dolomite kaolin.

In lesser quantities, the following products are also used: potassium nitrate sodium carbonate potassium carbonate magnesium carbonate titanium dioxide.

Such a variety of raw materials means there is a need to make basic choices in terms of storage and handling solutions. It should be noted, in fact, that some of the above-cited raw materials cannot be stored in silos (or, rather, not in a way that is straightforward). Moreover, materials such as barium carbonate, titanium dioxide and potassium nitrate are not available in bulk. Zinc oxide differs from other materials in that, in its light form, it cannot be stored in, or, rather, extracted from, silos: consequently, it is now common practice to employ so-called heavy zinc oxide, a material without any particular handling problems. Given the above difficulties and the simple fact that they are often utilised in small quantities these raw materials are, then, generally stored in sacks rather than silos, while others may be delivered in bulk by truck. Information regarding the type of minerals from which these materials are extracted, refining processes and chemical/particle size distribution limits that make these compounds suitable/unsuitable for fusion is given below (see p. 145).

140

Raw materials for frits and glazes

Technological aspects of frits and glazes Final ceramic surface coatings are, then, made up of thin vitreous layers. They are applied in aqueous suspensions obtained by water grinding the frit and any other raw components in the composition onto fired or dried unfired ceramic bodies. A certain glass type, known as cristallina, is, on some wall tiles, applied on already glazed and decorated tiles to obtain brighter and deeper colour effects. Another characteristic glaze is maiolica, highly opaque, usually applied thickly to give a rich white and glossy tile surface. The glaze get its opacity from the frit opacifier, usually, zirconium silicate. There are also frits for specific, finely-targeted use, such as those employed in monoporosa. In addition to opacity and brilliance, these must also have very high softening points (1020-1050 °C) so as to favour emission of the gas (CO2) emitted from the body during firing. With glazes, instead, the diffusion of vitreous single firing products has meant that, in addition to the concept of opacity other factors such as surface appearance (matt, semi-matt) should be considered. This chapter will deal first with the theoretical aspects of the nature of the glass, and then provide a review of the different types of frit and, finally, glaze compounds.

Glass formation and formulation mechanisms All vitreous masses come from the fusion of various constituents (see Tab. 8): vitrifying agents fluxes stabilizers opacifiers de-vitrifying agents.

141


Tab. 8.

The key physical characteristic of glasses is that they are isotropic while crystalline-structured solid bodies are known to be anisotropic. In the past, this led to the belief that glasses were amorphous substances; however, recent studies by Zachariasen and Warren have established that the characteristic tetrahedral coordination of silicon is maintained in glass too. While these tetrahedrons are, in crystals, arranged as a strictly regular geometric construction, in glasses they are arranged chaotically, without periodicity or symmetry. So while we can speak of a glass lattice, it must be born in mind that this is a disordered, contorted lattice essentially made up of silicon and oxygen. Two-dimensional illustrations of the tetrahedral arrangement in crystal-type silica and molten silica are shown in figures 88 and 89 respectively. Like silica glass, common glass also features this irregular arrangement of tetrahedrons; moreover, in the latter the ions of the other constituent elements fill the gaps left by the silicon and oxygen. The bonds in the glass lattice are not all equivalent as they are in crystalline lattices; consequently, the energy needed to break them apart is differentiated. Thus, as temperature rises the level of thermo-agitation energy increases until a level sufficient to break the weakest bonds is reached. Continued raising of the temperature results in the gradual breakdown of the lattice and the corresponding, progressive liquefaction of the glass. 142


Fig. 88. Crystalline structure.

Fig. 89. Vitreous structure.

For every temperature, then, there is a corresponding glass structure that characterises it. During solidification, as the temperature gradually falls, the glass reestablishes the bonds and tends to take on the structural state associated with lower energy levels. However, as the increase in viscosity around the transition point is somewhat rapid, the internal state of the solidified glass corresponds to that of higher temperatures. Consequently, structural instability arises: while this instability necessarily tends to evolve it does so over a very long time interval (Tab. 9). Anorhtite Gehlenite Sphene Gahnite Willemite Cristobalite Tridymite Spodumene Magnesium titanate Wollastonite Rutile Cordierite Forsterite Enstatite Diopside Zirconium silicate Zirconium oxide Celsian Leucite Tab. 9. Some crystalline components in de-vitrified frits.

143


The cations, which, in the oxide state, can be obtained in the vitreous state simply by heating and are thus known as lattice-forming cations are Si4+ and B3+. While silicon forms apex-linked tetrahedrons, the coordination 3 boron forms equilateral triangles at the centre of which lies the B3+ ion. Since the silicon ion has four bonds, while the boron ion has just three, it can be seen how boric glass is less viscous and thus more fusible. The fluxing cations, also known as lattice modifiers, split the links between the tetrahedrons once they are added in the oxide state: O O O Si O Si O O O

by introduction of Na2O

O O O Si O O Si O O Na Na O

These ions generally take up positions in the interstices between the siliceous polyhedrons. The greater the number of sodium ions introduced, the greater the number of splits, and so on, thus diminishing the viscosity of the glass. Furthermore, the high number of splits between the tetrahedrons ends up compromising the existence of the vitreous state itself because the more freedom the tetrahedrons gain, the more marked is their tendency to take on the regular structure of crystals and, consequently, of devitrified glass (Tab. 9). O O O Si O Si O O O

by introduction of CaO

O O O Si O O Si O O Ca O

Stabilising cations are lattice modifiers too: unlike alkaline cations, which, because of their weak ionic potential are only loosely linked to the lattice and thus easily removed, resulting in alteration of the glass, alkaline-earth cations have twice as much ionic potential and therefore reinforce the lattice structure of the glass and act as stabilisers. The replacement of a modifier ion (Na) with another of higher electrical charge (Ca), having more or less the same dimensions, causes: an increase in density, because the greater attractive force exerted on the adjacent oxygen ions gives rise to greater compactness an increased refractive index as a result of the increased density a reduction in electrical conductivity stemming from reduced mobility of the cations, in turn caused by increased bond energy increased viscosity for the same above-cited reason. It was previously affirmed that boron is a lattice-former, with a coordination number of 3. However, we know that it is never used as the sole vitrifying agent, while it is used extensively in siliceous glasses.

144


Therefore in exclusively boric glasses the lattice is made up of equilateral triangles, apex-connected by oxygen atoms which act as bridges; in silica-boric glasses, instead, increasing quantities of B2O3 first lead to the formation of BO4 tetrahedrons (boron coordination switches from 3 to 4), giving rise to a structure analogous to that of a strongly silica glass. When the B2O3 exceeds a certain threshold the triangular structures characteristic of pure boric glass, rather than BO4 tetrahedrons, are formed. Aluminum ions cannot, in themselves, be called lattice-formers; nevertheless, studies of feldspars have shown that they can, in the presence of electropositive ions, substitute the silicon ions, forming a corresponding number of tetrahedrons. Aluminum may behave similarly in the glass and this reinforcement of the tetrahedrons makes the glass more viscous, more chemically resistant and the vitreous state more stable. This behaviour of aluminum means that a net distinction between lattice-forming and lattice-modifying ions cannot be established. Under certain conditions even the latter may exist as lattice-formers. Numerous examples of such dual behaviour are provided by the coloured ions, depending on whether they exist as formers or modifiers. In this respect it can generally be stated (Dietzel) that cations having high coordination strengths with respect to the oxygen anion (Tab. 10) behave as latticeformers (Si4+, B3+) while those with the lowest values act as lattice modifiers (Pb2+, Ca2+, Ba2+, Li+, Na+, K+) and, finally, those that have intermediate values may perform both functions (Fe3+, Be2+, Mg2+, Ni2+, Zn2+, Co2+). Standard raw materials and their influence on the characteristics of glass 1 - Silica (SiO2) Introduced in the form of quartz, quartz sands, feldspathic sands and feldspars. Silica is the prime component of vitreous compositions as it has the property of vitrifying under the effect of the fluxes within a very broad temperature range. The fluxes or modifiers are: PbO, B2O3 , K2O, Na2O and Li2O. Glazes rich in silica are highly resistant to chemical agents and are extremely hard. The higher the silica content in a glaze the higher its firing temperature. 2 - Diboron Trioxide (B2O3) Introduced as boric acid, sodium borax, colemanite. Because of its vitrification properties, boron is, after silica, the most important element. However, it cannot be used on its own as it would give rise to strongly soluble glasses . It acts as a flux in silica glasses; indispensable in glasses that are lead-free with low melting points; dissolves many colorants, confers glossiness, reduces viscosity and lowers the expansion coefficient in those glasses to which it is added. 145


Degree of oxidation

Dissociation energy (Kcal/g.atom)

Coordination

LATTICE FORMERS

INTERMEDIATES

MODIFIERS

Tab. 10. Bond strengths calculated for certain oxides.

3 - Lead monoxide (PbO) Lead monoxide gives glass: high fusibility an increased refractive index increased density increased glossiness.

However, lead glazes also have: low viscosity 146

Strength of single bond (Kcal/g.atom)


high toxicity, proportional to lead content and as a function of the form in which the glass is bound high sensitivity to acid attack where oxide content exceeds a certain proportion. 4 - Alkalis (K2O, Na2O, Li2O) Introduced as nitrates, chlorides, carbonates or feldspars. Alkalis are lattice modifiers: their introduction weakens the lattice structure of the glass by lowering the melting point. The Na+ and K+ ions occupy positions in the interstices separating the tetrahedrons. The K+ ions, which are larger than the Na+ ions, form stronger bonds, thus giving rise to the easy alterability of sodium glasses. Highly sodium glasses are easily soluble. Alkalis generally increase glass expansion coefficients, except for lithium, which, as it is highly fusible, can produce the same results while being used in extremely low percentages (much lower than sodium or potassium). Alkali, especially lithium, give the glasses gloss; however, on their own they are unable to constitute the entire basic part of a glass composition owing to a tendency towards devitrification and the solubility of the formed silicates. 5 - Calcium oxide (CaO) Introduced in the form of calcium carbonate, dolomite, wollastonite, anorthite. Calcium oxide is a stabiliser: added to an alkaline silicate it eliminates the alterability of the glass. On its own it would form silicates with a high melting point (above 1400 °C); mixed with other silicates it gives rise to the formation of vitrified masses. It is thus obvious that high percentages of this oxide would lead to devitrification (matt CaO). Introduced in the right proportions (5-10% by oxide analysis), the calcium not only gives stability but also improves bending strength and body-glaze adhesion. Also lowers viscosity in glasses fired at high temperature. 6 - Alumina (Al2O3) Introduced as calcined alumina or alumina hydrate, feldspars, kaolin, corundum. In glazes in appropriate proportions (4-8%) it confers, in the case of low-temperature glazes, the following: increased viscosity reduced tendency to de-vitrify (crystallization) increased bending strength reduced expansion coefficient increased resistance to acids improved opacity (introduced in high percentages in concordance with the glaze firing temperature).

147


The quantity of Al2O3 introduced into a glaze increases or decreases in proportion to the firing temperature. Quantities will thus be higher in matt or satin-finish glazes and lower in high-gloss glazes. The quantity of Al2O3 introduced into a glaze also depends on its particle size distribution. As the particles become finer the percentage that can be introduced drops, while larger particles allow the introduction of higher quantities. Since it is an amphoteric substance this oxide has the capacity to combine as much with silica as with basic oxides. It is thus the most efficient of the stabilisers. 7 - Barium oxide (BaO) Normally utilised by introducing barium carbonate (BaCO3) into the composition. This oxide increases density and refraction, thus giving the glaze shine. It is also an excellent flux in the fusion of silicate glasses, a property that, in part, allows it to substitute lead oxide effectively; it is, however, highly toxic. Where present in significant percentages (over 0.3% in molecular equivalents), this oxide hardens glazes and induces de-vitrification. A barium glaze melts much more rapidly than a calcium one and is less viscous too. 8 - Magnesium oxide (MgO) Introduced by means of dolomite, magnesium carbonate and talc. Magnesium oxide behaves in glasses in much the same way as calcium oxide. The only difference is that it gives rise to more viscous glasses. It cannot be used in overly high percentages as it would otherwise raise the firing temperature of the glass. The magnesium reduces the expansion coefficient yet heightens the surface tension of the glass. 9 - Zinc oxide (ZnO) In acid glazes with a high alumina content, zinc oxide plays a fluxing role. Depending on the percentage used, this oxide has a range of effects: a) in low percentages: increases the brightness of glasses and colours except for greens and blues; together with alumina it improves the opacity and whiteness of the glazes as long as CaO content is low: in the absence of B2O3 it reduces the expansion coefficient. b) in high percentages: devitrifies from the vitreous mass, giving the glaze surface a characteristic matt finish, brought out where the glaze is basic. c) in very high percentages: crystallizes and individual crystals made up of ZnO silicates separate. Glasses rich in this oxide are extremely vulnerable to acid aggression. 10 - Titanium dioxide (TiO2) Titanium dioxide improves chemical resistance and offsets crazing. This last effect immediately becomes apparent even where only small percentages of the oxide are employed and remains steady even as percentages increase.

148


Adding TiO2 colours the glass: even at just 2% yellowing is observed. Simultaneously the surface of the finishing glaze takes on a matt appearance, becoming hard and coarse as oxide percentages increase. It has opacifying properties that improve in the absence of B2O3 and especially where the glass composition is rich in Al2O3 and added in the mill; under these conditions the colour fades. These characteristics are particularly evident if TiO2 is introduced into the glass as anatase, while, where introduced as rutile, it loses its de-vitrifying characteristics until, in high percentages at high firing temperatures, it gives rise to needle-shaped crystals. Crystallisation occurs mainly in high-fusibility glasses. 11 - Stannic dioxide (SnO2) Although a superb opacifier even at low percentages (6-10%) it is rarely used owing to its high cost. The opacity stems from suspension of the oxide in the vitreous mass as finely dispersed particles. Its opacifying potential thus depends on the purity of the oxide, the fineness of its particles and the nature of the vitreous mass to which it is added. Alkalines and boron may prove detrimental to good opacification of this oxide. The tin is best introduced into the mill during grinding stage and not in frit melting. 12 - Zirconium dioxide (ZrO2) Used in the form of zirconium silicates of varying particle size distribution. An excellent opacifier (not quite as efficient as stannic oxide but much more economic) and undoubtedly the one preferred by industry. High percentages of this oxide raise the firing temperature of the glass into which it is introduced. Zirconium silicates lend themselves as opacifiers in all glaze types firing between 940 and 1300 °C. Only a part of the introduced zirconium silicates combines with other components, the majority remaining as it is. The part that combines heightens the crazing resistance of the glaze. Zirconium also has the property of being a colour stabiliser. Calcium and/or barium oxide (no more than 0.2 molecular equivalents) help the zirconium improve opacification as do zinc and alumina. Three different types of zirconium silicate are commercially available: these largely differ in terms of particle size distribution: micronized zirconium silicates (very fine) zirconium silicate powder (coarser) zirconium silicate sand (very coarse). The most commonly used glaze opacifiers are micronized products, while powders are used in frit melting and sands are used as hardeners or fillers. Low firing temperature glazes, opacifiers with zirconates, often have non-smooth surfaces, undoubtedly in relation to their highly viscous nature. 149


Table 11 provides an overview of the mineral types allowing the introduction of different oxides. Table 12, instead, shows the most significant ceramic oxide parameters. Such parameters are: – molecular weight – surface tension – expansion coefficient. Types of frit As seen, the term “frit” is commonly used to indicate, in industrial production processes, a vitreous mix cooled suddenly in water.

TABLE SUMMARIZING THE COMMONEST MATERIALS FOR THE GLASSES AND GLAZES COMPOSITIONS Oxides

Used raw materials

SiO2

Quartz, Feldspars, China-clays Quartz, Feldspars, China-clays, Zirconium Silicates

B2O3

Boric acid*, Borax*, Colemanite

PbO

Minium, Litharge

Na2O

Feldspars, Borax*, Na2CO3, NaCl*

K2O

Feldspars, KNO3*

Li O LiO 2 2

Feldspars, Li2CO3

CaO

Wollastonite, CaCO , Feldspars, Dolomite Wollastonite, CaCO , Feldspars, Dolomite, Colemanite 3 3

BaO

BaCO3

MgO

MgCO3, Talc, Dolomite

Al2O3

Al2O3, Al(OH)3, China-clays, Feldspars

ZnO

Zinc oxide

SnO2

Tin oxide

TiO2

Titanium oxide, Rutile sand

ZrO2

Zirconium silicates, Zirconium oxide

* materials soluble in in HH22O, O,therefore thereforeusable usableininfusion frit materials soluble melting only only Tab. 11.

150


TABLE SUMMARIZING THE CHARACTERISTICS OF THE MAIN CERAMIC OXIDES Oxides

Molecular weight

Superficial tension

Expansion coefficient x10-7 ENGLISH TURNER

WINKELN SCHOTT

MAYER HAVAS

SiO2

60.1

3.4

0.15

0.8

0.8

B2O3

69.6

0.8

-1.98

0.1

0.1

PbO

223.2

1.2

3.18

3

4.2

Na2O

62

1.5

12.96

10

10

K2O

94.2

0.1

11.7

8.5

8.5

Li2O

29.9

4.6

-

-

-

CaO

56.1

4.8

4.89

5

5

BaO

153.4

3.7

5.2

3

3

MgO

40.3

6.6

1.35

0.1

0.1

Al2O3

101.9

6.2

0.52

5

5

ZnO

81.4

4.7

0.21

1.8

2.1

SnO2

150.7

-

-

2

TiO2

80.1

3

-

-

4.1

ZrO2

123.2

4.1

0.69

-

2.1

SrO

103.6

-

-

-

CoO

75

4.5

-

-

4.4

Fe2O3

159.7

4.5

-

-

4

MnO2

86.9

4.5

-

-

2.2

CuO

79.6

-

-

2.2

Tab. 12.

Frits are used as bases in finishing glazes and in low-temperature glazes to render the components insoluble. The market offers a wide range of frits having different fusibility, glossing, opacification and matting performance specifications. They can be grouped as follows: 1 - Glossy transparent frits a - for traditional double firing (a slow glost cycle) Characterised by low temperature fusibility. They are made up of a high per151


centage of SiO2 (50-60%) and a low percentage of fluxing elements (20-25%, including Na2 O-K2 O-PbO-B2O3). The rest of the composition consists of stabilisers (Al2O3-ZnO-CaO-BaO-MgO), almost always accounting for only very limited percentages (max 7-9%). These frits are mainly used in the preparation of transparent glazes although they are occasionally employed in low percentages in certain glazes fired at low temperature. At temperatures above 1100 °C these are employed albeit in low percentages in the preparation of almost all glazes so as to improve vitrification and confer fusibility. b - for fast double firing Frits used in fast double firing are substantially different from the previous types in order to provide the thermal specifications needed in the much faster firing cycles. The need to mature the glass in just a few minutes has led to the formulation of entirely new glasses. More specifically, the silica-boron-alkali-alkaline-earth ratios have been changed. Figure 90 illustrates the modifications to oxide content needed to make frits suitable for rapid firing. c - for monoporosa For porous fast single firing technology, the frit will have a radically different composition on account of specific needs of the bodies, which generally have a high carbonate content (8-14%). This has thus resulted in the development and fine-tuning of frits with high melting points that soften above 950 °C so as to allow complete emission of the carbon dioxide formed during decomposition of the calcite and/or dolomite. Examination under a heating microscope must show such frits to have a very high softening point together with a sphere point close to that of the previously described frit. To satisfy these requisites and aid/trigger this eutectic fusion it has been necessary to increase CaO, ZnO, MgO and BaO oxide content (with respect to double firing). At the same time the percentage of fluxing oxides has been reduced. The goal was to produce fast frits (i.e. capable of melting suddenly at high temperature). 2 - Glossy opaque viscous frits (known as zirconium whites or majolicas) a - for traditional double firing The only difference between these and the frits in the previous grouping is that they are opacified. In terms of characteristics and composition they are identical. Opacification is achieved with the aid of zirconium silicate, introduced into the composition in quantities ranging from 8 to 14%. 152


Na2O

→

decrease

K2O N

→

increase

Na2O+K2O

→

decrease

CaO

→

increase

ZnO

→

increase

BaO

→

increase

MgO

→

increase

B2O3

→

decrease

SiO2

→

decrease

Fig. 90. Oxide content changes to adapt frits to fast firing.

These frits are mainly used in the preparation of glossy white glazes of both the high and low-temperature firing type. Of course, in glazes fired at high temperature the percentage of frit diminishes considerably in favour of the raw materials. In the preparation of glazes other than glossy whites these frits are used only occasionally. b - for fast double firing Essentially the same as glossy transparent frits. Here too, rapid firing requires substantial modifications to formulation in order to make the fusibility compatible with rapid firing. The expansion coefficients for white, opacified and transparent frits suitable for fast double firing are slightly higher than those used in traditional double firing (linear coefficients range from about 60 to 70 × 10-7). 153


c - for porous single firing Apart from the utilisation of zirconium silicate, the concepts expressed above for the glossy frits apply. 3 - Matt frits (CaO-ZnO-TiO2) a - for traditional double firing This group includes all those frits where a component added in large quantities to the appropriate vitreous base shows clear devitrification characteristics. The devitrifying elements are: calcium, barium, zinc and titanium. The first two normally de-vitrify in alkaline-boric vitreous bases while the zinc and titanium devitrify in lead bases. Calcium and barium matts are normally lead-free; these frits are also viscous and opaque. Zinc matt frits are, instead, slightly fusible, leaded (PbO 25-30%) and semiopaque. Titanium matt frits are also slightly fusible, leaded, covering (opacified) and are always yellowish in colour. Normally used in the preparation of matt glazes or as correctives in other, not necessarily matt glazes. In many cases it is preferable to introduce ZnO-CaO-BaO-TiO2 in the frit as opposed to using the respective raw materials. This is to avoid the use of hygroscopic raw materials (ZnO) or those rich in volatile substances such as the CO2 contained in carbonates (CaCO3-BaCO3). b - for fast double firing These frits differ substantially from their traditional firing cousins in that while with the just-cited technology it was possible to employ formulas with high percentages of free alumina and zirconium, this is not possible in rapid firing. c - for porous single firing No substantial formulation differences exist between these frits and those suitable for fast double firing, except for high lead content (>10%), not used in single firing. 4 - Glossy, transparent medium fusibility frits These differ from crystalline viscous frits in that they are more fusible. The silica percentage, in fact, drops to 35-50% while the quantity of fluxing elements (Na2O-K2O-PbO-B2O3-Li2O) increases until it accounts for 30-40%. These frits are extensively employed in the preparation of virtually all glazes fired at low temperature. They are also occasionally used in low percentages in the preparation of several special glazes that fire at high temperatures (e.g. leather and pearled glazes). The widespread use of these frits stems from their fusibility which favours inmill introduction of high quantities of raw material and devitrification of all the 154


matting elements. In this way a range of glazes with significant technical and aesthetic differences can be obtained simply by using just one frit and varying inmill addition. 5 - Fluxing frits (leaded and lead-free) These frits feature high fusibility, hence the fluxing tag. Depending on the employed fusible element, they may be leaded (lead silicates) or lead-free (boric-alkaline or alkaline-boric fluxes). In some glazes these frits are used, in small percentages, as correctives, the aim being to add the fusible elements that would otherwise be impossible to introduce as raw materials in that they are water soluble (alkali and boron) or toxic (lead). The frits belonging to this group are used extensively in reagent silk-screen colours, in certain preparation bases and in some spray gun flame effects. Use of these frits diminishes rapidly as glaze firing temperatures rise and at high temperatures is abandoned altogether. Lead borosilicate fluxers and lithium lead-free fluxes belong to this group. Note that the lead borosilicates include Monoboron and Lustre (blue) frits, the approximate composition of which are illustrated below: MONOBORON:

PbO 68-70% B2O3 15-20% SiO2 10-15%

LUSTRE:

PbO 40-45% B2O3 18-20% SiO2 33-38%

In addition to high fusibility, these frits are characterised by strong reactivity, which shows up as a marked tendency to attack and penetrate the bodies and all the glazes that come into contact with them during firing. These frits are used only in certain reactive glazes fired at low temperatures and are always introduced into the glaze in very small percentages (except for special torn, stone and streaked glazes). Their use in the preparation of particularly reactive silk-screen colours is, instead, widespread. Where glazes and colours are fired at high temperatures these frits are used rarely, if at all. 6 - Frits coloured on fusion These frits differ from those in the preceding groups only in that they are coloured. In fact, were it not for the colouring they could quite easily be classified within groups 2 and 3. The usual colouring elements are iron, cobalt, manganese, copper, cadmium and selenium. 155


Cadmium and selenium frits are used on their own to obtain the respective glazes, unable to be produced in any other form. Coloured frits are only used to obtain certain coloured transparents or to introduce colorants in a stabilised form in flame or disc applications. Tables 13 and 14 show approximate compositions and key characteristics of the most commonly used frits in traditional/fast double firing and monoporosa technology (i.e. transparent, glossy and matt white). Classification of ceramic glazes A great many tile types produced using a wide range of technologies exist. Depending on that technology, glazes will have specific characteristics, their chemical make-up and physical parameters varying greatly. There are also a series of type-based requisites dictated by the market. The various commercial tile types require, in fact, specialised glazes to produce the desired aesthetic effect. Highly specific production technologies often correspond to well-defined, diversified commercial types. However, glazes that can satisfy several production technologies are much sought after. For clarity we shall illustrate the double classification system, which shows the correlation with both the production technology and the aesthetic needs of the marketed products (Fig. 91). Production technologies have, over the years, evolved as a function of that equilibrium which involves all the factors typical of industry in general. These include: technology renewal and relative amortization costs competitive benefits in terms of quality and aesthetics benefits in terms of potential added value and output general/economic conditions in the factory general state of the economy.

156


GROUP 2

GROUP 3

GROUP 4

ELEMENTS

A

B

C

A

B

C

A

alkali

8

6

4

8

7

5

1

1

1

12

3

6

9

5

11

1

6

13

28

MgO

CaO

3

8

BaO

3

3

ZnO

1

10

PbO

B

C

40

40

11

11

2

4

15

Al2O3

8

8

9

6

7

6

1

B2O3

19

6

3

15

11

5

10

6

6

SiO2

58

54

60

58

52

52

40

43

43

9

11

8

ZrO2 TiO2 Firing T

970

1040

1110

970

1040

1110

970

1040

1110

Softening T

890

950

1020

920

940

1040

750

1100

1100

Semi-sph. T

1160

1150

1120

1240

1210

1140

1000

1150

1150

Firing cycle

10h

40min.

50min.

10h

40min.

50min.

10h

40min.

50min.

Frit kiln T

1350

1450

1470

1350

1450

1520

1300

1350

1350

CT

CR

CM

BT

BR

BM

MT

MR

MM

Group 2

glossy transparent frits

Group 3

glossy opaque frits

Group 4

matt frits

A

used for traditional double-firing

B

used for fast double-firing

C

used for monoporosa

Tab. 13. Composition and characteristics of the most common frits.

157


GROUP 5

GROUP 6 A

ELEMENTS

B

GROUP 7 C

D

alkali

3-7

10-15

CaO

3-6

2-10

BaO

0-2

ZnO

3-6

PbO

30-40

Al2O3

2-4

B2O3

5-10

SiO2

40-50

E

GROUP 8 F

35

75

25

85

15

H

I

0-2

6-10

2-4

16-20

0-5 65

G

20-30 69

45

2-4

25-35

0-5

25-35

0-2

2-5

2-4

20-30

20

20

2-4

6-12

30-40

40-50

11

35

35-55

50-60

30-40

0-5

ZrO2

7-10

TiO2 Group 5

glossy transparent frits at average fusibility *

Group 6

(lead and leadfree) fluxes *

Group 7

reactive fluxes *

Group 8

frits coloured on fusion

A

bisilicate

B

rocaille

C

monosilicate

D

leadfree flux

E

monoboron

F

lustre glaze

G

ZnO matt

H

CaO matt

I

TiO2 matt

(* Use limited to low-temperature firing) Tab. 14. General characteristics of frits.

158


TRADITIONAL DOUBLE-FIRING

FAST DOUBLEFIRING

Õ

zirconium white glaze

Õ

alkaline transparent frit

Õ

engobe

Ô

Õ

matt

Ô

Õ

glazes for silk-screen printing

Ô

particular glazes

Ô

waxy glazes

Ô

majolica glazes

Ô

glazes with rustic and antique effects

Ô

cotto-like glazes

Ô

crystalline glazes

Ô

technical glazes

Ô

POROUS SIGLEFIRING

VITRIFIED SINGLE-FIRING

Fig. 91. Types of market-available glazes correlated with product-specific needs.

These factors have given rise to the present-day situation where different technologies are used to produce commercially overlapping products. Such technologies include: traditional double firing fast double firing porous single firing (monoporosa) vitrified single firing. 159


Each of these technologies, which can themselves be split into sub-groups, requires glazes of specific chemical-physical characteristics. Traditional double firing This technology virtually the only one up until the 1960s involved dry ground and pressed bodies fired in tunnel kilns at around 1000-1150 °C in a 15-30 hr cycle. The obtained ceramic biscuit, known as cottoforte, was characterised by high water absorption and porosity. This was sent on to the glazing line and then fired again at around 950 °C to 1050 °C in a further 12-hr cycle. Until just a few decades ago virtually all commercially available tiles glossy, matt, rustic and surface-effects glaze were produced this way. To avoid giving an out-of-date classification of double firing glazes, note that the glazes now employed in nearly all double firing operations are the following: Glossy zirconium white A white glaze almost entirely consisting of the characteristic frit, it has a high viscosity which gives mirror-like, covering surfaces. Plastic raw materials such as selected high quality white clays and/or kaolins are added in percentages varying from 3 to 10% so that grinding provides stable slips that do not settle quickly. Aqueous suspensions of this type, mostly made up of vitreous particles, often require a 0.01% - 0.3% addition of ionic systems, which, acting on the electrolytic characteristics of the vehicle (water), modify the rheology of the slip. Depending on requirements, they can also exert a suspending or fluidising action (increase or reduction of slip viscosity) to make application easier. A zirconium white should be selected mainly on the basis of the firing conditions dictated by the available technology. It is good practice to use whites having a sphere point close to maximum kiln firing temperature, in that the thermal change rate in kilns of this type allows us to hypothesize very low product inertias. No special preheating behaviour is required in that the body is inert. Linear expansion coefficients must fall between 55 and 60 × 10-7 °C-1 to maintain glaze compression and prevent crazing in the more challenging areas of product use. Transparent glaze This product is normally used to manufacture tiles with a transparent, deepglaze look. The application sequence usually involves a layer of engobe made up of frits of the zirconium white and/or transparent alkaline type to which feldspar raw materials, zirconium silicates (to improve covering), clays and kaolins (to adjust rheology) may be added on top of which the glaze is applied. In addition to these components, double firing engobes may also contain other 160


frits having specific functions: as fluxes to improve anchoring to the body, or special frits to increase the expansion coefficient of the layer of glaze in order to prevent excessive tile convexity. After the engobe the transparent glaze may be applied directly or there may be an intervening second layer of glaze which regulates the fusibility between them. Then the film of true transparent glaze is applied. Silk-screen printing can be before or after the transparent layer to obtain a deep or surface effect. Chemical-physical parameters are the same as those for zirconium whites, except for the absence of zirconium silicate which gives the finishing glaze a transparent appearance. Any introduction of additives is effected as with zirconium whites. In quantitative terms it should be born in mind that the engobes and semi-glazes usually have different viscosity characteristics from those of whites and transparents in that they have a higher plastic raw material content (clays and kaolins). Matt glazes These cover a decidedly smaller wall tile market share (10-20%) than glossy glazes. They are formulated as a function of devitrification with the development of a matt surface. Devitrification is generally caused by an excess of calcium or zinc. In this firing technology both alumina and zirconia can also be used as they contribute to the creation of a matt surface in addition to conferring their other typical characteristics. Zinc, titanium, calcium, barium and magnesium oxides matt by crystallisation while aluminum oxide and sometimes zirconium silicate matt by hardening. Glazes matted with zinc and titanium oxides normally consist of a fusible vitreous base and tend to be leaded. Aesthetically, they are an imperfect white, tending to grey where zinc is present and yellow where titanium is included. Glazes matted with alkaline-earth oxides are always white in colour and are normally highly viscous. Glazes matted by hardening always consist of fusible vitreous bases (see frit group 4) thoroughly hardened with alumina, corundum and zirconium silicate. When the matting element is alumina or corundum, satin-finish glazes are obtained while, with zirconium silicate as a matting element, stone-like glazes are obtained. Low-temperature matt glazes (ZnO-CaO) are directly made with the relative frits (see frit group 6) in high percentages: at higher firing temperatures, raw materials added to the mill are used. One of the most frequent problems is a varying degree of matt across the kiln cross-section. In some cases partially glossy tiles are produced where in-kiln burners are too close to the product and create microclimates unfavourable for devitrification. Matt surfaces are also influenced by in-kiln cooling conditions. Gradual cooling, 161


in certain cases, aids development of devitrification. The final aesthetic effect is provided by the successive applications of silk-screen products having chromatic or gloss-matt contrast effects. Engobes These are applied before the glazes and have a wide range of functions. For example, engobes can homogenise the degree of water absorption in subsequent applications. They are thus able to reduce the dimples caused by inconstant absorption across the surface of the body. Engobes also have the function of slowing down water absorption, thus preventing the formation of air bubbles in successive applications. Ceramic functions include the slowing down and, above all, splitting up of the gas bubbles that rise up from the body during firing because of pyrite, hematite or chalcopyrite (coquina) clusters in impure bodies. A further function is that of isolating the colour of the body, covering the base and thus making subsequent use of coloured transparent glazes or coloured preparation bases possible without any interference with the (often dark) body colour. Note also that engobes aid adhesion to the body without there being any problems of aggression with the latter and they also modify the average expansion coefficient of the vitreous state so as to avoid the problems cited earlier. Perfect calibration of all these chemical and rheological parameters with the optimum use of frits and raw materials is thus highly important. Today, more knowledge and skill is required for proper formulation of engobes than glazes themselves. The normally-used frits belong to the category of glossy whites, alkaline transparents, fluxes (used to a lesser extent) to aid adhesion and special high-expansion frits (to modify the dilatometric characteristics of the system). Overall percentages vary from 40% to 80%. Raw materials used to complete formulation all of which must be high quality include: zirconium (for covering performance), good quality feldspars and quartzes and plastic raw materials (for rheology) such as clays and kaolins. In this case very good fluidisation is required to ensure a good on-product spread during application. Fast double firing Note that this technology was largely developed to reduce the high production costs associated with traditional double firing; that goal was achieved by introducing a whole series of automation and firing technology changes that have since gone on to characterise the entire ceramic industry. Employment of new kiln construction materials with low thermal inertia and the adoption of single layer roller systems (thus doing away with the need for supports) allowed manufacturers to develop firing cycles (min. 25 minutes ~ max. 60 minutes) according to product size without failure of the ceramic body. Costs thus 162


fell on all fronts but glaze suppliers were forced to reformulate their entire range of products. Zirconium whites These glazes, even when used in a fast double firing context, continue to consist of at least 90% frit. However, the base frit making up these glazes has been substantially modified to adapt thermal characteristics to the much faster firing cycles. As regards the other technological aspects it can be said that the concepts expressed in the description of traditional double firing zirconium whites remain valid. Remember that it is, here, more difficult to attain high opacity owing to the difficulty in achieving a good degree of glaze maturation. Applying an increased percentage of zirconium and balancing out its refractory effect with increased boron and alkali content fails to provide good results in that it can cause partial solubility of the frit. Alkaline transparent glazes Essentially the same as zirconium whites. However, fast firing requires substantial formulation changes to make product fusibility compatible with short firing times. To ensure that the glaze spreads properly, it is good practice not to economise on the gram weight of the glaze itself or that of the engobe. Use of high quality raw materials in the transparent glaze guarantees results exempt from any partial opalescence or accentuated straw-yellow colouring even where thickness is high. Matt glazes Like the above glazes, these too differ substantially from the corresponding traditional double firing products. Observations as to the required modifications largely correspond to those given in previous sections. There is no progression in passing from a glossy surface to a matt one: the switch from satin-finish surfaces to gypsum-like ones takes place without any intermediate values. Experience teaches us that the best way of obtaining devitrified surfaces is to use calcium even in very high percentages introduced in the form of wollastonite or to use specific frits. Introducing calcium as a carbonate or as a dolomite is, instead, inadvisable, in that CO2 bubbles produced by thermal decomposition may be trapped. Zinc is also capable of giving very soft matt surfaces, but obtaining a glaze that remains inert in the face of chemical aggression is virtually impossible where this oxide is present in substantial percentages. Fluidity should be better than it is with traditional double firing glazes. It is possible to have both opacified and transparent products although control of crazing is more difficult on the latter. Application of an engobe is normally required. Engobes Fast double firing employs higher-expansion engobes to ensure flatness or slight convexity of the tiles: therefore more use is made of special high-expansion frits. The remaining parameters are similar to those of traditional double firing engobes. 163


Special glazes A whole range of wall tile glazes that were extremely successful in the past have now lost ground in terms of commercial potential to white and semi-transparent glazed products. A brief list of these follows below (Fig. 92). Waxy glazes Semi-covering glazes of average fusibility. Normally formulated with semi-fluxes to which small percentages of zirconium, zinc, etc. are added. These were used with coloured preparation bases and applied using techniques that led to uneven thickness. They are now mainly used in single firing and are currently making a weak comeback in both traditional and rapid double firing. Consequently, their composition is generally as follows: medium-fusibility vitreous base: type 4 frits or a mix of group 1 (A and B) and group 5 frits

waxy glazes

streaked glazes

torn glazes

iridescent

GLAZES PARTICULAR FOR DOUBLE-FIRING

flaked glazes

glazes with rustic effects

pearled glazes

crystalized glazes

leather glazes

Fig. 92. Different types of glazes once used in double firing (from the 60s to the 80s).

164


slightly opacified with zirconium silicate (4-8%) slightly matted with TiO2-ZnO-SnO2. Streaked and torn glazes These glazes have almost all disappeared from the technological-commercial scene. They were characterised by floating refractory islands usually made of zirconium which came to rest above the normally leaded fusible part, thus creating wide, irregular far-apart grooves in torn glazes and fine, closely-bunched furrows in streaked ones. These glazes are obtained by saturating highly fusible vitreous, reactive bases with high percentages of zirconium silicates, almost exclusively at low firing temperatures. To improve or enhance the torn effect a thin film of a fusible preparation base should be applied under these glazes. Iridescent glazes Glazes containing special components that give surface iridescence. Used rarely. The iridescence was obtained from boric vapour fallout (using leaded frits with a high percentage of boric oxide known as blue frits) or by using specific heavy metal oxides such as V, Bi, Mo, W. Flaked glazes Highly fusible glazes made up of special oxides (such as that of cerium) which tend to generate split-ups and produce an effect where white flakes are immersed in the transparent glaze. Used rarely, as they tend to craze. Rustic-effect glazes These glazes are currently regaining ground in the wall tile business while in the flooring sector they are already used extensively. They contain hard raw materials of large particle size such as zircon sands and corundums which give the tiles a rustic appearance. All those glazes having varyingly fusible vitreous bases strongly hardened with coarse materials can be classified in this group. These materials (various sands and corundums) are normally introduced into the composition at the end of grinding. Pearled glazes Glazes having a soft surface and slightly gloss-matt nuances that evoke the appearance of a pearl. These are normally leaded and opacified with oxides (e.g. stannic oxide). Not used on todays market. These could be categorized as matt glazes, but since they have rather unique composition and reactivity characteristics they deserve a separate description. A fusible vitreous base forms the necessary starting point for the production of this glaze (see group 5 frits for low firing temperatures). This base is then opacified with micronized zirconium silicate in medium-low percentages and then matted with the oxide mix (ZnO-TiO2-SnO2). 165


Leathery glazes Glazes resembling worked leather in terms of texture and colour. Normally of leaded formulation, they are rich in stannic oxide, zirconium silicate, and contain a little titanium dioxide. Colouring is provided by an orange pigment, accompanied by ferrous oxide. These were often applied on markedly leather-coloured bases and used flame effects to improve the overall chromatic effect. Used extensively until just a few years ago they have now all but disappeared, although they are still used in specific niches of the wall tile industry. Aventurines Glasses characterised by minute crystals in suspension (Fe-Cr-Cu), made highly fusible by a lead monoxide and/or alkaline-boric compound-based composition. Crystallization occurs by saturation (during firing) of these finishing glazes on the part of an oxide and relative separation of the latter during cooling. Preparation bases Made up of strongly-coloured, varyingly fusible finishing glazes. Never used on their own but always beneath other glazes. Porous single firing wall tiling This technology stemmed from appreciation of the technological benefits that came with the advent of single firing vitrified products. The goal (achieved in full) was to obtain products that would enjoy the cost-cutting benefits associated with a single body+glaze firing sequence while maintaining the characteristics required of wall tiling: low weight per m2, limited thickness, well-controlled sizing and flatness, aesthetic results on a par with double firing. Bodies consist of a mix of typical carbonate clays, or white firing clays mixed with calcite, quartz and feldspathoids. Once dried and glazed, single-stage firing takes place. Firing temperatures generally oscillate between 1060 °C and 1120 °C (depending on whether the body is red or white) while firing cycles last from 30 to 50 minutes depending on tile size. Zirconium whites With respect to fast double firing formulations it became necessary to increase percentages of oxides such as CaO, ZnO, MgO and BaO in order to promote and trigger this eutectic fusion. At the same time percentages of fluxing oxides such as alkalis and boric oxide were reduced. The aim is to have rapid frits that melt suddenly at high temperatures. The same concept is equally valid for the alkaline transparent frits. As regards other characteristics the observations made for fast double firing apply. Rheological aspects require, in addition to fluidisers, the introduction of 0.3% of CMC (carboxymethyl cellulose) to help bind the unfired glaze to the body. Without the CMC adhesion during preheating would be insufficient. 166


At present, low-viscosity CMCs are employed to reduce or even eliminate the need for fluidisers. Alkaline transparent frits The same concepts expressed in the above paragraph apply. In this case too a percentage of CMC (0.2-0.4%) is added to the suspension. Matt glazes Formulations differ little from those used in rapid double firing. The additives used in the whites and transparents are applicable here too. Engobes In this case too, the observations made vis-à-vis fast double firing apply. Perhaps more so than in other cases (i.e. fast and traditional double firing), much attention needs to be paid to viscosity. Appropriate percentages of fluidising materials (0.05-0.3%) are, in fact, added to the engobes because they are formulated using percentages of clay and glaze weights higher than those used in fast double firing. Silk-screen printing products Nearly all wall tiling products, whether they employ zirconium white, alkaline transparent or matt glazes, employ silk-screen or rotary decorating techniques to meet aesthetic demand. A specific series of glazes has thus been developed for this purpose. The glazes in question are dispersed with short or medium-chain polyglycols to give pastes just the right viscosity and simultaneously ensure decomposition of the polyglycols themselves in the first stage of preheating, thus avoiding interference at the glaze softening phase. The most frequently used glazes are the following (see Fig. 93). Range of colours These are coloured glazes and coloured transparent frits that can be applied under various conditions. To obtain neutral effects alkaline transparent bases are generally used. For slightly deeper effects the use of semi-fluxing materials is preferred. Subsequent definition of matrix parameters refers to the number of threads per cm. Silk-screen printing matrices are generally made with nylon, polyester or, in special cases, steel fabric. Highly-covering white glazes Formulated with leaded bases that have been opacified with ceric oxide and stannic or zirconium oxide. The matrix fabric mesh varies from 45 to 75. 167


range of colours

highly-covering white glazes

alkaline transparent frits

white relief glazes

SILK-SCREEN PRINTING GLAZES FOR ALL TECHNOLOGIES

heavy glazes

lustre glazes

iridescent glazes

reactive glazes

Fig. 93. Glaze types most frequently used as screen printing bases (all production technologies).

Alkaline transparent glazes Used for gloss-matt effects. The matrix fabric mesh ranges from 21 to 77. White relief glazes Formulated using zirconium whites with the addition of opacifying raw materials: zirconium, zinc, tin. Fabrics with a low number of threads per cm2 are used. Their use in third-firing is currently popular. Heavy glazes Used for incision effects on zirconium white or transparent glazes. Formulated with vanadium frits or bases with high PbO percentages. Printing is usually effected with matrices having 51-77 threads per cm2. 168


Lustre glazes Formulated with alkaline bases to which tungstic oxide or metallic tungsten are added. Can also be formulated with special cerium frits. Matrices with 45-100 threads per cm2 are used. Iridescent glazes Formulated with boric bases or heavy metals such as molybdenum, bismuth, vanadium. Matrices with 21-45 threads per cm are used. Reactive glazes Used under the glaze (not just in screen printing) to obtain a mixing effect with the overlying glaze during firing. Of high fusibility and applied using matrices having 18-45 threads per cm. Single firing floor tiling This is currently the most widespread floor tile production technology. In the Sassuolo area this technology was largely based on the use of local dry or wetground red clays for the higher quality products. Red bodies were then replaced by wet-ground, spray-dried ones and, more recently, these have, in part, been replaced by stoneware or porcelain. Over the years the technology has been refined to provide products of evermore sophisticated aesthetics and ever-better quality. Glazing is effected on a hot body after treatment in a drier (80-120 °C). It is essential to glaze the tile while it is hot enough to evaporate the water in the glaze suspension and thus aid drying. The addition of carboxymethyl cellulose in the glaze suspension is indispensable in ensuring proper adhesion of the glaze-body interlayer during preheating. Light coloured bodies are generally fired at a temperature of around 1160-1220 °C with cycles lasting 35-55 minutes. Red bodies are generally fired at temperatures of around 1120-1130 °C with cycle times similar to those cited above. Finally, high-porosity red bodies are fired at 1080-1100 °C with firing cycles of 30 to 50 minutes. Descriptions of the most commonly used single firing floor tile glazes are given below. Waxy glazes Refers to a category of glazes used extensively in single firing of a silky matt appearance. Their covering capacity lies midway between that of a white and a transparent. Sometimes, completely transparent products are also defined as waxy (but only in single firing). Fusibility is always medium-to-high and under-glaze decorations or applications may be applied. Above-glaze applications are generally limited to silk-screen deco169


ration only. CMC is added to these glazes to improve on-body adhesion. Formulation generally makes use of averagely leaded frits. The added raw materials, which make up 10-30% of the glaze are: zirconium silicates, zinc and stannic oxides, feldspars, wollastonite, clays and kaolins. Opacified white glossy glazes These differ markedly from their double firing counterparts. Generally formulated using frits with a high calcium or zirconium content and very low levels of lead. Added raw materials generally consist of micronized zirconium silicates, clays and kaolins. These glazes can be applied immediately after the engobe and before application of other glazes (waxy, crystalline etc.). Decoration is always performed on top of the glaze. Application is generally of the disc type. Rustic-effect glazes In recent years this category has become extremely diversified. These glazes generally have non-gloss surfaces, are soft to the touch and opaque to reflection of light or rough and strongly satined. They may also be gypsum-like and markedly opaque. Formulation varies widely. The employed frits are of the semi-fluxing type, such as, for example, disilicates, accompanied by various other frits. Added raw materials too constitute a highly varied group, ranging from corundums and sands to confer hardness and coarseness to zinc oxide, tin and wollastonite for softness (calcium is also introduced as carbonate and/or dolomite). Feldspars and nephelines are also employed to give the glazes glassiness and clays and kaolins to regulate rheology. Zirconium silicate is used in low percentages. Covering capacity is highly variable and depends on the need to utilise highly elaborate (and often under-glaze) applications. Items produced with these glazes are commercially referred to as rustic. Such products may require up to 15 successive applications that involve disc application, brushing, airbrushing, washing with water, air-jet blowing and dry applications. The resulting aesthetics are superbly sophisticated, especially when one bears in mind that pre-structured bodies can extend decorative opportunities even further. Cotto glazes These glazes were created to imitate classic, unglazed cotto. They have since been diversified into a range of products to meet more specific aesthetic needs such as rustic cottos etc. Composition largely falls into line with that of the rustic glazes, yet a wider range of colours is available. These products are normally off-shade and matched with rustics to obtain floors that fall within the extensive rustic category. Application is as described for rustics. Crystallised glazes Currently regaining lost ground after a prolonged absence from the market. Formulated using fluxing frits and high percentages of titanium dioxide. The 170


above-cited oxide crystallizes during firing under the form of rutile. The use of coloured bases gives rise to deep-colour effect products. The greatest inconveniences stem from the level of crystallization which is difficult to keep constant. Modern fast firing processes should aid uniformity of firing across the entire kiln cross-section. Engobes These serve the same purposes described for the other technologies. Formulated with a high percentage of clayey raw materials and kaolins. Coloured bases May be formulated with fluxing frits coloured on fusion, mixes made with high percentages of pigments and colorant oxides such as ferrous and manganese oxide. These coloured bases are used to obtain special under-glaze effects by way of their reactivity with the overlying glazes. The applied glaze layer can also be brushed to concentrate the colour in structured body cavities, thus creating highly interesting deep-colour effects. Matt glazes These glazes are formulated using frits of high calcium, zinc and/or magnesium content. Semi-fluxing frits are sometimes used too, while zirconium silicate, wollastonite and plastic materials are added in the mill. Matt glazes are generally applied using the disc method. Technical glazes These glazes feature excellent abrasion resistance and porosity both on the surface and inside the vitreous mass. Generally formulated using hard raw materials such as corundums. Special frits that give rise to devitrification-type crystallisation during firing are sometimes used too (glass-ceramic). Application begins with a specific white base followed by fairly thick (e.g. 12-thread) silk-screen printing that provides the product with marked technical characteristics. Improvement of on-product aesthetics is achieved via widespread utilisation of coloured silkscreen printing products applied between the base and the thick technical glaze layer. Dry-application grains Dry application techniques have come a long way over the last few years: these were originally conceived to meet both the continually evolving aesthetic demand and satisfy a pressing need within the ceramic industry to reduce the environmental impact of glaze application and preparation. It should be born in mind that the grinding and wet application of glazes generates substantial quantities of by-products (i.e. ceramic sludge) which need 171


to be disposed of in compliance with standards. The cost of doing so is high and results are not always satisfactory. Dry glazing, on the other hand, does not involve any grinding (which, as we shall see, is carried out “dry” or wet by the glaze manufacturer). Glazing thus requires no vehicle and does not produce any sludge. Dry glazing products have so far enjoyed extensive success in single firing floor items and have evolved to form the three main categories that simultaneously exist on today’s market. Initially unsatisfactory attempts made in the 70s which consisted of dry-applying standard glazes in powdered form were followed by the first proper products known as frit grains. These products are made via controlled crushing of the frits (described in greater depth further on). The resulting grains are put through particle size selection, the out-of-range part being rejected. Frit grains have been – and still are – used extensively on numerous commercial tile types ranging from the glossy to semi-glossy and even the “technical”. Application begins with an initial wet-applied base glaze followed by drop-application of the dry grains. If the manufacturer has the forethought to apply the grains while the underlying glaze is still damp (i.e. has not yet dried) the grains adhere to the tile surface extremely well. Nevertheless, producers often prefer to apply a glue suspension over the grains to prevent them being detached by in-kiln air flow. An alternative to the glue suspension solution is application of a further layer of glaze: this improves adhesion and can also yield certain aesthetic effects (interpenetration or transparency). Frit grains can also be used on a base that has been coated with a high viscosity glue beforehand using silk-screen printing techniques. Of course, the grains only “stick” where the glue has been applied and even a weak air flow will carry the rest away: a whole range of effects can be obtained this way. One of the main problems with frit grains is that of “unmelted particles”. These are foreign bodies which sometimes exist in miniscule percentages within the frit. The nature of this “contamination” often varies. It may consist of small quartz grains or pieces of refractory material that have detached from the melting kiln lining. The presence of mixed unmelted particles in the grains frequently generates dots on the surface of the finished tile. Last but not least, note that the particle size distribution range of these grains lies between 0.15 and 0.6 mm. A second type of grain – sintered grains – was subsequently launched on the market. These are produced from a ground, dried glaze that is then treated so as to form agglomerates obtained by way of extrusion or pelletization. The agglomerates are then put through a 24-hr heat treatment in discontinuous or fast kilns similar to those used for the tiles themselves. Temperature must reach the glaze softening point so that well-vitrified agglomerates of good consistency form. These agglomerates are then crushed repeatedly to produce the grains. 172


The final stage consists of particle size selection. These grains offer a series of advantages with respect to frit grains. The first is the total absence of infusions in that sintered grains are produced with ground glazes. A second advantage lies in the possibility of using just about any type of composition, an alternative unavailable where frits are used “as they come”. The particle size range for these sintered grains spans from 0.18 to 1.5 mm. While sintered grains cost much more than frit grains their use is justified by the excellent aesthetic effects they produce. The third mainline grain group takes the name of pelletized grains. These are produced from ground, dried glazes that are pelletized and hardened with chemical additives. Once again, proper particle sizing is essential, the typical range spanning from 0.18 to 3 mm. Used on all tile types, these grains provide a range of advantages with respect to their sintered cousins, not least of which is the lower cost. The use of additives acting as binders may cause degassing problems (especially where firing conditions are borderline and products are already loaded with other organic substances that have to be decomposed). Glossy white Used on the market in all three above-described types. The frit grain is usually produced in a less fusible form while sintered and pelletized grains use high-fusibility, generally leaded white glazes opacified with zirconium, tin, zinc. Transparent glaze Here too, the most commonly used products are highly fusible. Suitable for types that require the grains to melt until the product spreads completely. In other cases tiles with only partial grain fusion are produced. White matt This product is difficult to use in pelletized form in that it is harder to eliminate the adhesive degassing problems. As a frit grain there are often persistent infusionrelated problems. Another generalized problem is that it is difficult to obtain flat surfaces without using zinc (albeit in small percentages) as a devitrifying agent. Remember that in these cases zinc often lays the tile surface open to attack by acid. Special grains These include glazes with lustre effects (those obtained from the frit being especially effective), crystallized effects and deep-colour transparent effects (sintered grains are in a league of their own here). Note that it is also possible to mix different grain types (even where they belong to different families) and production techniques, thus multiplying the range of ceramic and chromatic effects. It should, however, be born in mind that particle size differences must be limited otherwise separation and layering will occur during both storage and application. 173


Technical grains Usually of very fine particle size, these are frit or pelletized grains. Applied on specific base glazes in very thin layers. Made from frits belonging to the technical family, they thus feature excellent abrasion resistance and good compactness even inside the vitreous mass. Capable of passing even the ISO 10545/7 Class V test. Fig. 94 provides an overview of grain types, showing how they split into the three main production technologies.

FRIT GRAINS

SPECIAL GRAINS

PELLETIZED GRAINS

TECHNICAL GRAINS

WHITE GLOSSY GLAZE

LUSTRE GLAZE

COLOURED TRANSPARENT FRIT

CRYSTALLIZED GLAZE

Fig. 94. Various types of grain.

174

SINTERED GRAINS

TRANSPARENT FRIT

WHITE MATT GLAZE


Chapter V PHYSICAL AND STRUCTURAL PROPERTIES OF CERAMIC RAW MATERIALS

Many properties of raw materials, whether fired or unfired, correlate with their chemical or mineralogical composition. Yet other parameters, such as size and shape of particles, their arrangement and the way they come into contact with each other the very structure of the ceramic body are equally important. Such factors strongly influence the chemical and physical changes that take place in the presence of the energy input associated with firing. For this reason, most ceramic materials need to be broken down to a size that will optimise subsequent production processes: this goal can be achieved via a series of methods, collectively known as COMMINUTION. A comminution process consists of applying energy to particles until they break or separate to form smaller particles: the reactivity of individual particles, of course, will vary as a function of compression and abrasion resistance, hardness, elasticity and the nature of the particle as dictated by the type of deposit, how it is fractured and the whole geological history behind it (pressure, temperature) etc. Without entering into an in-depth description of the principles behind the grinding process (a subject that will be dealt with in Volume 2), let us examine the main means of comminution. Crushing or compression of particles, between the very hard surfaces in the crushing or milling machine. In theory, the effectiveness of this comminution process varies as a function of the compression resistance of the ceramic material, yet the irregular shape of the particles also considerably influences the forces that shatter them. Crushing generally yields relatively uniform sized particles and little powder. Grinding of particles, in which comminution occurs by rubbing and shattering the particles against each other and use of fixed or mobile grinding media, resulting in gradual reduction of average particle size and simultaneous production of finer powder and a correspondingly wider particle size distribution. Impact fracturing, instead, involves new comminution principles: force is applied on particle edges/corners and thus works efficiently along the fissure lines in the particles themselves. Breakage usually occurs along the natural lines of weakness of the structure (associated with the mineralogical nature of the material) or where different minerals meet. Generally speaking, raw material comminution machines (to be described in greater depth in Volume 2) employ a combination of the above-described principles. 175


The main problem is the sharp drop in grinding efficiency that occurs as the particles get smaller. It is thus common practice to employ return-feed screening systems that filter out already-ground material as soon as it forms. Particle size distribution A raw material made up of particles of different size and shape may be defined as a particulate system: size distribution of the grains in such a system is enormously important, yet very difficult to measure. Were they perfectly spherical, it would be easy to sort them via dimensional or gravimetric selection (assuming constant sphere density): yet with a clayey system the matter is somewhat more complex, making it difficult to define exactly what is meant by the word particle. While it has already been described how a clay is made up of extremely small particles (micelles), it is also a known fact that these micelles, in the ceramic bodies and raw materials themselves, are aggregated and agglomerated, thus making them difficult to isolate in their free form. Particles produced by grinding clays, then, will essentially be made up of many small agglomerates of irregular form and variable surface/peripheral charge, thus influencing behaviour in aqueous suspensions and during firing. The first problem, then, in attempting to measure the dimensions of ceramic particles and their distribution, is to decide whether to concentrate on the dimensions of the individual particles or, vice versa, the agglomerates. Similarly, the choice of pre-analysis dispersion systems will be decisive, as these could break down the agglomerates/particles or even cause further agglomeration: particle size analysis on a material dispersed in water with rheological additives gives results that differ substantially from measurements made on the same powder in its dry state. Further complications are found in the enormous variety of granulometric sizes within a ceramic body, especially where it has been dry-ground: however, even wetground materials contain particles ranging in size from just a few tenths of a µm, to tens of mm, and there is no instrumentation capable of providing adequate accuracy over such a wide range. Finally, the particles of a ground ceramic material, especially where clayey, are far from spherical, often plate-like. These particles are mixed in with other materials that break down to form particles that, while spherical, are of greatly differing density. Nevertheless, use of statistical evaluation and particle shape correction equations, which give spherical-equivalent diameters, makes it possible to effect reliable particle size distribution measurements of solids in a suspension (usually in an aqueous solution) so as to evaluate the behaviour of the combinations during grinding, forming, drying and firing.

176

Physical and structural properties of ceramic raw materials

The main dimensional classification methods are listed in the following table, which orders the methods according to minimum and maximum detectable particle size. Method

Detection interval (µm)

Sieving Micro-sieving/filtration

> 50 0.2 - 50

Optical microscope SEM - Electronic microscope AFM - Atomic force microscope

25 - 2500 0.5 - 1000 0.001 - 5

Sedimentation Elutriation Centrifugation

1 - 50 2 - 50 0.05 - 5

X-ray dispersion Laser diffraction

0.05 - 100 0.05 - 200

Gas permeability

0.1 - 300

Powder classification is always expressed as per national standards on the basis of equivalent sieve classification, according to ISO (international), BS (British), UNI (Italian), DIN (German) or NF (French) values. However, the MESH ASTM classification (USA) is used extensively all over the world and the equivalent sizes are given on the following page:

177


Mesh A.S.T.M. 5 6 7 8 10 12 14 16 18 20 25 30 35 40 45 50 60 80 100 120 170 200 230 270 325 400

Micrometer (µm) 4000 3350 2800 2360 2000 1700 1400 1180 1000 850 710 600 500 425 355 300 250 180 150 125 90 75 63 53 45 38

Mesh / cm2 3 4 6 9 11 14 19 26 35 50 75 100 140 200 270 380 590 1100 1600 2400 4700 6200 9500 12350 18200 26000

The above table highlights the limitations of sieving methods, their use being confined to rather approximate discrimination: on the other hand, hydrodynamic methods, long used for finer resolutions, are complex and lengthy (especially so for very fine particles). The best compromise is to use short wavelength incident radiation (X or laser) interaction instruments through the application of modifications to Stokes law. This is expressed by the following equation: V=

2 (ρ - ρ ) g r 2

1 2 ____________________

9η

where V, the speed of a falling particle in cm/sec, is obtained from the values: η viscosity of the suspension medium [water = 0.013 (10 °C), 0.010 (20 °C), 0.008 (30 °C) poises]. ρ1 density of the particle ρ2 density of the suspension medium g gravitational acceleration (981 cm/sec) r radius of the particle, assumed to be spherical (cm). 178


At constant temperature, where all constant parameters are known and the sedimentation speed of a set of particles is measured, it will be possible to calculate the average dimensions of the particles themselves. By applying Stokes equation values to a particle with a density of approximately 2.5 g/cm3 and a diameter of 1 mm, it can be deduced that it will drop through water at 20 °C, in the absence of any interaction with the water itself, at just under 1.2 cm/hour. The appropriate practical modifications to the equation, as a function of a real particle, reduces this speed to about 0.3 cm/hr, thus giving a clear idea of just how long hydrometric particle size analysis can take where a sedimentation of at least 20 cm (and thus a wait of some 64 hours) is required. Whichever system is used, particle size distribution can be represented by two graphs (see Fig. 95): an actual % greater or less than a particular size or a cumulative % against a logarithmic scaled size on the x axis. Industrial production involves a series of particle size distribution tests designed to ensure the desired outcome: it is, in fact, necessary to: a) check the dimensions of the raw materials destined for grinding so as to optimise control of the process itself. b) remove coarse (sieve-retained) materials resistant to the grinding process, eliminating them or recycling them back into the grinding process.

Fig. 95. Example of differential and cumulative particle size distribution curve.

179


c) run a series of checks, with appropriate sieving, so as to ensure smooth operation of the spray drier pumps, the spray drier, the presses, and the glaze/engobe/silk-screen application stations along the production line. All this is essentially based on a no larger than principle through use of appropriate sieves. At this point it is important to highlight, with particular reference to grinding (and homogenisation), the influence of variations in particle size both throughout the process and on final product characteristics and classification. The understanding and control of particle size is essential to ensure optimum body texture and homogeneity and the required porosity, specific weight, modulus of rupture. The texture of the material is strongly influenced by the type and degree of grinding. This is because grinding alters the size and shape of the individual grains and causes variations in the way in which they associate, thus influencing particle packability and final density. Various studies on maximum attainable packing density have been carried out with regard to both spherical particle systems with no particle deformation and for isomorph sphere systems: these studies have yielded an ideal distribution that gives rise to two possible solutions: open packing with spheres arranged by cubic symmetry, where empty gaps account for 48% of the entire occupied volume and closed tetrahedral-symmetry packing with gaps accounting for just 26% of total volume. Of course, systems featuring mixtures of different, always-spherical grains will give rise to ever-more complicated theoretical models which optimise use of space enormously: for example, a tri-modal system based on appropriately calibrated spheres (having an assumed diameter ratio of 50:8:1) of large (62% by volume), medium (24%) and small (9%) size will, in theory, fill the space very efficiently, leaving an air-gap residue of just 5% (62 + 24 + 9 + 5 = 100%). While the individual particles in a ceramic body are generally angular rather than spherical the above-explained general principle may be taken except where particles are particularly elongated in one direction as being applicable, albeit inexactly as there are many other inconveniences linked to the fact that the irregular shape of the particles stops them sliding against each other. On the other hand, that irregularity will, statistically, give rise to the real possibility of forming more extensive areas of tighter agglomeration and, all in all, lowering average porosity. In general, grinding processes that involve initial impact crushing followed by abrasive comminution (as in an Alsing-type ball mill), give optimum yields in terms of attainable density, especially where the treated materials are of different hardness and density: this thus permits the co-presence of relatively coarse particles (feldspars, quartz etc.) and others of smaller and smaller particle size distribution (e.g. clays). Clearly, a particle size control system consisting of just one simple sieving (so180


called sieve residue) serves only to establish whether production activities completed thus far have remained within operational parameters. Fig. 96 shows two bodies (the particle size distribution curves of which are illustrated) with an identical residue of about 2% at 63 µm: yet they clearly differ substantially in terms of particle size composition, with body A having a Gaussian distribution around 15 µm, and body B having bimodal distribution centred around 0.6 µm, with an average dimension of about 9 µm. Slip sample A (AVG. DIAMETER = 15.3 MICRONS)

Particle Size (Microns)

Slip sample B (AVG. DIAMETER = 9.1 MICRONS)

Particle Size (Microns)

Fig. 96. Particle size distribution analysis of two bodies having same residue but different particle size distribution.

181


The technological behaviour of the two bodies, which seem identical in the light of residue-only analysis will actually be radically different, as B is much more compactable than A. The unpackability of ground raw materials and the forces applied during the forming process (pressing, extrusion) make the production of a ceramic body without any empty spaces at all virtually impossible, no matter how accurate selection and particle size distribution may be. All ceramic materials, except for some glasses, have pores or empty spaces: the porosity of a material is, then, defined by the quantity of air it contains. In scientific literature there are six different kinds of porosity (see Fig. 97):

Fig. 97. Different types of porosity. From: Grimshaw, Chemistry and Physics of Clays-Benn.

a) b) c) d) e) f)

closed pores capillary canals connecting closed pores blind pores interconnected pores open or ink bottle pores micropores (so small they prevent entry of water or any other liquids).

There are two types of closed pore: those formed by pressing the semi-finished item and those originating from open pores that are sealed by material that melts or forms during firing. In both cases, but mostly the latter, closed pores are full of air, water vapour, SO2, CO2 or CO, or even O2. When a ceramic body is heated, the pressure inside the pore increases and exerts considerable force on the surrounding solid material: this can cause fissuring, lamination and even explosions as long as the material remains rigid (low temperature), but when the latter starts to soften at temperatures close to and 182


beyond the vitrification point, the pressure exerted by the enclosed gas may cause expansion of the material itself and create bubbles which, in turn, form new macroscopic pores. Dimples or pin-hole defects seen on fired glazes often stem from closed pores too: these are formed by carbonates, sulphates and other pyrolysis-affected compounds that emit gas at high temperatures: if this occurs when the glasses in the glaze are about to become plastic the glaze will block the escape and elimination of such gases. Open pores (i.e. those connected to the exterior by capillary channels of varying length and width) may be formed by the original particle packing configuration, elimination of water vapour during drying or the initial stages of firing, elimination of gas during firing, micro-structural alterations in the pieces during drying etc. The presence of such pores is closely tied to the size of the body particles and how they are packed. In general, since fluids (rainwater with regard to freezing resistance and inks, oils etc. for staining and cleaning parameters) can enter and exit open pores, it is preferable for a fired product to have open pores of either a very large average diameter (>300 µm, to allow easy evacuation of water or introduction of detergent solutions) or a very small one (<40 µm, the smaller the better) so as to make fluid entry difficult and prevent any water contained therein from freezing as a result of extremely strong capillary forces: it thus follows that materials with a substantial amount of medium-size open pores (from 40 to 300 µm) remain at risk. Such a wide range of scenarios makes it possible to measure two different body parameters: total porosity and apparent porosity. The true (or real) porosity of a body is the ratio of all (open and closed) empty space volumes to total volume, while the much more frequently-used apparent porosity parameter is defined as the ratio between the volume of water the body can absorb under certain conditions (during prolonged boiling or vacuum treatment) and the total volume of the piece. Such porosities will clearly be influenced by the dimensions, shape and packing of the particles, the chemical nature of the materials making up the body, and the technological treatment applied to materials during production: an often-underestimated parameter is the intrinsic porosity of the individual particles. As stated in the introduction, depending on their mineralogical formation, clays possess widely differing structures with interlayer spaces ranging from 5 to 20 or more Å; in this respect they differ greatly, of course, from quartz or other compact material particles. For this reason a clay-based body containing large quantities of sands and feldspars, especially where the latter are of suitable particle size, will be less porous than that of a predominantly clayey body. In addition to these morphological features, body texture and particle size distribution will play an important role in physical and chemical reactivity during firing, as will be illustrated further on. 183


Any particle size variation, in fact, will have an influence on the specific surface area of the powders and an immediate influence on the inter-particle surfaces, sharply increasing that reactivity surface area, which, via solid-solid or solid-viscous liquid reactions (high-temperature ceramic bodies), is of primary importance for proper sintering. The example in Fig. 98 shows a ceramic body matrix made up of relatively coarse feldspar (F) and quartz (Q) particles contained in a brick-like framework of clayey particles:

Fig. 98. Progressive fusion-reaction hypothesis.

The sintering process begins at the particle contact points, where the fluxing action of the feldspars first acts (the black dots in the far left drawing). Following initial reactivity, which tends to involve the smaller particles in closer contact with the fluxing agents, a liquid phase forms: this goes on to involve, partially, the surrounding clayey structure. During the process, and especially during cooling, crystalline neo-phases form and reinforce the sintering phenomena in the mass. The importance of particle size distribution in ceramic manufacturing is not just limited to the need for proper grinding: there must also be proper control of the powder agglomerates sent on to the pressing department. As seen in the introduction, ceramic technology allows us to form ceramic tiles by pressing powders of low moisture content (3-6%). Such powders may be produced by dry grinding (i.e. with dense, irregularlyshaped and often sharp-edged solid particles), followed by appropriate addition of water, or wet processing followed generally by spray drying that produces (sometimes hollow) smooth spherical particles which can agglomerate. In both cases it is essential that powders flow smoothly from the silo to the press feeder without segregation and fill the press cavity as evenly as possible: they must therefore be properly granulated. Particle size distribution of granules is extremely important in attaining a postpressing green tile that is compacted evenly all over and free from defects. The differences between dry and wet ground powders will be dealt with in depth in Volume II of this work. 184


In both the former case (particle size distribution either natural or induced by raw material grinding) and the latter (particle size distribution of semi-finished items), it is essential to evaluate the relative amount of the different granule classes so as to obtain the best possible mix and optimum compaction.

185


HOW WATER INFLUENCES CERAMIC SYSTEMS Water plays a vital role in ceramics. Most ceramic raw materials are sedimentary in nature and thus formed in an aqueous environment. Many key minerals are of secondary origin, created by water-induced weathering of primary rock. Furthermore, water is added at most stages of the ceramic production process to aid completion of shaped semi-finished items. Although water has a significant effect on any type of raw material, it plays a particularly important role in clays owing to the already-described structural characteristics: these include inter-particle spaces and electrostatic charges that allow intense interaction with the water molecules, which, because oxygen and hydrogen have different electronic affinities, may be seen as small dipoles with a partial negative charge on the Hδ+ O and a partial positive charge on the H δ-O < δ+. H When water is mixed in with the powders in quantities equal to approximately one fifth the weight of the powder itself, plastic bodies are obtained. Where that proportion rises to between a third and a half, suspensions of extremely differentiated rheological characteristics result: given the dimensions of the clayey particles, these may be considered and treated as colloidal suspensions. Given the complex nature of water these basic interlayer electrostatic interactions are accompanied by others dependent on structural or chemical irregularities, generally influencing the surface of the particles: a) distorted ionic groupings b) broken surface bonds, unsaturated bonds, sliding or crumbling planes c) cations of unsaturated charge, or cations which create charge defects in the crystal lattice etc. If a basal plane of the crystal lattice is theoretically in electrical equilibrium and the surface, vice versa, has a excess/insufficient charge (generally caused by the high number of oxygen2- atoms in the ceramic material), a fracture will greatly disturb that equilibrium, causing different attraction/repulsion scenarios involving the water molecules (Fig. 99). In general, though, the particle surfaces are deemed to be sufficiently charged to attract at least one layer of water (see Fig. 100), which remains attached by a strong electrostatic bond and differs from free water in that it is denser (like ice): this first layer of molecules, oriented according to the attraction of the particle on the dipole, is known as a double electric layer and has been studied and characterised extensively. It varies in thickness from approximately 7 to 60 Å. In-depth analysis of double layer interactions is outside the scope of this publication. Many experts interpret it as a sort of molecular condenser, the (zeta) potential of which can be calculated by studying variations in the potential energy of the system (see Fig. 101). Here we shall simply examine the macroscopic 186


Fig. 99. Kaolinite crystals with positive and negative edge charges (A and B respectively), deflocculation (C), edge-to-face flocculation (D), face-to-face flocculation (E), flocculated Ca clay (F), deflocculated clay -Na (G).

free water free water

bound water

Fig. 100. Distribution of charges and water, free and bound, around clay particles in suspension. The cross-hatched area in the middle is the clayey core.

187


variations caused by disturbance of these electrostatic interactions at molecular level. Before examining these variations in more detail it is worthwhile highlighting a parameter often neglected in academic works but of great importance in industrial practice: time. That is, the contact time between water and the ceramic material. It is a factor on which the completion of many reactions depends: such reactions can have an enormous influence on the final characteristics of the suspension to be spray-dried or cast, the body to be extruded or the semi-dry powder to be pressed. The time factor involves both physical and chemical changes. These range from the degree of hydration significantly influencing the plastic properties to the solution of substances present. The latter can result in the freeing of ions, alteration of pH or aiding the oxidation of cations in iron and sulphides. Another important yet often neglected alteration of significant macroscopic implications is the emission of hydration heat, which, in clayey raw materials may vary from 0.1 to more than 600 calories per gram yet is generally negligible in other ceramic raw materials (0.0015 c/g for quartz, 0.04 for Feldspars). Zeta potential:

e = charge D = dielectric constant d = distance between rigid and diffuse layer rigid layer

diffuse layer

potential resultant R = repulsion

distance A = attraction

Weak double layer: R weak à attraction prevails à flocculation Strong double layer: R intense à repulsion prevails à deflocculation

Fig. 101. Attraction and repulsion potentials in a radial direction starting from the surface of a grain dispersed in a solution.

188


As regards aqueous suspensions of ceramic bodies, the main events which need to be controlled and guided are flocculation and its opposite, deflocculation. In the former, rapid sedimentation of flaky particle aggregates of greater density easily separated from the system are created from within a homogeneous in-water particle suspension. In the latter attempts are made, vice versa, to aid dispersion of aggregates using appropriate additives, thus obtaining particles that remain in suspension for long periods even in the absence of agitation. This last characteristic is, of course, particularly sought after in the wet grinding of clayey bodies. Both effects are the result of disturbing the solid-water electrical particle layers: flocculation is usually caused by the introduction of small and therefore high charge-density bi or tri-valent cations; these can seriously alter the electrostatic equilibrium of the suspension. Deflocculation, instead, is generally achieved by adding organic or inorganic polymeric molecules that isolate the particles from each other by separating them on account of their structural characteristics. Here too, the presence of appropriate cations (Na+, NH+) that alter the equilibrium of the double layer aids the effectiveness of the additive. Because ceramic particles are negatively charged, they easily attract positively charged cations. The tendency of cations to be adsorbed has been shown, for the most common, to be: H+ > Al3+ > Ba2+ > Sr2+ > Ca2+ > Mg2+ > NH4+ > K+ >Na+ > Li+, so the useful cations, such as sodium and potassium, have little chance to act. The effect (which will depend on the ratio of their charge and volume) may be used to vary the electrical characteristics of a suspension or by adding excess to act on the exchange equilibrium (CLAYn-. Xm+) + Ym+ ⇔ (CLAYn-. Ym+) + Xm+ where X and Y are the different cations.

189


190


Chapter VI RHEOLOGY: BASIC CONCEPTS

The word RHEOLOGY is a relatively recent addition to the vocabulary of ceramics: yet it concerns problems that the ceramist has to deal with all too frequently. Rheology from the Greek words reo meaning flow, and logo meaning study is, then, the study of flow. This includes flow in both solids (e.g. the rheology of powders) and fluids (e.g. the rheology of slips or even gases). Such matters are of enormous interest to the ceramist, as they are of relevance at nearly every stage of the production cycle, especially during the wet grinding of bodies and glazes and during on-tile application of the glazes. One of the biggest challenges facing the ceramic technician is translating that colourful jargon surrounding the practical aspects of rheological phenomena into more rigorous terminology. Remarks such as the glaze flows like oil, its bubbly, its settled and hardened in the tank, its thinned out too much, the mills jammed and the slip has frozen are all common factory floor comments. These empirical expressions actually reflect considerable knowledge of how rheological factors affect the production process. Yet it is a knowledge that has been built up on a trial and error basis and which, to a large extent, has not been rationalised in a way that allows it to be transmitted. Nevertheless, the ceramic industry continued to rely on this rather uncertain wealth of knowledge for years, and it has to be said that production processes were, by and large, kept under control; when production cycles were slower and product change-overs less frequent, this level of understanding was more than good enough. Now, though, with the advent of new technologies and the fast-changing flurry of products and glazes there is no longer time to build up an empirical understanding of events. Instead, the glaze or body slip must comply with precise production standards, defined according to standardised units of measure. Let us begin by shedding some light on the concept of viscosity. This term has become a part of everyday vocabulary and is now mentally associated with a greater or lesser ease of flow. To arrive at an exact definition it is first necessary to define the exact physical measurement one is referring to. A rather crude definition of viscosity might be the effort needed to make a fluid attain a certain speed. 191


The first thing is to clarify that one cannot refer to the speed of a moving fluid as if it were a racing car. If that were the case then all the fluid, like all the components of the car, would be moving in exactly the same way. For example, we can talk about the speed of a drop of water as it falls into an empty glass: the entire drop moves with respect to the sides of the glass and thus its speed is defined by the ratio between distance covered and time taken as with the racing car. Yet a fluid flowing in a tube is subject to quite a different set of conditions: the part of the fluid in contact with the wall of the tube moves much more slowly than the fluid in its middle. The absolute speed of a fluid molecule thus differs according to its position inside the tube, reaching a maximum in the centre and a minimum up against the walls. In shifting from the wall to the centre, then, speed increases constantly; this is because none of the molecules are free to move in the same way as a car on the race track or the drop falling into the glass. They are, instead, obstructed by the other molecules in the fluid as a result of internal friction. The wider the tube the easier it is for the fluid to flow because the particles in the central zone are less affected by friction than those up against the walls. Or, in tubes of equal diameter, the lower the friction between the molecules of the fluid the higher the speed of the fluid in the central position. Velocity gradient To define viscosity we need to know all the parameters affecting it. One of these is the velocity gradient. If, in a certain zone of a moving fluid mass we measure maximum speed (vmax), and in another the minimum (Vmin) and then the distance separating the two points at which the measurements were taken (l), we obtain a gradient value valid for the moment of measurement. The difference between the two speeds divided by the distance is defined as the velocity gradient and is indicated by D: D = Velocity gradient = (Vmax - Vmin)/ l As we have seen, this value is directly affected by the degree of friction within the fluid. If, all conditions being equal, a fluid has a high velocity gradient this indicates that its molecules generate low-level internal friction. The physical dimensions of D are: v [LT -1] ___ _______ D= = = [T 1] = reciprocal seconds L [L] This is the inverse of a time. Reference is always made to the velocity gradient expressed in reciprocal seconds. 192

Rheology: basic concepts

Shear stress The second element affecting viscosity is the stress the fluid is subject to. Let us go back to our example of the tube for a moment: tube diameter and influid friction remaining equal, it can intuitively be seen that increasing pressure increases the velocity gradient (i.e. the fluid will flow faster). However, the concept of pressure, as a description of the stress to which the fluid is subject, is inadequate in that it is not necessarily connected with flow. In a closed system there is no flow at all, whatever the applied pressure. This is because pressure is the ratio between a force and a surface perpendicular to the force itself and does not necessarily generate movement. Example: a concrete block resting on a floor generates a pressure defined as force per unit of surface area. Getting the block to move requires considerable force, but such force must be applied parallel to the floor, not perpendicular to it. At a certain point, once the applied force reaches a certain threshold, the block will start moving along the surface of the floor. The force needed to start the block moving will be considerable as cohesive forces that bind two objects in contact when at standstill need to be overcome: such motion-opposing forces, which tend to stick two immobile objects together, are defined as static friction. In mechanics the ratio between the orthogonal forces acting on the two surfaces in contact, when movement begins, is defined as the coefficient of static friction, µ (mu). Note that, since the coefficient of friction is an intrinsic characteristic of the two surfaces in contact, surface area is non-influential: both forces act on the same surface even though they are orthogonal. When the body, still subject to the pressure of its own weight, finally starts moving, a traction force needs to be applied: this will be proportional to speed and surface roughness. This applied stress, like the above-mentioned static friction, is needed to overcome the forces generated by dynamic friction (generally lower than static friction). Of course, one cannot speak of friction between two surfaces subject to orthogonal forces where fluids in motion are concerned: the relevant concept in this case is internal friction, and thus the only significant value when studying the flow of a fluid is the force applied tangentially to the unit of surface. This is defined as shear stress and is indicated by the Greek letter τ (tau) and is measured as a pressure: Ft [LMT-2] ___ ________ τ= = = [L-1 MT-2] = N / m2 = Pascal 2 A [L ]

193


Viscosity Referring to the above measurements it can be said that applying a force to a certain area of a fluid mass to keep it moving causes the fluid to respond by generating shear stresses that result in different zone-to-zone speeds. Within the mass, then, a velocity gradient is created. Its value depends on the internal friction of the fluid, a factor closely related to its viscosity. Viscosity is defined as the ratio between the shear stress and the velocity gradient. It is indicated by the Greek letter η (eta) and has the dimensions of a pressure over time: τ [L-1MT-2] ____ _________ η= = = [L-1 MT-1] = Ns/m2 = 10 Poises -1 -1 [T ] [T ] This essential formula tells us that viscosity is directly proportional to the shear stress needed to obtain a given velocity gradient in the fluid. On the basis of the units of measure adopted for D and τ, viscosity is usually defined as: η = pascals per second This however, is too large for many practical applications, so a sub-multiple is preferred: millipascals per second = mPa . s = centipoises = cP Other methods of defining viscosity are to be found in more specialist literature. In the context of ceramics, then, viscosity varies enormously, spanning from just a few centipoises in the case of disc or spray gun-applied glazes to hundreds of centipoises for bell-applied ones or slips prior to spray drying. The range of velocity gradients to which ceramic suspensions are subject is even wider, extending from tens of reciprocal seconds in slow-stir tanks to hundreds of reciprocal seconds in pumps and several thousand inside nozzles. Glazes may be subject to large and sudden variations in the velocity gradient: final on-tile flows, in fact, starting from a situation that varies widely depending on the type of applicator, always end with a zero velocity gradient. At first glance such a wide velocity gradient interval is no cause for concern in that once the viscosity η of our glaze suspension is defined it should be possible to forecast its behaviour whatever the flow situation. Unfortunately, this is true only in a minority of cases of ideal or Newtonian viscosity where the ratio between shear stress and velocity gradient remains constant for any velocity gradient value. In these few fortunate cases (e.g. Vaseline oil), the behaviour of the fluid can be 194


fully described by the value η as it remains constant whatever the flow speed and depends on temperature only. A graph showing τ (shear stress) values against variations in D (velocity gradient) gives us the line illustrated in Fig. 102. This way of illustrating rheological behaviour, generally defined as a rheogram, is extremely useful. Viscosity is represented by the ratio between the two quantities, that is, the angle of the line (i.e. the tangent of the angle α). The greater the angle the higher the viscosity. Predicting behaviour in fluids like this is easy: it will continue to flow at decreasing speeds as shear stress decreases and will stop flowing altogether when shear stress is completely eliminated. Example: oil spilt on a floor seems to spread forever. While flow speed depends on the viscosity of the oil, flow will continue until the shear stress produced by the hydrostatic pressure (determined by the level of the liquid) is counterbalanced by surface tension. Since oil has very low surface tension the slick will spread a long way. Yield point Experience teaches us that ceramic suspensions are decidedly non-Newtonian. In attempting to define those divergences from ideal behaviour that significantly affect the technological cycle, it should first be pointed out that one of the most important aspects of such a suspension is that it will often start flowing only when applied shear stress exceeds a certain threshold. Similarly, if it is already moving, it will only stop doing so once shear stress drops back below that threshold. Returning momentarily to the example of the concrete block resting on a floor, there is a clear analogy with the tangential force needed to shift the block. The level of such force depends on friction. Getting back to the flow of fluids (i.e. rheology), this threshold value is defined as the yield-point. This is an initial shear stress indicated by τ0 (tau zero) and is measured in pascals. Where a fluid has a yield point and flow only begins at a threshold level of shear stress, the velocity gradient remains zero until the τ0 level is reached as in the graph in Fig. 103. This important characteristic is common in ceramic fluids where flow stops though still subject to shear stress. It is this parameter that determines the spread of disc-applied glazes, the form of the droplets in the droplet method, the thickness of the dipped glaze or the thickness of glaze applied to a vertical surface before it starts to run.

195


Fig. 102. Standard rheogram for Newtonian-type fluid where the following relationship applies:

τ D

η = _____ = tan α.

Fig. 103. Rheogram for a fluid with an initial yield point τ0 (Bingham plastic). 196


The importance of time Certain types of non-ideal rheological behaviour are important in ceramics because of both their practical importance in industrial processes and the fact that they sometimes give rise to quite striking phenomena. For instance, it is known that ceramic suspensions, especially those with considerable percentages of clays, can become rigid if left undisturbed for a time and then turn fluid when agitated. This characteristic is particularly evident in the bentonites. The latter, in fact, are used in the oil industry to seal drilling wells: the well is filled with an easily pumpable slip that, when left undisturbed, hardens to form an efficient plug. The advantage of this system is that the well can be reactivated simply by directing a water jet into it, thus agitating the bentonite and returning it to its fluid state. This property stems from the fact that, at standstill, the structure of the fluid changes as a function of time: three-dimensional bonds form between the particles, making the structure rigid. These bonds are destroyed by the application of shear stress and need time to reform once the fluid is again motionless. Hence viscosity changes as a function of time. While most body or glaze slips behave this way, time-dependent viscosity is not generally a problem as it occurs at very low flow rates and ceramic suspensions are usually stirred constantly at much higher velocity gradients. Note that the key difference regarding the yield point is that its effects are seen as soon as the fluid drops below a given shear stress value; whilst this is independent of time the viscosity of such suspensions will still be time dependent. The difference is that the yield point remains a characteristic of the fluid even when it is agitated, while, if the fluidity is also time-dependent, the τ0 value can increase drastically if the fluid remains at standstill. This poses a problem as to how the yield point is measured. Since ceramic suspensions are also time-dependent, it is necessary before constructing the rheogram to subject the suspension to the highest possible velocity gradient for a period long enough to destroy all the structures that may have formed; then the velocity gradient is brought to the minimum D and τ is measured to produce the τ rheogram as D varies. As might be expected, where time-dependence is absent the D/τ rheogram will, as D values are increased and then decreased, turn back along the same path to give a single curve valid for both the outward and return scenarios. Where, instead, behaviour is time-dependent the rheogram obtained by decreasing D is different from the increasing one. The only way of measuring time-dependence in itself is to take viscosity measurements as a function of time while keeping the D value of the sample constant, thus constructing a rheogram.

197


In constantly-stirred suspensions this measurement is of little interest: however, if the extent of the time-dependence is abnormal it can cause difficulties in the production process. For example, the suspension may thicken so quickly as to make mill emptying slow or impossible, or the viscosity of the suspension may increase steadily even as it is slowly stirred. In this last case the slip can be unloaded, pumped and sieved easily enough but as soon as it is introduced into the tank it tends to freeze because the agitator turns too slowly to prevent bonds forming in the fluid. Without the addition of fluidizers the situation can soon become critical and the contents of the tank thicken so much that only the area actually swept by the agitator paddles remains fluid. Unfortunately, at this point the addition of fluidizers becomes awkward as they can no longer be distributed efficiently. The best solution is to dilute the fluidizer in water and add it to the slip slowly, employing a pump to create a top-to-bottom return flow within the tank contents. It may take a while to completely homogenise the fluid and more solid parts of the suspension. Thixotropy A third type of rheological behaviour frequently encountered in ceramics is thixotropy which, if excessive, can cause difficulties. This is a tendency that leads some suspensions to change their rheological behaviour as a function of the yield stress they have previously been subject to. The most common scenario is that in which viscosity is lower where the fluid comes from a situation of higher stress, and is often when a yield point exists. In practice a graph illustrating τ and D for initially higher and then lower velocity gradients (D) yields a rheogram with hysteresis (Fig. 104). This hysteresis effect is defined as thixotropy; however, a unit of measure that allows for absolute evaluation does not exist. The only way of estimating the phenomenon is to take a series of measurements at higher and higher and then lower and lower velocity gradients and then trace a graph of the results. Thixotropy is represented by the area enclosed by the curve. It is evident that the obtained values will depend on measuring procedures: the use of standardised procedures is thus essential if one is to obtain results that can then be compared. Since, in ceramic suspensions, this effect is always linked to the presence of a yield point and a certain time-dependence, it is highly advisable to use the same procedures at all times. This allows that type of thixotropy which is most damaging (i.e. that which remains after the suspension has been stressed at maximum velocity) to be highlighted.

198


(a) thixotropic behaviour (e.g. kaolin suspensions in H2O)

(b) antithixotropic behaviour (e.g. magnesium hydroxide in H2O)

(c) rheopexic behaviour (e.g. vanadium oxide in H2O)

(d) antirheopexic behaviour

(e) thixo-rheopexic behaviour

(f) antirheopexic-antithixotropic behaviour

(g) thixo-antithixotropic behaviour

(h) antirheo-rheopexic behaviour

Fig. 104. Rheograms showing hysteresis and corresponding definitions of the behaviour that produces them.

199


From an application point of view, interest generally focuses on evaluating the behaviour of fluids subject to high stress levels and which progressively slow to a halt (e.g. a droplet launched from a disc). The most relevant tract of the rheogram (Fig. 104) is thus the one showing the decreasing velocity gradient, which, with glazes, is often perfectly linear. The rheological behaviour of the suspension can thus efficiently be described in terms of plastic viscosity and yield point. If thixotropy is excessive, application problems can arise, especially where low viscosity and a certain yield point are required. Thixotropy hinders the control of viscosity because it varies depending on how the suspension is agitated. This particular rheological effect is generally caused by in-suspension particles of a shape that leads them to align according to direction of flow (e.g. like platelets, sticks or filaments). As the velocity gradient steadily increases the initially disordered particles gradually align. When the velocity gradient is subsequently lowered, the alignment remains and the fluid flows more easily. Thixotropy is always undesirable, yet can largely be ignored. However, certain circumstances will require efforts to minimise it. As explained, thixotropic behaviour is linked to the geometry of the particles in the suspension and it is clayey minerals, in fact, that have the dimensional characteristics most likely to trigger it. If a glaze is excessively thixotropic it is good practice to modify clayey component ratios, trying out different types of clay if need be. Should the thixotropic behaviour of a body be excessive it is, instead, necessary to modify the deflocculant, perhaps using components with a synergetic effect.

200


Rheology of clays The rheological behaviour of clay slips generally characterises them as nonNewtonian systems. More specifically, they can be classified as plastic and pseudoplastic fluids of the thixotropic type. Note that the same clay may behave in either a plastic or pseudoplastic manner depending on the percentage of solid in the slip, as will be illustrated further on. Depending on the industrial application in question, such effects may be positive or negative. In ceramics the presence of thixotropy and yield points linked to plastic behaviour are undesirable. However, it should be pointed out that, given the technological cycle of clay slip production, effects are limited except in sieving and where the slip is left in the storage tank for a lengthy period. From the information provided in preceding chapters it should be fairly clear which chemical-physical parameters control rheology of clays. In particular, it can be said that all those factors directly influencing the zeta potential especially cationic exchange, contact surface area between liquid and solid (which depends on particle size distribution of the clays) and the in-clay presence of soluble salts, especially sulphates and calcium salts are of primary importance. The effect of the above factors on the rheology of the clays can be controlled in a number of ways: the simplest is to vary the percentage of clay in the slip. Alternatively, the quantity and type of deflocculant can be adjusted or the grinding process can be modified. While the role played by grinding in determining the rheological qualities of a slip is one of the lesser known aspects in ceramics, it is probably one of the most important factors in the rheology of systems. Grinding (especially its duration) directly influences system particle size distribution and slip temperature. In practice we could determine the rheological characteristics of a slip by adjusting grinding, bearing in mind that the longer the mix is ground the greater the solubility of the salts, the greater the contact area between water and solids and, consequently, the higher the level of cationic exchange. This normally leads to decreased system fluidity and an increase of apparent viscosity, accentuation of thixotropy and the appearance or increase of a yield point. These effects are enhanced even further as temperature rises when grinding times are increased. Temperature increase plays two roles: the first directly influences the rheology of the system in a generally negative manner, in that temperatures of > 60 °C are reached (as with porcelain tiles), while the second aids the kinetics of salt dissolution and cationic exchange, leading to the above effects. The percentage of clay in a slip causes rapid and often obvious changes in the rheology of the slip itself: to this end note that all clayey systems feature some degree of thixotropy, seen more as the percentage of solid increases.

201


A water-clay system can also be controlled via the use of deflocculants: the same ceramic material deflocculated with different percentages of sodium tripolyphosphate (STPP) will show significant system rheology differences, with the appearance of a yield point and accentuation of hysteresis. Less favourable rheological conditions reign with higher tripolyphosphate percentages, indicative of an over-deflocculation that causes incorrect system fluidization. Analogous situations arise when the type of deflocculant is changed. Given our present level of understanding it can be said that the phosphates (especially tripolyphosphates) tend to give good results right across the known range of clays by providing sufficient control of system thixotropy. Silicates yield good results with clays that are lean or of high kaolin content, but control of thixotropy is more difficult than with phosphates. Organic products are generally good dispersants and stabilizers but are less efficient in controlling apparent viscosity. As regards the ratio between deflocculants, the yield point and plastic viscosity in water-clay systems, little information is available. This is particularly so in industrial-type systems, owing to the difficulty of discriminating between the various factors that influence the two parameters above. Mineralogy and rheology of clays The most common clayey minerals in clays used by the ceramic industry are kaolin, illite, the smectites (montmorillonite, nontronite) and chlorites. The rheological characteristics of a clay thus depend on the type and quantity of clayey minerals it contains, though they cannot neatly be summed as a function of their in-clay percentages. This is because of the already-cited presence of accessory minerals, which can be important in defining the rheological characteristics of a clay. There follows a brief description of the rheological characteristics of the abovecited minerals: however these observations refer to the minerals in their pure state, while for clays used in industry other parameters influence. Kaolin: of a mineralogical structure that precludes the presence of interlayer water. No cation exchange capacity. There are usually no soluble salts as kaolins are normally very pure. These factors result in non-plastic rheological behaviour. Consequently no yield points are observed except where rheological conditions are particularly unfavourable (e.g. high concentrations, unsuitable deflocculants, extensive grinding-though among the various clayey minerals, kaolin is the one least affected by grinding).

202


The pH of kaolins normally falls between 5 and 6, allowing for optimum deflocculation with sodium silicates, sodium carbonate or, under very favourable conditions, sodium hydrate: products that shift system pH from 7 to 9. Of coarse, phosphates and organic salts also have an optimum effect on the water-kaolin system, while the use of acid pH products is to be excluded. Whilst the above observations apply to pure kaolins, they nevertheless remain largely valid for both white and red kaolinitic clays as long as contamination of the system by soluble salts or clayey minerals of a different nature is limited. Illites: as illites of a purity comparable to that of kaolins do not exist, one can only reason in terms of clayey systems containing varying percentages of illites. This fact, taking into account that rheological characteristics cannot neatly relate to the percentage of minerals present and the difficulties in determining the percentages of such minerals with accuracy, shows how difficult it is to interpret the rheological characteristics of an illite clay. Let us evaluate, for example, an illite-kaolinitic clay in which other minerals and impurities are assumed to be absent. In this case, compared to a kaolin, all other conditions being equal, there is an increase in the plastic behaviour of the system, with more accentuated thixotropy and greater apparent viscosity. On the other hand, comparing an illite-montmorillonite to a montmorillonite clay, the apparent viscosity, yield point and thixotropy of the system will be lower, but compared to an illite-kaolinite clay they will on average be higher. All in all it can be said that illite is a clayey mineral having rheological characteristics less desirable than those of kaolin as there is the appearance of a yield point and thixotropy. Certain grinding conditions emphasise these characteristics as illite is more sensitive to grinding than kaolinite. The most effective deflocculants are tripolyphosphate and sodium metasilicate, while products that act on pH only, such as carbonate and sodium hydrate, are of limited use. This is because the natural pH of illite is less acid than kaolin. The above observations will obviously require modification where minerals of a different nature predominate in the illite clay system. Montmorillonite: from a rheological viewpoint this mineral is without doubt the worst, at least as far as the clay slips used in ceramics are concerned. In other industries, however (e.g. as a sludge in oil drilling), it is actually prized on account of those characteristics. Montmorillonite is characterised by high thixotropy and a particularly marked yield point, features that correspond with plastic behaviour of the slip. Both plastic and apparent viscosity are high even where the percentage of solids is low. A certain differentiation in the behaviour of these clayey minerals is ob-

203


served as a function of the type of cations in the layers. There are, roughly speaking, three main types of montmorillonite: acid, sodium and calcium. The acids are montmorillonites obtained by treating natural minerals generally extracted as sodium or calcium montmorillonite. Generally speaking, the acids are extremely difficult to deflocculate, unless deflocculants that raise pH sharply are used. On the other hand the sodiums are the easiest to deflocculate and can be fluidized with STPP and polyacrylates. Calcium montmorillonite is the mineral commonly found in clays for ceramic use; it is more difficult to deflocculate than the sodium type and generally requires phosphate or polyacrylate-based mixes. In any case, use of this mineral must be tightly controlled and it is good practice to limit quantities to no more than 6% in lean bodies and 4% in plastic bodies. Chlorites: with respect to the above-described minerals, defining the rheological characteristics of a chlorite is extremely problematical. This is because chlorite, while one of the most common clayey minerals, is given its genesis always accompanied by illite or smectite. It is thus extremely difficult to discriminate between the rheological behaviour of these minerals and chlorite. One can only affirm that, given the structure and mineralogical similarities, the presence of chlorite in a clay is generally to the detriment of rheological characteristics with the appearance of thixotropy and the presence of a yield point. In concluding this brief analysis of the rheological characteristics of clays, it should be underlined that only laboratory tests can provide accurate knowledge as to the rheological behaviour of a clay. Mineralogical and chemical data only allows us to hypothesize general behavioural rules useful for evaluation on the basis of the mineralogical characteristics of the clays: yet such rules will always be approximate in nature. Rheology of ceramic bodies Ceramic bodies are essentially made up of a plastic clayey fraction, a hard quartzose (inert)/feldspathic (fluxing) fraction and limited quantities of chamotte, carbonates, recycled materials and sludge, colorant oxides and opacifiers. The appropriately batched set of raw materials is loaded into a mill with the required quantities of water and deflocculant and ground to a pre-determined residue. Since the clayey fraction spans a particle size distribution range below that of normally required residues it is simply broken up or dispersed to its fine phase, while the harder elements are subject to a true grinding action by impact and rolling.

204


On the basis of the mineralogical and chemical composition of a body, we can hazard a rough rheological classification of the components as follows: a) colloidal clayey fraction, the effects of which were discussed in the previous chapter. b) inert or fluxing materials virtually insoluble in water, such as feldspars, quartz, zirconium silicate, alumina... etc. c) soluble materials in appreciable quantities: alkaline salts and earth-alkalines contained in the raw materials and in any additives in the form of waste, sludge. Non-plastic insoluble products only act on slip rheology as number of particles per unit of volume or by crowding the system with particles having a high specific surface area (which increases enormously as particle size reduces). Their effect is largely that of corpuscles which, from a chemical-physical viewpoint, are substantially inert, yet active in a rheological sense in that they occupy a space already crowded with masses and electrical charges; they therefore oppose a certain resistance to layer flow and influence rheological properties. As already seen, soluble ions can approximately be divided into flocculating ions (small and highly charged) such as Ca2+, Mg2+, Fe2+, Fe3+, Al3+, Zn2+ and other cations such as Na+, K+ and, less frequently NH+4, as well as the anions: sulphates, chlorides, carbonates and bicarbonates and borates. They may come from clays, sludges and, above all, grinding water. The importance and effect of the flocculating ions is easily understood: they contribute to a reduction of the zeta potential between the micelles by reducing their repulsive forces. Somewhat less obvious is the rheological relevance of the other ions, which can be explained as follows: overcrowding of charges, and thus a decisive increase in the total ionic strength of the aqueous phase, has a negative influence on the zeta potential and obstructs the exchange processes. The presence of the other ions is, then, a hindrance to deflocculation even where the concentration of true flocculating ions is relatively low. With regard to the relative importance of the three main ceramic body components, it can certainly be stated from a rheological standpoint at least that the plastic clayey fraction plays the lead role in slip behaviour, although the influence of soluble salts is often decisive. The effect of the non-plastic inert materials, in comparison to that of the other two components, is decidedly secondary. Influence of grinding water Lab tests using de-ionised water have confirmed that soluble ion slips prepared with de-ionised water show much better viscosity than all the others prepared with

205


potable or industrial water (other variables remaining equal) for reasons already cited (presence of flocculating ions, high total ionic charge). Furthermore, ceramic body grinding water is often the worst in the whole factory, containing runoff and waste from other departments (i.e. glazes, various organic substances and so on). However, even clean water recovered from waste beneficiation can contain small quantities of flocculants (ferric chloride, aluminum chloride, lime, polyacrylamides) and can have worse effects from a rheological viewpoint than untreated waste water. A further grinding water factor is pH: as already mentioned, there is an optimum deflocculation pH range (normally 8 to 9). Water with pH values outside this range will make the whole process more difficult. From these observations it can be concluded that periodic inspection of the grinding water may be useful, if not necessary. Rather than a true chemical analysis, which would be difficult to carry out inside most ceramic production plants, it might be preferable to run the following tests as they can be completed quickly and do not require the use of costly equipment: a) pH: using a pH-meter. The importance of pH is cited in the preceding paragraphs. b) electrical conductivity: using a conductimeter (simple hand-held versions are available), giving an approximate idea of total ionic concentration. c) total hardness: determined by titration, measures the quantity of flocculating ions (Ca2+ and Mg2+). d) COD (Chemical Oxygen Demand): a measure of the oxidizable organic substances in the water. This assessment, while more complex than the others, is done by titration. Periodic and frequent control of these parameters will provide sufficient (but incomplete) grinding water quality data, allowing correlations to be drawn between changes in such parameters and any worsening of the rheological characteristics of the slips. It also allows, albeit in an entirely empirical fashion, the establishment of a good water range outside which action needs to be taken (e.g. dilution with non-recycled water). The wet grinding process and spray drying of the ceramic bodies from a rheological standpoint The entire process can be summarised as follows: a) batching of main and auxiliary raw materials, water, deflocculant and loading of mills. b) homogenisation of the mass inside the mill as it starts to rotate.

206


This part of the process is practically ineffective in terms of grinding, and is of extremely variable duration. c) true grinding: this stage reduces solid particle size and adjusts the particle size distribution curve, shifting it towards lower and lower values. d) unloading of mills. e) screening of the slip using a series of vibrating sieves. f) in-tank slow-stir storage. h) transfer into smaller service tanks, high-pressure pumping and spray drying. During the settling and homogenisation stage, characterised by extremely variable density, the deflocculant dissolves (where this has been introduced in a solid state) or is, in any case, dispersed within the aqueous phase. The efficiency of this first stage often determines the outcome of the entire grinding process, and the development of negative phenomena at this stage can have long-lasting effects, sometimes causing serious unloading problems or low grinding effectiveness. Jamming of mills, observed in systems with a high plastic material content, is a classic example. The rotary movement creates compact spheroidal masses (blocks) of greatly differing size, which in worst-case scenarios actually envelop the grinding media: these generally consist of clay or other materials lined with layers of plastic clay. Where they are small they tend to float on the suspension; sometimes, instead, they snowball, gathering up solids as they roll and when grinding is over they have to be broken up mechanically. The result is that the jammed material is not ground and final residue is abnormally high. Exactly why this happens is not clear, although the poor wettability of certain raw materials, overly irregular loading (sometimes causing stratification) and the accumulation of deflocculant in only certain zones (thus delaying its dispersing action) are all thought to be contributing factors. The appearance of such phenomena can be prevented by introducing or improving pre-mixing of all the body components, adding water simultaneously with the raw materials during mill charging and, above all, using deflocculants that are already in solution and/or dissolved in the grinding water. If this line of action fails to resolve the problem then the trouble may lie with the grinding media, in which case the load will need to be reviewed in terms of quality, quantity and media size distribution. A combination of these actions will shorten this initial phase and rapidly bring the system to a degree of homogenisation that renders the action of the grinding media effective, thus reducing the total grinding time.

207


As soon as a sufficiently homogeneous, constant-density slip is formed the rheological characteristics of the system take on more significance. At this point the slip may be described, by and large, as an unstable, pseudoplastic, thixotropic suspension with a yield point. In an analysis of the grinding process true and proper, variations in viscosity and residue (taken as an index of particle size fineness) are illustrated in the graph in Fig. 105. Grinding residue (curve A) decreases and finally stabilises: extending the grinding time beyond this point yields no advantages whatsoever and merely wears the grinding media. Viscosity, instead, initially declines sharply, stabilises during a secondary variable-duration phase and then recovers over the final phase as grinding continues. This behaviour can be explained as follows. The initial falloff in viscosity is caused by the action of the deflocculant; the constant-viscosity phase is linked to the fact that the clayey portion, which is more involved in the interactions with the deflocculant, is already, at this point, in a homogeneous system of a particle size distribution well below that set by the residue; the final increase is a result of various factors:

Viscosity Residue

time Fig. 105. Changes in viscosity and residue during wet grinding.

208


a) particle size reduction of the inert fraction causes a higher concentration of particles of higher specific surface area. b) system temperature may increase beyond the optimal range (40-50 °C). c) grinding time, temperature and physical-mechanical stress can lead to partial deterioration of the deflocculant or have a negative effect on exchange mechanisms. At this stage of the process, where the system is subject to continuous and intense shear stresses, viscosity is the most important factor: it must be low enough to allow proper movement of the grinding media but not so low as to cause their rapid wear, which would result in poor grinding efficiency. This is particularly true where high-density grinding media are used. Some examples of viscous grinding have, in fact, led to a reduction in overall grinding times. Unloading difficulties can effectively be resolved by adding small quantities of deflocculant (generally 10-20% of the total quantity) a few minutes before the end of grinding: this lowers the yield point and results in smooth slip outflow. Highefficiency liquid deflocculants and automatic or semi-automatic dosing systems make the task considerably easier. Emptying or unloading the mills normally takes from 15 minutes to 1 hour; the slip is subject to low shear stress and factors such as yield point and thixotropy count for more than viscosity in itself. It is at this stage that thixotropic slips with a high yield point tend to take on a pseudo-solid structure (in jargon, they are said to ice or gel), forming a surface crust and leaving significant quantities of residue on the grinding media and the mill lining, thus making unloading tasks long and difficult. It is interesting to note how this effect is never seen while the slip is kept moving, yet increases progressively under standstill or slow-flow conditions. Systems which, at the end of grinding, reach and exceed 70-80 °C are a case aside; in addition to having higher viscosity, these often form surface crusts because of the increase in local density induced by water evaporation. Note also that excessive end-of-grinding temperatures are not only useless (or detrimental) as regards the rheological characteristics of the slip: they also imply wasteful use of energy and may be indicative of problems or irregularities in grinding media size and/or the ratio between the latter and the mill load. Sieving is sometimes an integral part of unloading or is sometimes carried out afterwards. It will be fast where yield points are low since an important trend is towards finer mesh sizes to maximise the removal of potentially damaging impurities. Sieving is often the slowest part of the entire production process. Consequently, a number or series of sieves are used.

209


At this point the body slip goes into holding tanks with low-speed agitation. Holding times vary greatly depending on the size of the plant, ranging from 6-8 hours to 12-15 days. Under these conditions (very low shear stresses and extended duration) timedependent phenomena such as thixotropy become crucial, especially where tanks are full and there are nearly-still areas of slip. Consequently, the formation of surface crusts often a tell-tale sign of impending gelling or thickening of the entire mass are sometimes observed: this event is known and feared by all ceramists, as it is the forerunner of a whole series of difficulties. To recover from this problem pre-diluted liquid deflocculant is added but it may take time to disperse throughout the tank. Ceramic suspensions have inherently poor stability and should not be held for prolonged periods unnecessarily. Where unloading and sieving problems or thixotropy-related in-tank viscosity increases occur continuously and frequently an in-depth rheological analysis and review of the entire process may well be needed. This chapter does not aspire to give a precise, exhaustive explanation of (or easy solutions to) problems that often stem from a host of concurrent factors. It does, however, intend to scrutinise the various factors contributing to the rheological characterisation of a body slip and provide guidelines that allow producers to predict (if not prevent) the phenomena and take effective countermeasures: 1) 2) 3) 4) 5)

Increased viscosity and other phenomena may be attributable to: raw materials grinding water additives deflocculants particle size distribution.

On closer analysis, we can state: 1) An increase in viscosity and plastic behaviour may be attributable to changes, in a plastic sense, in one or more raw materials (mostly the clayey ones); the more frequently the raw materials are topped up, the more probable this becomes. Variations in the moisture content of the clayey components can also lead to changes in viscosity, yet such changes are always accompanied by easily recognisable increases in slip density. If, instead, the mineralogical and/or chemical composition of some of the body components has altered, identifying the guilty party will be extremely difficult without the aid of accurate chemical and diffractometry analyses, and even these sometimes fail to clarify the problem. These plastic increases can sometimes be identified through repeated measuring of unfired bending load or shrinkage on both unfired and fired pieces. The only way to prevent the problem is to run accurate rheological tests

210


on the raw materials at the same time as the usual technological ones (absorption, shrinkage, firing colour, presence of carbonates etc.). The same phenomena occurs when errors in formulation, weighing and so on result in combinations overly-rich in plastic components being fed into the mills. 2) Variation in the composition of the water (see preceding pages). 3.A) Addition (or increase) of plasticizers or glues. Any introduction/changes to quantity of such products in a body must always be preceded by laboratory tests. 3.B) Addition of sludge, glazes and recycled materials from glazing plants: once again, laboratory testing is essential. 3.C) Addition (or increase) of exhausted lime from kiln filters. As calcium ions cause flocculation in slips, dosing must be calibrated with care. With respect to lime-free loads extra deflocculant must be added according to the quantity of lime that has been introduced: in the most difficult cases this ranges from 50 to 100%, and has led some companies to abandon the use of recovered lime in grinding altogether as it is anti-economic with respect to other solutions (e.g. assistance from specialised companies). 4.A) Drop in the percentage of deflocculant caused by erroneous dosing. 4.B) Drop in percentage of deflocculant as the latter itself absorbs the water. A fairly rare phenomenon that is usually easily visible to the naked eye. 4.C) Wrong type of deflocculant (supplier error). Identification of this cause is fairly easy in that the difficulties coincide with change-overs in the product being used. 5.A) Variation in particle size distribution in that there is an increase in the fine or super-fine fraction owing to an extended grinding time (drop in residue). A marked increase in grinding time generally leads to increased viscosity. 5.B) Like 5.A but brought about by the addition/increase of very fine recycled dust or unfired scrap. In these cases too it is highly advisable to effect lab tests before making any definitive changes. The stocked and sieved slip is pumped to the spray drier at a pressure of a few tens of atmospheres (on average about 30 Atm) using piston pumps.

211


From a rheological viewpoint this is perhaps the least understood stage of the entire process, as the physical stress the system is subject to is virtually impossible to reproduce using laboratory instruments. On the whole though, it seems safe to state that the system, where subject to very high yield stress, behaves in a Newtonian manner in a laminar fluxing situation, at least as far as the nozzle outlet where, aided mechanically, it rapidly transforms into a disordered cloud to form the droplets that, following interaction with heat, will become spray-dried grains. If this fails to happen (or happens too late) the jets released from the nozzles are projected and build up on the inner surface of the chamber, which if left can seriously affect the performance of the drier. Density, viscosity and other rheological parameters undoubtedly influence the quality of the spray-dried product, but correlations between these parameters are difficult and individual experiments are necessary in each case.

212


Rheological additives The main rheological additives are given in the table below: ADDITIVES FOR DEFLOCCULATION INORGANIC

ORGANIC

Na2CO3

Humic acids

Na2O. nSiO2

Tannins

Polymeric phosphates (NaPO3)6

Derivatives of acrylic acid CH2=CHCOOH

Na5P3O10 BaCO3

Derivates of oxalic acid (COOH | COOH)

In everyday language, the term deflocculant refers to a substance which, when added to an aqueous suspension of dispersed colloidal powders, causes greater fluidity of the suspension, diminishing its apparent viscosity. Deflocculant is a synonym for fluidizer and merely correlates the deflocculation process with a decrease in viscosity, neglecting its function as a dispersant and a preventer of the rapid aggregation that would cause precipitation of the solid particles suspended in the water. Deflocculants are compounds which impede flocculation by way of a higher zeta potential and a relative increase in the repelling forces between the particles. A good dispersing action creates very high zeta potentials, thus giving minimum viscosity. It follows, then, that the term deflocculant includes both the basic concept of dispersion and that of fluidization. A dispersant is that prevailing fluid phase into which the non-mixable solid or liquid fractions are dispersed (such fractions thus represent the dispersed phase). In most applications the dispersant is water.

213


Deflocculation is that process in which the solid colloidal particles in the dispersant fluid (e.g. a slip) move away from each other yet remain in suspension because of the repelling electrostatic action that the deflocculant substance induces (by increasing their zeta potential). A fluidizer is a substance capable of making a fluid flow more smoothly (i.e. lowering its viscosity). It is the exact opposite of a thickener, an agent that increases the consistency of a fluid mass. A deflocculant is a substance which, when added in small quantities to a fluid mass, is capable of preventing agglomeration of the colloidal particles and thus their precipitation. Since this effect can be achieved via dispersion (i.e. by addition of the fluid phase) the two can easily be confused; while it is certainly true that a deflocculant also acts as a fluidizer, it is not necessarily true that all fluidizers are also deflocculants. The fundamental mechanisms that explain the deflocculating action may be explained as follows: l. shifting of the pH towards basic values by introducing OH- ions into the H2Osolid system via an addition of monovalent bases or basic electrolytes which, by hydrolysis, give OH- ions (an excess causes over-deflocculation). 2. substitution of other cations constituting the positive side of the diffused double layer with Na+, K+, Li+, NH4+. 3. an increase in the negative charge on the clayey particles by adsorption of certain types of anion (adsorption is preferential for anions of higher valence with a strong electric field). 4. increase in the total negative charge of the solid-liquid system, assuming a nonionic colloid carrying a negative charge. 5. addition of a shielding colloid that shields the suspended particles from the action of flocculants. 6. elimination of any flocculants: by precipitation of the flocculating ion Na2CO3 + Ca2+ → CaCO3 BaCO3 + Ca2+ → CaCO3 BaCO3 + SO42- → BaSO4

↓ ↓ ↓

+ 2 Na+ + Ba2+ + CO32-

via formation of coordination complexes [XA]B where X = flocculating cation (Ca2+, Fe3+) (Polyphosphates). The most complex deflocculants act via a combination of the above-described mechanisms. Deflocculants may be inorganic or organic. Inorganic deflocculants are electrolytes (monovalent bases, basic electrolytes, carbonates, silicates and sodium phosphates). Organic deflocculants may or may not be electrolytes. Whether organic or inorganic deflocculants are used, it is necessary to take into consideration not only their

214


undesirable side effects, but also the effect they have on the dehydration rate, plasticity and bending strength of the body. The influence of the deflocculant on thixotropy and its effects during the downstream forming phase (adhesion to dies, chemical aggression on mould) must also be evaluated. Inorganic deflocculants are more sensitive to the nature and the quantity of ions already present in the suspension, are not eliminated during the drying and firing phase and do not generally have any thixotropic effects. Concentrations being equal, organic deflocculants are more effective, less sensitive to interference from other ions and have greater stabilising power. They volatilise during firing, are usually thixotropic and are more expensive. Choosing the right deflocculant is a case-by-case affair and will necessarily involve practical tests. The most complex problems can be resolved by using compatible mixes of organic and inorganic substances that provide good synergic performance. Inorganic deflocculants Na2CO3 Clay Ca2+ + Na2CO3 → Clay Na+ + CaCO3↓ Ca(HCO3)2 + Na2CO3 → CaCO3 ↓ + 2NaOH + 2CO2 These fundamental reactions make sodium carbonate an excellent deflocculant: the more Ca2+ is present the better the effect. The property of precipitating the calcium, while advantageous for the purposes of deflocculation, is, for example, somewhat less so as regards the life of the plaster casting moulds. CaSO4 + Na2CO3 → CaCO3 ↓ + Na2SO4 The aggression of the salt Na2SO4 lowers deflocculating power and damages the plaster. Na2CO3 , which melts at 850 °C, increases plasticity and the dry bending strength of bodies but reduces the drying rate. Na2O - n SiO2 The sodium silicate, which hydrolyses easily with an alkaline reaction (pH > 11) gives Na+ and OH- by separating the colloidal silica Na20 - n SiO2+ H2O → n SiO2 + 2Na+ + 2OHThe silicate is not as aggressive towards the plaster as the carbonate and has binding properties. Where sulphates are present its deflocculating power is reduced. 215


The mix: Na2CO3 + Na2O - n SiO2 has the advantage of acting in a good four of the six ways described in the preceding paragraph (namely 1, 2, 5 and 6). It thus has a higher deflocculating power and results in fewer defects being caused by the aggression of Na2CO3. These are stable solutions that let gelatinous silica separate. Additions usually fall in the 0.2-0.6% range (by weight). Phosphates, Polyphosphates Those with the most remarkable deflocculating power are sodium hexamethaphosphate (NaPO3)6 and the polyphosphates, especially sodium tripolyphosphate (Na5P3O10). The mechanism exploits the fact that the phosphoric anion is one of the anions that is preferentially adsorbed by the clay particle, increasing its negative charge. The consequent increase in the zeta potential makes a key contribution to a greater dispersant effect. They are also capable of seizing flocculating ions to form coordination complexes. An example is provided by hexamethaphosphoric acid (H6P6O18) which, with the Ca2+, Mg2+, Fe3+ ions forms complex compounds of the following type: [CaP6O18]4-, [Ca2P6O18]2The considerable deflocculating power of sodium hexamethaphosphate can be attributed to the action of the Na+ cation, the strong anionic adsorption and the seizing of the flocculating cations. The polyphosphates behave in a similar fashion: some of the most interesting of these are: Na5P3O10 (sodium tripolyphosphate) in alkaline oxide Mn+2 Pn O3n+1 is the general formula with M = monovalent metal If n = 1 M3PO4 (orthophosphates) If n = 2 M4P2O7 (pyrophosphates) If n = 3 M5P3O10 (tripolyphosphates) If n = 4 M6P4O13 (tetrapolyphosphates) For n > 4 vitreous-structure mixes form. These compounds too seize the flocculating cations under the form of complexes. The polyphosphates have a chain-like structure and have many uses; even where dosed at < 0.4% they are still effective as dispersants on clayey materials. Deflocculating suspensions with polyphosphates break down over time and increase their viscosity when heated. BaCO3 The presence of SO4= obstructs the processes of deflocculation (reduction of

216


the zeta potential) especially in the presence of the ideal flocculant Ca2+, Mg2+, Fe3+ etc. This is because the SO4= anion (unlike other anions such as the phosphoric one) is easily adsorbed by the clayey particle even in substitution of the OH- ions. This happens even if SO4= is present in the form of a strong electrolyte (e.g. Na2SO4): other strong electrolytes such as NaCl act similarly. The ideal thing is thus to remove the SO4= groups. As mentioned above Na2CO3 reacts with the sulphates (e.g. CaSO4) and CaCO3 precipitate forms, removing the Ca2+ flocculating cation, but failing to eliminate the SO4= anions which always remain (like Na2SO4 soluble in H2O) even where silicates or sodium phosphates are present: CaSO4 + Na2SiO3 → CaSiO3 + Na2SO4 3CaSO4+ 2Na3PO4 → Ca3 (PO4)2 + 2Na2SO4 To make Na2SO4 insoluble it is necessary to add barium compounds: Na2SO4 + BaCO3 → BaSO4 ↓ + Na2CO3 ↓ The latter acts as a deflocculant. In place of BaCO3 (almost insoluble), soluble BaCl2 could be employed, but this is only possible via stoichiometric dosing of the sulphates present in the materials to be deflocculated in that the excess Ba2+ ion acts as a dangerous flocculant. Hence BaCO3 is preferred on account of its limited solubility. BaCO3 should be added before the deflocculants in quantities ranging from 0.02 to 0.1%. Organic deflocculants Organic deflocculants are also electrolytes or polyelectrolytes in the form of sodium salts and act according to the mechanisms described in the inorganic deflocculants section. Their anion has a colloidal property and aids the deflocculating power of the various compounds. Humic acids and derivatives Extracted from humus, peat and lignites, these are oxycarboxylic acids of high molecular complexity with furanic rings and different functional units with the following properties: negative radical of a colloidal nature, solubility in alkalis even if diluted, insolubility of their salts with bi or trivalent metals. It is believed that humic acids wrap up the clay particles to form a clayey-humic compound, thus inhibiting the action of the flocculating cations, and so involving other mechanisms. Just how much the protective sheath that wraps up the mineral particle disturbs the action of the typically deflocculating Na+ is, however, unknown. 217


The colloidal clay-humic acid mix is very stable and thus difficult to disperse. A commercially available humate-based deflocculant has been obtained by treating the lignite with soda. Tannin compounds Tannins of a colloidal nature and those of vegetable origin such as the humic acids (used in tanning processes owing to their insolubility and because they prevent the leather gelatine from rotting), have deflocculating properties. The prototypes are tannic acid and gallatannic acid (a mix of glucose and gallic acid esters), the latter belonging to the vast group of the phenolic acids, organic compounds having a molecule that simultaneously houses more than one carboxylic group and one or more phenolic hydroxyls. From this compound for the elimination of CO2, pyrogallic acid is obtained. The alkaline salts of such compounds (sodium tannate, sodium gallate, pyrogallates) are all good deflocculants on a par with the humic derivatives. It has been hypothesized that the tannin molecules are adsorbed at the surface of the dispersed particles with the aromatic part protruding into the dispersant. This would alter the exterior of the particles with variations in hydration and consequent lowering of viscosity. Acrylic derivatives Derived from acrylic acid CH2 = CHCOOH, with an acid force greater than that of acetic acid, CH3 COOH. Na+ ed NH4+ salts can easily be obtained via substitution of H+. Acrylic acid gives rise to a wide range of polymers with which it is possible to form acrylic resins, important groups of thermoplastic resins. Furthermore, the double ethylenic bond is another point capable of reacting. Also of interest are compounds of the CH2 = CHCOOR type where R = metal, alkyl If R = CH3 (acrylic ester) If R = C2 H5 (polyester) Others of importance are: CH2 = CHCH acrylonitrile CH2 = CHNH, acrylamide Acrylic acid and its derivatives polymerise easily under the action of heat, light and peroxides (benzoyl), giving rise to (polyanionic) polyacrylic acids with the corresponding polyelectrolytes, the polyacrylic derivatives, polyesters, polyacrylnitriles etc. COOR CN | | - CH2- CH - CH2 - CH - CH - CH2 - CH - CH2 | | COOR CN 218


Polyacrylic acid dissolves fairly slowly in water and its alkaline salts are easily soluble. Polyesters and polyacrylnitriles are insoluble. It should be noted how certain electrolytes obtained by way of a partial alkaline hydrolysis of the polyacrylnitrile have considerable aggregating power. Others are effective dispersants. The length of the chain is fundamental. Those of greater molecular mass act as flocculants. Polyacrylates are electrolytes and perform their deflocculating action according to the rules typical of the deflocculant; the polymer anion is easily adsorbable by the clayey particles, thus guaranteeing an exceptional dispersant action and excellent stability over time. Acrylic deflocculants appear to be better than traditional ones (silicate, carbonate, NaOH, polyphosphates), seeming less sensitive to overdeflocculation and interfering cations. Their thermo-plastic nature, however, can make their employment difficult. Fig. 106 compares acrylic deflocculants with other deflocculants such as silicate and sodium tripolyphosphate which alter the viscosity of a slip as a function of the quantity introduced.

Brookfield viscosity (epoises)

Polypropylate Na

Acrylic RLP/4 Acrylic R727/S

Acrylic R747

Percentage of deflocculant

Fig. 106. Comparison of how various deflocculants, applied in different percentages, affect the viscosity of a white tile body (by G. Malandrino).

219


A powdery product consisting of a compatible mixture of organic and inorganic fluidizers in quantities ranging from 0.16 to 0.22% is known to confer better bending strengths than those obtained with standard deflocculants such as NaTPP, one of the most effective fluidizers for inorganic suspensions available. It is, however, overly sensitive to small variations in water content and thus a mixture of organic deflocculants having optimum ionicity characteristics so as to obtain maximum stability of the dispersion has been proposed. Ammonium derivatives Substituting NH3 (ammonia) with one or more organic radicals gives methylamine (RNH4) and diethylamine etc. The amines are basic substances which dissociate in water. Aliphatic amines are more dissociated in aqueous solutions than NH3 and are thus more basic. Aromatic amines are less basic. Many substances such as ethyl amine, diethylamine, polyvinylammine and piperidine are good deflocculants but are used only rarely. Oxalates

COOH These derive from oxalic acid | which is quite a strong acid and can form COOH complex salts. Sodium and ammonium oxalate are used little as deflocculants: these are both soluble in water, unlike the other oxalates which are insoluble. Sodium and ammonium oxalate have the property of making the Ca2+ precipitate completely. The oxalate anion may also be adsorbed by the surface of the clay particles. These compounds are also used together with others having a higher deflocculating power. Other substances Other deflocculating substances worth remembering are the sodium salts of the polymerised alchil-naphtalen-sulphonic (pH 8-10,5) acids, lithium citrate and the sodium derivatives of carboxymethyl cellulose (CMC), an electrolyte with colloidal anion (Fig. 107). The CMC anion is irreversibly adsorbed at the surface of the mineral particles, thus increasing their negative charge and dispersion capacity. The smaller the degree of polymerisation, the greater the observed deflocculating effect. CMC is used in ceramics as a binder while, however, bearing in minds its deflocculating properties. Main additive classes Effects, advantages and disadvantages of practical use. Side effects So far, then, we have seen that a correct rheological approach with regard to the

220


Cellulose

Caboxymethyl cellulose

Fig. 107. Cellulose and carboxymethyl cellulose (CMC) structural formula.

aqueous suspension systems used in ceramics is of great importance and the main operative mechanisms of rheological additives have been illustrated. As seen, rheological additives are used because it is necessary to work with aqueous suspensions of a high solid content; yet we must be able to modify and/or maintain precise characteristics of viscosity, plasticity, thixotropy etc. vis-à-vis the manner in which the solution will subsequently be used. For this reason various classes of rheological additives have been defined as a function of the role they play in the preparation and control of clayey bodies and glazes. Ceramic spray drying bodies While costly, the technological process of wet-grinding raw materials in Alsing ball mills and then spray drying the obtained slip is the best because it produces bodies that are homogeneous, impurity-free and with the optimum granulometric characteristics required for subsequent pressing and firing in rapid-cycle kilns. This set of processes involves an enormous quantity of water (30-40% of the total mass, reduced to 5-7% during spray drying). The use of rheological additives, which in ceramics means deflocculants or fluidizers, is thus essentially linked to the need to operate with as low a percentage of water as possible without compromising the viscosity, plasticity etc. that will make emptying of the mills, in-tank holding, slow stirring, sieving, pumping and spray drying practical. The criteria are thus economicproductive: a reduction in the water content, other conditions remaining equal, results in a lower volume of water to be evaporated and greater plant productivity.

221


Of coarse, the whole process also depends on other factors such as the grinding time and the cost/performance ratio of the deflocculant. The main body deflocculants, then, are: sodium carbonate sodium silicate various sodium polyphosphates natural organic compounds (humic acids and humates, tannins, ligninsulphonates...) 5) synthetic organic compounds (polyacrylates, polymethacrylates and derivatives). 1) 2) 3) 4)

Already described in the preceding paragraphs, we shall now summarise their main characteristics. Sodium bicarbonate used only rarely in spray drying bodies and only in those bodies with very low plastic clayey fraction content (white body double firing) because of its low fluidizing effect. Sodium silicate a colloidal protector often used, for example, in sanitaryware casting slips, together with sodium carbonate which shifts the calcium from the slip. The effect of the sodium silicate is to coat the clayey micelles in mixes without polyvalent ions; it is thus highly suitable for slips without organic colloidal-protectors (i.e. those containing low-plasticity clays). Natural organic products used only very rarely. Polyphosphates and polyacrylates used extensively, especially sodium tripolyphosphate. Evaluating which product (or products) is the best is somewhat difficult. In any case it is good practice to effect comparative industrial and laboratory tests for each individual body, which has its own specific characteristics. Side effects Common to all deflocculants not just those used in ceramic bodies is the phenomenon of over-deflocculation. This is an increase in apparent viscosity caused by an excess of deflocculant. Organic products, where introduced into bodies where a situation of compromise or risk already exists, tend to emphasize those black core reduction phenomena caused by poor permeability or problems connected with a firing curve that is unsuitable for oxidation. This tendency is much less marked in synthetic products. Another effect (this time a positive one) sometimes observed, although not yet fully understood, is a possible increase in the unfired strength of the pressed items. This is mainly connected with silicates or sodium polyphosphates.

222


Ceramic glazes As seen, in the ceramic industry the use of rheological additives is not so much governed by economic-productive aspects as by the absolute need to work with suspensions having certain chemical-physical properties that are suitable for on-tile application and compatible with different application technologies which often involve machines with extremely different operating principles. In this case, rather than sub-dividing the most commonly employed additives on the basis of characterisation, they have been classified according to the effects they produce: RHEOLOGICAL ADDITIVES FOR GLAZES: 1) Deflocculants 2) Suspension agents 3) Glues (binders) 1) Deflocculants the main effect these have on a glaze slip is to lower its apparent viscosity. These products can be added at the grinding stage or on the glazing line (the latter being suitable for liquid or easily soluble products only). The main product classes are the already-seen ones of the sodium polyphosphates (tripoly, tetrapoly...), the sodium or ammonia derivatives of polyacrylic or polymetacrylic acid or copolymers of more complex structure. The advantages of high density grinding (i.e. with a high percentage of solid) are as follows: reduction of grinding times greater uniformity of particle size distribution better reproducibility greater versatility in subsequent applications. These additives are used in varying percentages, depending on the efficiency of the product and the desired effect: generally, percentages range from 0.03% to 0.4%. Additions along the production line are mainly to correct viscosity to the specification required by the application machinery and the aesthetic effects instantaneously. Another (positive) effect sometimes observed, the mechanism of which has yet to be fully understood, is a possible increase in the unfired strength of the glaze, largely connected with utilisation of silicates and sodium polyphosphates.

a) b) c) d) e)

Side effects Can be summed up as follows: glaze dries more slowly improved spreading, especially with disc application slower short-term sedimentation rates formation of extremely tenacious sediments (or cementing) in the medium and long term reduced plasticity and thixotropy in high-density glaze suspensions where bell application systems are used. 223


Overdosing effects a) viscosity too low and flow excessive: may cause irregular application (e.g. accumulation at sides) b) overdeflocculation. SUSPENSION AGENTS form another category of rheological additive. These act by way of electrostatic effects on the inter-atomic or inter-molecular bonds where soluble salts in polar solvents (water) are concerned, or by increasing viscosity (CMC) or the colloidal load in suspension. In any case, they have the effect of reducing sedimentation rate: this, in turn, depends on a complex series of factors such as the specific weight and particle size distribution of the suspended solids, the density and viscosity of the slip and the presence of colloids.

The most commonly used products are: sodium chloride electrolyte mixes in solution cellulose derivatives (high-viscosity CMC) bentonite clays colloidal silica.

The final group of additives is the glues (binders). Certain glues are used in ceramics as glaze adhesives. There are various types: methylcellulose, ether and ester starches etc. Undoubtedly the most commonly used glues are the carboxymethyl celluloses (CMC). Use is made of their properties as binders, in that they improve the cohesion of the unfired glaze particles and the adhesion of the latter to the tile body, as evaporation regulators (they slow down evaporation of the liquid phase), and they improve the spreading characteristics of the glazes. Rheological effects strictly depend on the type of CMC used: a rough classification, based on the viscosity of the solutions at 2%, can be made as follows: low-viscosity CMC (between 5 and 50 mpa.s approx., where mpa.s = milli pascal × second) medium-viscosity CMC (between 100 and 1000 mpa.s approx.) high-viscosity CMC (between 1000 and 10,000 mpa.s and beyond) The corresponding rheological effects, at standard usage percentages of 0.10.6%, are: low-viscosity CMC causes a slight drop in viscosity in the suspension and tends to lead to cementing medium-viscosity CMC limited effect on viscosity and a light suspending effect high-viscosity CMC causes an increase in viscosity and has a high suspending effect. Overdosing, of course, causes excessive increase in viscosity, extended drying times, irregular absorption of the glaze by the body. Some interesting rheological examples are given in Figs. 108-113. 224


Fig. 108. Rheological behaviour of a bell-applied double firing glaze: high viscosity (indicated by graph gradient), zero yield point. (Curve obtained after stressing the fluid at the maximum value of D for 1 minute).

without additive

Fig. 109. Effect of adding small quantities of sodium tripolyphosphate (NaTPP) and carboxymethyl cellulose (CMC) to a glaze with high yield point. (Curve obtained after stressing the fluid at the maximum value of D for 1 minute).

225


without additive

Fig. 110. Effect of adding high molecular weight CMC to a low-viscosity glaze. (Curve obtained after stressing the fluid at the maximum value of D for 1 minute).

Fig. 111. Rheological behaviour of a disc-applied glaze. (Curve obtained after stressing the fluid at the maximum value of D for 1 minute).

Kaolin Without additive

Fig. 112. Effect of adding kaolin and bentonite to a glaze suspension.

226


Fig. 113. Effects on rheological properties of increasing the apparent density of a glaze: as density increases there is a marked increase of all parameters: viscosity (gradient), thixotropy (area of hysteresis), yield point (shift from origin).

Glazes: non-rheological side effects a) b) c) d) e) f)

Effects of a non-rheological nature are linked to the presence of: polyphosphates sodium chlorine soluble salts other non-volatile inorganic products (SiO2 , Al2 O3 ...) various organic substances.

The effects may be catalogued as follows: a) During firing the polyphosphates lead to the formation of P2O5 glasses, insoluble in SiO2-based glasses: this may lead to the formation of micro-bubbles and black specks. b) The sodium becomes a part of the glasses, exerting a certain fluxing action and giving rise to over-firing pin holes. c) Thermal decomposition of chlorides may lead to formation of gaseous chlorine which damages the rollers, metal parts and the kiln lining because of corrosive condensation. Gaseous chlorine is also environmentally harmful. d) Soluble salts give rise to the phenomenon of migration towards the tile sides, especially during drying: localised staining, marking, pin-holing and bubbling may result. e) Any inorganic compound which is non-volatile at firing temperature becomes part of the finishing glaze, thus contributing to varying degrees to the modification of its composition and alteration of its fusibility and surface aspect. 227


f) Organic substances decompose under the effects of heating, giving rise to gaseous substances which exit via the glazed surface. If there is poor compatibility between the quantity and quality of the organic substances and the firing cycle there may be degassing and black speck problems. The presence of certain organic products (starches, CMC and other cellulose derivatives) also causes biological decomposition via micro-organisms (moulds, bacteria). This mostly occurs in solution. The main effect in a glaze slip is loss of adhesive power, which immediately translates into pin-holing, dimples right across the surface, and then micro-fractures and fissures on the sides. This problem occurs with glazes containing CMC applied some few days after grinding. In truth, though, it should be noted that the rheological additives form only a tiny proportion of total chemical auxiliaries, represented by silk-screen printing products. The latter, in fact, are the main cause of fume filter clogging and increased COD in waste waters etc. Most commonly used additives As far as the BODIES are concerned there is a widespread tendency for producers to use low-cost additives, even where they are less effective. Producers also tend to invest little in new product research. The most widely used product class is that of the fluidizer-deflocculant polyphosphates. In particular, sodium tripolyphosphate is used extensively in quantities of 0.2 - 0.4% (by dry raw material weight). These salts exert an excellent chelating (sequestering) action on the polyvalent ions in the solution and their action is immediate and effective. Another important advantage is the completely inorganic nature of the salt, which is unaffected by pyrolitic phenomena during firing that can have undesirable effects on the ceramic body (e.g. localised reductions where pieces have low permeability, black core etc. and discharge emissions - condensation, fume depuration etc.). Polyphosphates require a well-defined pH range in suspension, which they themselves help determine. This should be between 8 and 9, outside of which they are no longer as effective because of changes in the electrostatic forces between the particles in suspension. The main disadvantages of these additives are that they decay rapidly and soon lose effectiveness. Furthermore, the rather difficult in-water solubility of sodium tripolyphosphate may create non-homogeneous fluidity inside the mill. The polyacrylates are another frequently used class of product; these are generally sodium or ammonia-based and have an extremely intense deflocculating action, especially where added after grinding. These are normally used in percentages of around 0.1%. Since they are perfectly 228


soluble (the products are commercialised in solution), their action is always well distributed and stable over time. The main disadvantages, though, are high costs and the fact that the polyacrylic chain breaks up if added during grinding, thus reducing effectiveness. Furthermore, as they are organic compounds they are less suitable for addition to the body as they produce, by pyrolysis, damaging reducer compounds and ammonia (or, generally, nitrogen-containing) salts which makes the lime in purification units malodorous. If dosage is excessive their effects are irreversible but an increase of sodium and/or ammonium may create undesired effects. Other additives which have been used or tested in a production context are humates etc., applied in percentages varying from 0.1 to 0.2%; while considered to be highly active, they have very slow reaction times owing to their low solubility. This can create serious problems at the mill before they are effective. While generally inexpensive, these additives contain high percentages of carbon, with all its possible consequences. Sodium silicate has the advantage of being very low cost, but its effectiveness is not generalised and it does not exert a truly deflocculating action. Overdosing can cause serious problems, and optimisation curves are very narrow indeed. We shall now take a look at the additives used in the preparation of GLAZES. In general it can be said that double firing glazes do not require any particular additives (sometimes suspension agents are used), single firing glazes always feature the addition of glues and sometimes suspension agents, while monoporosa glazes generally require the addition of glue, fluidizer and sometimes a suspension agent. Particle size distribution of the glaze itself (and thus mill load and grinding media) plays a key role in glaze rheology as does solubility of the frits, which can place cations of enormous rheological interference in the solution, with consequent aging problems. The fluidizers used in actual industrial production belong to the same categories as those used in bodies. One of the already-mentioned suspending electrostatic-action ionic compounds is sodium chloride, normally used at 0.1-0.3%, which gives the glaze more body and increases its viscosity. This prevents dry dips occurring as the glaze tends to yield water more slowly. While its low cost is highly advantageous it should be pointed out that the salt can tend to migrate to the sides of the tile, thus making the unfired glaze layer more fragile (lateral scratches). It is hygroscopic and less effective than other similar-action salts, such as barium chloride, which also sequesters undesired anions highly effectively. There also exist specially produced combined-action salt mixtures with very good storage properties. Excessive dosing of these products results in excessive viscosity increases with deleterious in-tank (ricotta) effects. 229


Other often-used additives include clays and kaolins, normally introduced in quantities of 2-4%. These aid application as they increase viscosity. In monoporosa it is necessary to keep the frit grains separated, therefore the above-mentioned clays are added as they carry finer particles which work their way in between the grains. Adding excessive quantities of clays interferes with the quality of the opaque and/or gloss aspect of the finished product. Other special additives such as ethylendiaminictetracetic acid (EDTA) salts, tetramethylammonia salts, various complexing agents etc. have no practical industrial use, although their effects are studied and they are occasionally tested in an industrial setting. Summing up: optimum rheological characteristics in a ceramic body or glaze suspension can be said to depend largely, for ceramic bodies, on the type and quantity of the clayey materials, the employed grinding technology and accurate control of additive optimisation curves. With glazes, optimum rheological characteristics greatly depend on the application method, the formulation of the glaze itself, the solid/liquid ratio and the quality/quantity of the additives used.

230


Chapter VII THE REMOVAL OF WATER

The water-clay system and how it affects moulding Plasticity allows a ceramic body to be moulded into a shape that will subsequently be fixed during firing. This plastic state may be seen as an intermediate condition in which the material is neither a solid nor a viscous liquid. Applying pressure to a liquid causes it to flow according to the limitations and specifications described in the preceding chapters, yet applying pressure to a solid generates no appreciable changes until the ultimate load is exceeded and the piece breaks. A body with plastic properties behaves in an intermediate manner, as is evident in the case of extruded materials prepared from bodies with a high water content: outflow speed is proportional to applied pressure and continues as long as that pressure is applied, yet the material maintains its shape and cannot therefore be considered a liquid. With dry pressed ceramic materials (with a moisture content of about 3-6%) the body does not have plastic properties as such; nevertheless the mineralogical composition of the components and water content are essential for good pressing results. The very structure of the clayey minerals crystallised strips separated by well-defined gaps allows the lattice sheets to slide against each other when the interlayer water acts as a lubricant. Plasticity in these semi-plastic systems depends, then, on the formation of a film of water around each individual grain of material. The thickness of that layer is crucial: if it is excessive a second layer of free water which does not interact with solid particles or cations in fracture zones forms, generating a different type of flow that is actually detrimental to the plasticity of the system. On the other hand, insufficient quantities of water will cause the particles to touch each other, generating attrition and destroying any plastic properties. System plasticity thus depends on a certain equilibrium which, in turn, depends on water content and wettable surface area. The thickness of the film of water enveloping the clayey particles when plasticity is at its height, while difficult to calculate, has been defined (for an applied pressure of 8 Kg/cm2) as: China Clay Ball Clay Bentonite Brick clay Clayey soils

2100 Å (1Å = 10-8 cm) 2400 3400 3100 700 231


This ideal value can be significantly altered by innumerable factors, which, one way or another, change the state of disturbance and electrostatic interaction between the water and the clayey particles. Such factors include: water content particle size distribution size, shape and structure of particles particle composition aggregation surface area and intermolecular attraction the effects of additives (including natural additives such as salts dispersed in the raw materials) particle orientation. Leaving aside an in-depth examination of the many methods used to measure the plasticity of a ceramic system (most of which apply to extruded products), it should be pointed out that a good indication of plasticity can be obtained from the bending strength of the pressed material (green or dry). Thus a body with good plastic qualities results in better pressing, higher compaction, higher apparent density and, finally, a higher bending strength (consequently, it will also be more difficult to dry). Because of the low water content the pressing process is unable to bring about all those preferential particle alignment phenomena seen in extruded products. The compaction mechanism during pressing will largely be influenced by proper size distribution of the particles and their extreme smallness, which allows almost all the gaps to be filled: furthermore, the film of water lying between the particles must have a consistency that sticks the latter together, thus aiding, among other things, subsequent sintering reactions during firing. It is also important that the material does not stick to the mould surfaces as a result of one-way re-absorption of the film of water eliminated from the body at the ceramic-metal or ceramic-rubber interface. In particularly clay-rich bodies orientation phenomena may occur, with water separating out onto planes perpendicular to the direction of pressing and causing lamination, a possible cause of problems during drying or in-kiln prefiring. Optimisation of these compaction processes through the use of isostatic presses or punches is described in the chapter on ceramic products. Removing the water Water added to raw materials for refinement (e.g. production of pure kaolins), purification or grinding purposes must subsequently be reduced to allow moulding.

232

The removal of water

There are several methods: sedimentation (e.g. sludge clarification) centrifugal separation (little used in ceramics) filtration (e.g. filter-pressing in the manufacture of sanitaryware or tableware) electro-osmosis (e.g. in some types of glazing) evaporation (usually spray drying) selective absorption, on solid vehicles that absorb more water.

Independently of the need to eliminate water from a ceramic material suspensions, in wet grinding, where the spray is slip-dried (see Vol. II on ceramic products) undoubtedly the process that removes the greatest quantity of water it is important that the production process leaves a certain amount of water in the body to lubricate the particles during moulding and, because of the applied forces, produces the preliminary, partial inter-particle adhesion that aids subsequent sintering processes. Extruded products are formed in a highly plastic state and thus the pre-forming residual water content is necessarily high (15-25% by weight). However, in so-called dry pressing the most common tile production technique required water content varies as a function of required plasticity and is generally much lower at just 3-6%. Nevertheless, even this water will have to be eliminated, together with that absorbed by the pieces during decoration and glazing, before the tile enters the kiln. Hence pre-firing drying is necessary. Even though the amount of water to be evaporated is rather modest, the drying phase is an extremely delicate one because of the ease with which fatal mechanical stress can affect the still-inconsistent material. Water, like all liquids (and, to a lesser extent, solids), has a tendency to become gaseous, thus absorbing energy. Vapour emission does not continue indefinitely: when the vapour reaches saturation point it ceases. An equilibrium is thus established between the number of molecules that evaporate and the number of vapour molecules that re-liquefy: this equilibrium corresponds to a certain pressure known as saturated water vapour pressure. Water vapour pressure rises as temperature rises. When, as temperature increases, vapour pressure equals atmospheric pressure, boiling occurs (i.e. the vapour escapes not just from the surface of the liquid but from its interior too, forming bubbles). Boiling point diminishes with pressure and every substance, at a certain temperature, has its vapour pressure. Water (see Fig. 114) has a saturated vapour pressure of 1 Atm at 100 °C. The energies in play, then, taking into account that which is absorbed during the liquid-vapour transformation phase, are often underestimated.

233

Atm p 105 Pa

Water vapour pressure


Temperature °C Fig. 114. Saturated water vapour pressure at different temperatures.

For example: (1) Energy needed to increase the temperature of 80 g of clay containing 20 g of H2O from 20 °C to 100 °C, eliminating the water by evaporation: 20 g H2O from 20 a 100 °C: evaporate 20 g. H2O: 80 g clay from 20 a 100 °C:

m. Cs. ∆T = 20. 1. 80 m. L = 20. 540 m. Cs. ∆T = 80.0.2.80

= 1,600 calories = 10,800 = 1,280 ________ 13,680 calories

(2) Energy needed to increase the temperature of 80 g of clay from 100 °C to 1000 °C. 80 g of clay from 100 to 1000 °C: m. Cs. ∆T = 80. 0.2. 900 = 14,400 calories Cs = specific heat (1 for H2O, 0.2 for clay) L = latent evaporation heat The above example illustrates that in a production process such as extrusion, drying, although generally considered to be energetically modest, actually absorbs almost as much heat as the firing process. Drying can essentially be seen as a process conditioned by three factors: 234

Drying rate


time

Drying rate

Fig. 115. Variation of drying speed against time (according to Y.H. Perry, from A.G. Verduch).

Moisture Fig. 116. Variation of drying speed against moisture content (according to Y.H. Perry, from A.G. Verduch).

the drying power of the work area, relative humidity, air speed forces acting at capillary level shrinkage caused by loss of moisture. The drying process, then, involves transfer of heat from the surrounding environment to the ceramic body and the simultaneous movement of water vapour in the opposite direction. The heat (energy) needed for the former may reach the piece via convection, radiation or conduction: normally, all three are involved. 235


The most common effect of drying (by hot air convection) can be described by the sequence: transfer of heat from air to piece transformation of water from liquid to vapour removal of vapour from surface of piece transport of liquid water from piece interior to the surface. When surface evaporation speed exceeds internal capillary transfer capacity another drying phase begins during which the evaporation front moves inside the piece itself. All this is usually represented by the classic description of a convection-type drying process for porous and hygroscopic materials, according to a curve split into at least three phases that are governed by different drying speed gradients at different depths. During the first drying phase water evaporation occurs as per the laws that govern evaporation from a body of water such as a lake, where a constant surface area is in contact with an air flow of constant temperature and relative humidity: drying speed is thus constant too. During the second stage, drying speed decreases rapidly as the volume of water evaporating from the surface has to be drawn from the interior of the piece by capillary action: the rapidity with which this takes place will depend on the extent to which the piece is heated and its dimensional variations (which influence the vapour pressure of the fluid and thus its transfer speed). As the evaporation front moves deeper and deeper into the piece, drying speed falls as capillary resistance to diffusion increases: this resistance, of course, is inversely proportional to the size of the capillary ducts themselves. The smaller the average diameter of the capillary the higher the resistance (according to the function x = k √ rt, where x is the speed of diffusion of the liquid, r is the capillary radius, t is time and k is a constant that takes into account the surface tension and viscosity of the liquid according to Poiseuilles law). As a rough guide, vapour pressure values for water as a function of capillary diameter are given in the following table: Radius (mm)

Vapour pressure 0.1 0.2 0.4 0.8 0.9 0.95 0.98 0.99

5 10-5 7 10-5 12 10-5 48 10-5 105 10-5 209 10-5 558 10-5 1070 10-5 ≅ 10mm

236


According to Norton, the diffusion of water in the interior of the ceramic piece is, then, a function of various parameters dV k (U2 - U1) p ___ = ___________ dV/dt dt dη U1-U2 p d η

= flow rate of water through the piece = difference in moisture content between core and surface = permeability of the ceramic body = distance from surface of interior point U2 = water viscosity

in which the most representative variable of the choice of raw materials, technological process and characteristics of the ceramic piece is undoubtedly p (permeability), greatly affected by particle size, plasticity and compaction etc. Clays and ceramic bodies subject to drying, then, undergo consistent modifications in terms of shrinkage and weight loss caused by the evaporation of water. Barellatographic analysis, employed to see just how these variations progress, yields Bigot curves which illustrate the behaviour of the material in terms of shrinkage/ weight loss at a constant T, or other curves derived from these. The drying process revealed by the barellatograph corresponds to theoretical models put forward by Bourry (see Figs. 117 and 118). These diagrams represent the percentages of body (by volume) occupied by clay, water and pores as a function of drying time. The ABC curve illustrates shrinkage: at time T1, for example, a body with an initial volume of 100 consists of up of 56% clay + 15% water + 8% pores, and has undergone a shrinkage of 21%. Such curves, initially developed for extrusion technologies employing very slow drying cycles (at ambient temperature and relative humidity), can also be traced making the appropriate modifications for products of low water content subject to extremely fast, high-temperature drying cycles. As seen, one of the most significant factors affecting evaporation during the drying of ceramic materials is the surface area of the material to be dried, which influences the permeability of the material itself. Evaporation speed is also influenced by hygrometric parameters in the drying area: a flow of dryer or completely dry air is particularly effective in increasing evaporation during the first stage of drying, while its influence will decrease at the capillary transfer stage as surface evaporation comes to an end. Once the critical humidity point is reached (point of inflection on Bigot curve), the ceramic body assumes an almost-definitive consistency and rigidity and adapts to tension less easily. As shrinkage in the first stages of drying is much more marked than in subsequent stages, the resulting tension may lead to differential shrinkage, deformation 237


Kaolin

Montmorillonite

Shrinkage

Pores

Water

Volume (%)

Volume (%)

Shrinkage

Water

Pores

Clay Clay

Hours

Hours

Fig. 117. Drying or Bourry diagram for a kaolin and a montmorillonite.

Shrinkage

Water Volume (%)

Pores

Clay

Hours

Fig. 118. Bourry drying diagram.

238


or crazing when an excessive humidity gradient exists between the interior and exterior of the ceramic body. The distribution of water around the particles and the way in which the latter gather closer together as drying progresses are illustrated in the diagrams in Fig. 119. The differing orientation of the clayey particles and the presence of differently sized particles of other hard materials has, then, an enormous influence on drying. For example, clayey structures with a playing card castle configuration dry faster than those arranged as a deck of cards. This is because the latter are more compact and oppose greater resistance to evaporation.

Figure A Before the start of drying the individual particles are separated by a thin film of water (grey area).

Figure B As soon as the surface water has evaporated and water starts being drawn from the interior by capillarity, the particles begin moving and may start coming into contact with each other.

Figure C Even after the second drying stage significant quantities of water still remain between the particles, in the pores.

Figure D In the final stages of drying temperatures must exceed boiling point to remove any residue. At this point there is a partial collapse of the particle structure.

Fig. 119. Structural rearrangements during following drying.

239


Illustrating, then, what happens during the removal of water in terms of its interaction with solid particles as opposed to its position in the piece, it can be said that drying removes the water in pores or interstices, water adsorbed on the surface of the mineral (especially clayey) particles or water introduced into their structural interlayer. None of these types of water form part of the crystalline structure of the clayey minerals. The drying (and plastic) properties of clays are mainly determined by the size and shape of particles, by composition and structure, ionic exchange capacity, exchangeable types of ion and, finally, by their current state of hydration. To these intrinsic qualities must be added extrinsic ones such as spatial relations between the particles, pore size and interstice size distribution, degree of compaction etc. The water surrendered at modest temperatures that which is neither part of the crystalline structure nor attributable to the OH- hydroxyls in the composition (the so-called chemical water expelled between 500 and 800 °C) can be divided into three categories: water in the pores, on the surface and at the edges of the particles interlayer water, between elementary clayey sheets (e.g. in smectites) water in structural channels (e.g. in zeolites). Eliminating water of the first category requires low yet constant energy as seen in Fig. 115 and thus relatively low temperatures (even ambient temperatures will do where relative humidity is low). Temperatures in excess of 100 °C are used to accelerate the process and increase the thermal gradient between the interior and exterior of the piece. On the other hand, waters of the second and third category are eliminated with lesser or greater difficulty as a function of various other parameters closely tied to the texture and compaction of the materials. Interactions between clay and water depend, in fact, on the physical and chemical reactivity of the surface of the clayey mineral, where oxygen, hydroxyls and in the fracture zones other elements are present; the distribution of these active sites determines the surface activity of the clayey minerals (exchange capacity, ionic selectivity...); specific surface charge, surface area and pH also play an important role. Clayey minerals are not perfect crystals. On the one hand, they always have fractured edges and consequent bond breaks that remain unsaturated. On the other hand isomorphic atomic substitutions bring cations of different oxidation states into play; these alter the natural electro-neutral structure of the crystal. These factors, then, cause the solid particles to behave as if they were large, weakly (mostly negatively) charged insoluble ions, as isomorph substitutions of insufficient charge are more common. These charged particles are, then, distributed in contact with and within the water, which is, of course, impure (i.e. exempt from ionic charges due to saline solubilisations). 240


All this contributes to further alteration of the structural arrangement of the material and complication of water bonding phenomena. Consequently, the drying release mechanism becomes more complex. Thus far, we have examined concepts that regard drying by convection. Drying processes that use only conduction are few and have little relevance here owing to the poor thermal conductivity of ceramic materials. Yet the drying possibilities offered by radiation are, instead, certainly worth a brief look as they offer the option of a more selective kind of drying that acts almost only on the water actually contained in the piece. These systems can accelerate the drying process significantly and are extremely energy-efficient, using electromagnetic radiation at the infrared (IR) and microwave (MW) wavelengths (see Fig. 120). In both cases, since the majority of water in the piece is contained in its innermost layers and ceramic materials are virtually radiation-transparent, heating takes place from within. This is because it is the water which first transforms the radiation energy into heat energy. Of course, this also aids evacuation of vapour because, unlike convection heating, surface porosity is unaffected by structural shrinkages (except, perhaps, at the end of drying). Infrared driers exploit the fact that water has an almost unitary absorbance factor (0.92) for radiation of wavelength λ = 2.8 µm, and a still-excellent absorbance factor (0.91) at λ = 5.95 µm (see Fig. 121). At these wavelengths air is completely transparent and its presence becomes irrelevant: yet in the context of drying, the fact that it does not heat up is a negative one. The water vapour generated by selective heating inside the piece is a negative factor too, as it absorbs part of the incident radiation. The ceramic body, depending on its composition, colour and degree of compaction, will absorb a varying amount of the incident radiation and generally tends to heat up. These systems, which reached a peak of popularity in the 80s, have proved highly successful, reducing drying times from the once standard 40-45 minutes to 12-15 minutes. The main problem with such radiation systems is that power levels limit penetration capacity, thus making it difficult to use them with particularly thick products where the bulk of the energy is absorbed by the surface of the piece. However, no penetration problems exist where hyper-frequency 2450 MHz (122 mm) microwave radiation is used: this acts selectively on the water molecules thanks to the selective absorption caused by the high dielectric constant of water, which has a decidedly dipolar molecule. The main obstacles to its use, especially in Italy, lie in the high cost of electricity 241


Fig. 120. The electromagnetic spectrum.

(such hyper-frequency radiation cannot be generated from any other source) and the difficulty of controlling the drying process effectively, as it tends to be too rapid, thus causing the pieces to break. Moreover, the working life of microwave generators (magnetrons) is relatively limited and their maintenance is somewhat costly. A description of the most commonly used drying systems employed by the ceramic industry is given in the relevant section of Volume II of this work, which 242


Fig. 121. IR absorption of water as a function of wavelength.

focuses more on the ceramic products themselves. That section will also examine the most important technological parameters for proper control of the drying process, namely: temperature distribution contact surface area airflow vapour extraction.

243


244

Description of ceramic products

Chapter VIII DESCRIPTION OF CERAMIC PRODUCTS

Although we have now clarified the main characteristics of the individual raw materials and stated how they are selected, combined and prepared, we have yet to provide information on the most suitable combinations and mixtures for the production of ceramic tiles, which cover a wide range of technological characteristics. In fact, depending on the context in which they are used, tiles must have good-toexcellent load, abrasion and stain resistance (floor tiles) or, where they are to form regular geometric patterns, need to be sized extremely accurately (wall tiles). Tiles may sometimes be laid outdoors where they are subject to freeze/thaw cycles or extremes of heat and humidity where different absorption and long term reactivity to water or humidity are required. These characteristics are listed in the International Standards contained in the Appendices at the end of this volume. Two of the most important final properties required of a tile, whether glazed or unglazed, are bending strength and good dimensional characteristics (size and flatness), as illustrated in Fig. 122. Yet these properties, with limits and tolerances that depend on the intended type of product, are not in themselves sufficient to define the formulation of a good body. There are, in fact, plenty of other conditions that need to be satisfied, such as: a plasticity sufficient for moulding and unfired handling purposes, the ability to dry

PRIME GOALS

OTHER NECESSARY GOALS

Composition Grinding Pressing Firing

Fig. 122. Factors determining ceramic tile body quality.

245

Limited moisture expansion

Good glaze match

FINAL TILE PROPERTIES

Absence of black core

Rapid firing cycles

Ease of drying

DURING MANUFACTURING PROCESS

Unfired bending strength

Sizing and flatness

Bending strength

FINAL TILE PROPERTIES


the material properly (and perhaps very rapidly), suitability for the firing process etc. Should manufacturers fail to meet these conditions the final product may suffer from defects such as lamination, cracking, incompatibility with the glazed surface, black core, post-expansion and so on. Going back for a moment to bending strength, it should be pointed out that products must meet minimum strengths as defined by relevant standards (Fig. 123); to this end the ISO standard requires that not only Modulus of Rupture (MOR) but also Breaking Strength be measured so as to ensure product bending strength independent of thickness and size. While, in fact, MOR is an intensive measurement that characterises the quality of one material with respect to another, effective breaking strength indicates the working behaviour of the specific piece. To illustrate: a thin vitrified product has in any case a very high modulus, yet the effective breaking load might actually be less than that of a standard-thickness wall tile. Another example that highlights the difference is the behaviour of unfired tiles of identical thickness but differing sizes (e.g. 20 × 20 cm and 50 × 50 cm tiles): their modulus will, of course, be the same, but the effective breaking load (and thus resistance to stress on the production line) in the bending strength test will be much lower for the larger tile.

MOR (kg/cm2)

This explains why finished product bending strength requisites set by International Standards establish minimum Modulus of Rupture from 15 N/mm2 (wall tiles - class B III) up to 35 N/mm2 (stoneware - class B Ia), while minimum breaking strength values range from 60 to 130 Kg.

Fig. 123. Bending strength requisites according to ISO standards (1 kg per cm2 ~ 0.1 N/mm2).

246


Meeting these requisites is fairly easy, in that the same body can be used for more than one product class; yet things become a little trickier when all the key properties of the tile are taken into consideration.

% WA

Above all, body behaviour during vitrification needs to be examined (Fig. 124): vitrification curves (shrinkage and water absorption as a function of maximum firing temperature) for different product types [red (R) and white (B) body wall tile, red (R) and white (B) body floor tile, porcelain tile (P)] show just how great the differences can be.

wall red wall white floor red floor white porcelain tile

Fig. 124. Vitrification curves.

Wall tile bodies have high porosity within a fairly wide range at low-temperature and then vitrify too rapidly to be used as flooring. A body intended for total vitrification (water absorption = 0), such as stoneware, has, unlike standard glazed floor tile bodies, wider stability at low porosities. Such behaviour needs to be understood not only for production process optimisation but can also be explained, forecast and modified as a function of the chemical-physical characteristics of raw materials and body formulations. As with porosity, shrinkage curves highlight an enormous diversity of behaviour between the various bodies (Fig. 125). The graph shows that white bodies are generally more stable than red, which are more clayey and have a less balanced composition, in both vitreous and porous compositions. As far as porous products (typical of wall tiling) are concerned, note the limited shrinkage (< 1%) at low firing temperatures: this was once deemed necessary to ensure maximum size uniformity. Bending strength (Fig. 126) is, then, for the most part linked to processes that take place during sintering. 247

Shrinkage %



Bending strength

Fig. 125. Shrinkage curves.


Fig. 126. Bending strength curves.

Consequently higher strengths are seen for the red bodies. which tend to be more vitreous, except for porcelain tiles, the composition of which is specifically intended to develop the glassy phases. Interpreting the behaviour of ceramic bodies So far only those macroscopic factors that are usually monitored and controlled during the production process have been considered. It is, though, possible to make a more in-depth analysis of the structural modifications that take place in the mate248


rials (especially during firing), thus allowing us to observe the root causes of the phenomena illustrated above. This more detailed examination highlights even sharper differences between the compositional and behavioural differences of porous and vitrified bodies. Starting with the former (Fig. 127), it can be seen how wall tile bodies are characterised by an abundance of calcium and/or magnesium in the form of carbonates that decompose at temperatures above 800 °C, freeing the corresponding oxides. These oxides subsequently react with the rest of the ceramic matrix to form (at relatively low temperatures) new crystalline compounds, Ca and Mg silica-aluminates that provide a certain firing stability and develop bending strength before giving rise to an uncontrollable fusion. The numerous crystallines that can be created by a CaO - Al2 O3 - SiO2 system are shown in the tertiary diagram in Fig. 128. Note the formation of gehlenite, pseudo-wollastonite and anorthite, which, together, can lead to marked lowering of melting points, thus fuelling further reactions via the formation of liquid phases. The following table (Fig. 129) shows some of the properties of the crystalline compounds already in both the original compositions and the neo-formed composi-

% WA Shrinkage MOR

Decomposition of carbonates

Degassing through glaze Tendency towards post-expansion Increased bending strength Dimensional stability Increased shrinkage Deformation

Presence of free CaO

Formation of crystalline phases: Gehlenite Diopside Wollastonite Anorthite

Uncontrolled Uncontrolled formation formation of of glassy glass phases phase

Fig. 127. Wall tiles: the symbols correspond to A.A. (water absorption), MOR (bending strength) and shrinkage.

249


Fig. 128. Tertiary diagram for calcium silica-aluminates. (10-7/°C)

Compound

T

(°C) fusion

Quartz Albite Orthoclase Anorthite Gehlenite p-Wollastonite Diopside Forsterite Spinel Mullite

Fig. 129. Expansion coefficients and melting points for crystalline compounds existing or formed in the bodies.

250


tions (i.e. those formed during firing); these take part in the firing process and influence the final structural properties of the products. Note the generally high expansion coefficients of the new compounds containing calcium and magnesium which serve to inhibit shrinkage. Theoretical melting points for pure compounds are, however, very high and their formation from solid-solid or solid-liquid reactions is thus difficult: therefore the percentages detected in fired products (especially where rapid firing cycles are used) are generally low. This is particularly true for mullite, which is unlikely to form even at very high maximum firing temperatures. XRD analysis of fired wall tile powders (Fig. 130) still show, however, significant traces of crystallinity of starting minerals, especially quartz, and partial formation of neo-phases (e.g. Diopside). Where a composition has an incorrect balance of components of poor particle size distribution the firing cycle may be too fast to allow stable crystalline phases to develop. Though this structure may be frozen in a meta stable phase it will still survive for the lifetime of the product. If, for example, we consider the formation mechanisms of a vitrified product, based on the fluxing action of sodium and potassium compounds, it can be seen that, compared to wall tiles based on calcium-magnesium systems, reaction temperatures shift to a higher level: the low-temperature mechanical characteristics of vitrified products are therefore generally poorer than porous ones.

1

#

Fig. 130. X-ray diffraction analysis of wall tile body (Q= quartz, C= neo-formed comp.).

251


Moreover, intermediate neo-formed compounds that start developing in an asyet insufficiently bonded matrix result in under-fired vitreous bodies that tend to have higher moisture expansion than porous wall tiles. Consequently they are unsuitable for use as wall tiles, even where there are no particular final product shrinkage and thus size requirements. The expansion curve for an unfired vitreous body (Fig. 131) highlights, in addition to the α - β quartz transition, a relatively gradual vitrification process: optimum firing temperature is clearly identifiable via the derivative of the expansion/ shrinkage curve which peaks where the shrinkage is fastest. At this temperature vitrification takes place extremely quickly: use of higher temperatures not only wastes energy but can also give rise to less dense structures owing to the onset of expansion (Fig. 132). The Na2O - K2O - Al2O3 - SiO2 tertiary diagram (Fig. 133) for sodium silicates, potassium silicates and silica, shows how, in certain compositional fields, the development of a liquid phase at decidedly low temperatures is possible (1118 °C for sodium feldspar, 1150 °C for potassium feldspar). The tertiary diagrams refer, of course, to systems in thermodynamic equilibrium and do not provide any information on reaction speeds and necessary vitrification time which, especially in fast firing cycles, must be brief.

Fig. 131. Expansion curve for a wall tile body.

252


Absorption Shrinkage MOR

Phase transformations: Metakaolin-spinel Partial crystallite destruction (albite, orthoclase, quartz) Formation of glassy phase

Increased bending strength Increased frost resistance, Reduced permeability, shrinkage Deformation Increased bending strength

Formation of new crystallites (mullite)

Fig. 132. Totally vitrifiable bodies (for symbols Absorption and MOR see fig. 127).

Fig. 133. Tertiary diagram for sodium-potassium feldspars.

253


However, the reaction speeds of solid state compounds depends greatly on the available surface area and, therefore, particle-size distribution. Completion of a crystalline solid particle reaction is unlikely inside a ceramic body. Yet it is possible for reactions, mostly of the cementing type, to take place between the individual grains aided by imperfections in the crystalline lattices and impurities, which, even in small quantities can act as catalysts or give rise to low-fluxing eutectics (the same quantity of material, at a density of, for example, 2.5 g/cm3, has a reaction area of 24 cm2/g where its average particle diameter is 1 mm; if this size is reduced to 50 µm, the reaction area increases to 480 cm2/g). Taking into consideration all the above factors and arguments and, most importantly, production needs, plant limitations, market demand, the availability of suitable raw materials and their cost, producers need to decide which type of ceramic tile to manufacture. There are, in essence, three main product classes:

WALL TILES

FLOOR TILES

PORCELAIN TILES

ACCESSORIES AND TRIMS

254


Chapter IX WALL TILES

Introduction Tiles were one of the earliest man-made technological materials. From the very start they provided hygiene and fulfilled artistic and decorative functions as well. One of the oldest examples of glazed tiling is to be found on the Ishtar gate of the Babylonian king Nebuchadnezzar II (575 B.C.), a restored version of which is now on show in the Pergamon museum of Berlin. Then there are the palaces of Nimroud and Khosabab, those of Cyrus, Darius, Xerxes, several monuments in India, China and Turkey. The ceramic tile, then, has been with us throughout nearly all human history. In Italy, tiles were first manufactured during the Middle Ages, absorbing Arab and other influences. From these tentative beginnings came modern ceramics with its well-defined, standardised products such as majolica, cottoforte, earthenware, gres, etc. In Italy, the ceramics industry took off in the post-war reconstruction period of the 1950s, with the manufacture of red gres and majolica and the advent of cottoforte, a classic Italian product made using double firing techniques. It was also during the 50s that ceramics was transformed from a network of small-scale workshops into a true industry. The technical dynamism of the industry proved decisive in promoting continuous evolution of processes and materials. The result was a shift from traditional double firing to fast double/single firing. Tiles also became larger and their aesthetic/decorative effects were constantly upgraded. Most of these developments have taken place over the last twenty years, stemming from experience in single fired floor tile products. During the 70-80s the fast firing process, as far as wall tiles were concerned, focussed almost entirely on glazed tiles and was initially a simplification of traditional tunnel firing systems using refractory saggers on cars. Later, during the energy crisis at the beginning of the 80s the single firing process was developed and adopted for porous body tiles too. Single and double fast firing technology quickly became commonplace all over the globe. Its spread has been (and is) dependent on a number of factors: local levels of technical and professional skill, energy costs, labour costs and cultural traits.

255


The market Figure 134 compares the evolution of the Italian market for double firing and monoporosa wall tiles from 1990 to 1998. Note the slight fall off in double firing output in 1993, followed by a more positive trend which lasted until 1998, stabilising at about 85 million m2 in the period 1996-97-98. Monoporosa, however, showed a constant upward trend from 1991 onwards, reaching a peak of 48 million m2 in 1995. The graph in Figure 135 shows percentages of output accounted for by single and double firing respectively against the total quantity of wall tiles. Product classification General wall tile classification includes products obtained by both double and single firing. International ISO UNI EN 13006 classifies ceramic tiles into several groups on the basis of their forming method and finished product water absorption (see Fig. 137). On the basis of current ISO 13006 standards, porous wall tiles are included in the group B III Absorption > 10%

Fig. 134. Evolution of Italian monoporosa and double firing output from 1991 to 1998 (millions of square metres source: Ceramic World Review n. 32/99).

256

Wall tiles

Fig. 135. Percentage of total wall tile output accounted for by single and double firing products (source: Ceramic World Review n. 32/99). Note: the term double firing includes all materials obtained with two fires whether fast or traditional).

Fig. 136. Shows output increases in percentage terms for both double firing and monoporosa with respect to 1991.

Water absorption

Shaping

extrusion

dry pressing

GROUP I ≤ 3%

Old EN

GROUP IIA 3% - 6%

Old EN

GROUP IIB 6% - 10%

Old EN

GROUP III > 10%

Old EN

GROUP A1

EN 121

GROUP AIIa1

EN 186/1

GROUP AIIb1

EN 187/1

GROUP AIII

EN 188

GROUP AIIa2

EN 186/2

GROUP AIIb2

EN 187/2

GROUP BIIa

EN 177

GROUP BIIb

EN 178

GROUP BIII

EN 159

GROUP BIa ≤ 0.5%

EN 176

GROUP Bib BIb 0.5% - 3%

Fig. 137. Classification of tiles according to ISO 13006.

257


Technical features Wall tiles generally have the following characteristics: high dimensional stability during firing, with almost no shrinkage (less than 1%) porosity between 13% and 18% (expressed as percentage of water absorbed) MOR between 200 and 250 kg/cm2. These properties are only indicative, mainly helping us to classify the product from a commercial viewpoint that takes into account its field of use. The specifications provided for under ISO 10545.1 - 17, group BIII, include the dimensional, physical and ceramic properties of the products. Fig. 138 charts these last two, without taking size into account. For more realistic comparison purposes the table shows, in addition to the values defined by the standards, some characteristics of products actually available on the market. Any description of wall tiles must necessarily include the types and characteristics of the glazes used in the new fast firing technologies. It should also be observed that, as far as surface aesthetics (i.e. glazes) are concerned, the use of such technologies has neither modified nor materially improved the final results, since aesthetic quality was already very high. Transparent (crystalline) glasses and opaque glazes (mainly white), both of which are glossy, remain the most popular glaze bases to this day. While glazes have evolved to suit different firing technologies (traditional, double fast, monoporosa), their technical performance in terms of resistance to staining, chemical agents and abrasion have remained largely unchanged. In the case of white and crystalline glazes, the limitations and the potential provided by the new technologies have required new frit and glaze formulations, with a shift from viscous low-fluxing materials to high-fluxing compounds with eutectic melting. This evolution in composition has affected both double fast firing and monoporosa glazes. Aesthetic features Size During the 60-70s, traditional firing processes mainly offered 15 × 15 and 20 × 20 cm tiles. With the introduction of fast double firing and porous single firing (monoporosa), tile size increased to 25 × 33, 33 × 45 and in certain cases 40 × 60 cm and beyond. The most common sizes, however, are those in the middle of the range: 20 × 20, 20 × 25, 25 × 33 cm etc. Large tile production is much more limited but now increasing rapidly, such tiles no longer being seen as just niche products. Such large tiles generally require sin258

Wall tiles

FEATURES

STANDARD

MAXIMUM STANDARDPRESCRIBED VALUE

REAL VALUE OF MARKETED PRODUCTS

WATER ASSORBATION

ISO 10545.3

>10 ,< 20% - min. 9%

13 - 18%

MODULUS OF RUPTURE (MOR)*

ISO 10545.4

12 - 15 N/mm2

> 20 N/mm2

BREAKING STRENGTH*

ISO 10545.4

min 600 - min 200 N

min 700 - min 300 N

ABRASION RESISTANCE**

ISO 10545.6/.7

SPECIFIED BY THE MANUFACTURER


EXPANSION COEFFICIENT

ISO 10545.8

TEST AVAILABLE

6,5 - 7,5 . 10 -6

THERMAL SHOCK RESISTANCE

ISO 10545.9

TEST AVAILABLE

CONFORME

MOISTURE EXPANSION

ISO 10545.10

TEST AVAILABLE

< 0.06 mm/mt

CRAZING RESISTANCE

ISO 10545.11

REQUIRED

MEETS SPECIFICATIONS

FROST RESISTANCE

ISO 10545.12

TEST AVAILABLE


RESISTANCE TO HOUSEHOLD CHEMICALS

ISO 10545.13

CLASS GB min

CLASS GB min

RESISTANCE TO ACIDS AND ALKALIS

ISO 10545.13

TEST AVAILABLE


RESISTANCE TO STAINING

ISO 10545.14

CLASS 3 min

CLASS 3 min

Pb AND Cd LEACHING

ISO 10545.15

TEST AVAILABLE


COLOUR DIFFERENCE

ISO 10545.16

TEST AVAILABLE


FRICTION COEFFICIENT**

ISO 10545.17

TEST AVAILABLE


Standards 10545.1 and 10545.2 are not shown in the table owing to limited space. Nevertheless, the reader is reminded that they contain sample acceptance methods and size and surface quality characteristics. * Depending on thickness (< 7.5 mm ≥ greater than 7.5 mm) ** To be effected only where used as paving. For more detailed information see UNI publications (Italy).

Fig. 138. Technical characteristics of porous wall tiles according to ISO standards.

gle-layer kilns and the advanced automation/ handling systems now available on new fast-firing plants. Trims Special pieces, considered complementary to standard tile production, have gained much ground in recent years. These pieces feature relief work obtained by pressing (inserts, festoons, bull-nose pieces etc.). Production of such items is generally outsourced to other companies and they are made using a glazing process followed by decoration and third firing. Moreover, there is also an increasing demand, especially in American and Asian countries, for trims acting as a technical complement: these are used to finish outside or inside corners or frame the perimeter of a tile-covered surface. Generally speaking, these pieces are obtained using presses of limited power and special dies. They are normally double-fired. In most cases, biscuit pieces are covered with standard glazes.

259


Surface appearance Traditionally, wall tiles have always been glossy. Consequently, glazes are generally transparent (crystalline) or opaque (basically white), rarely matt. In the 60-70s, in traditional firing, frits with a high content of lead, boron and cerium were used to obtain such decorative effects as torn, reactive, mother-ofpearl, etc. From a ceramic viewpoint, the effects were fascinating, yet their resistance to abrasion and chemical aggression was very limited. The adoption of new formulations and the partial replacement of lead with other fluxes has now provided manufacturers with crystallines and glazes on a par with their predecessors but better suited to the new technologies. Recently, the influence of developments in architecture and interior design has given rise to new effects known as antiqued, which mimic aged plaster, chalky surfaces, etc. and are sometimes highlighted by special elements built into the tile, such as structured surfaces or intentional chipping around the edge. The bulk of production, though, still uses glossy-effect glazes, enriched by decors that take their cue from natural stones such as marble and breccia etc. Raw materials for bodies General features In recent years wall tile compositions have evolved considerably, mainly to adapt to fast firing cycles where body and glaze are fired together. With monoporosa there may be interference between the degassing of certain raw materials in the biscuit and the molten glass: this can lead to surface defects on the glaze. The key factors in the evolution of body compositions from those suited to traditional firing, then fast double firing and finally porous single firing have been: reduction of clayey material percentages introduction of higher percentages of fillers and complementary raw materials (feldspars, feldspathic sands, quartz) limitation of minerals which give off gaseous phases at high firing temperatures (calcite and/or dolomite), especially in monoporosa. In pursuing these goals it has been possible to employ tried and tested materials by changing the quantitative composition. In the case of monoporosa, it is preferable to use raw materials with a high degree of purity and a fine particle size distribution. Generally speaking, finished wall cladding products must have a high degree of size stability.

260

Wall tiles

The formation of crystalline compounds, such as wollastonite, gehlenite, anorthite and diopside, not only guarantees that stability: they also condition and stabilise other features, such as post-expansion and expansion coefficient, etc. Characteristics of raw materials for bodies According to commercial classification, wall tiles can be divided into red and white. In both cases, the raw materials are of two basic types: clayey materials complementary materials (feldspars, feldspathic sands, quartzes, calcites). Figure 139 shows the chemical-physical features of the raw materials most commonly used in bodies for both fast double firing and monoporosa. Some notes about the overall nature and mineralogical features of these materials follow. Plastic raw materials Marl-carbonitic clays: mineralogical associations may be of the illite-chlorite or sometimes illite-kaolinitic type (the latter in lesser quantities). The quantity of calcite in the clayey matrix varies and is sometimes considerable. These clays give the body plasticity. Generally speaking, after firing they take on a beige-orange colour due to the presence of ferrous minerals. Vitrifiable plastic red firing clays: almost entirely carbonate-free. Clayey minerals may be associated with illite-chlorite types. Their function is to confer plasticity on the system, making it possible to achieve good breaking load values on green, dry and fired pieces too, since they are generally vitrifying materials. Ball-clays: generally white firing. The clayey matrix is of a kaolin type, with little illite. Particle size distribution is generally fine, so good plastic behaviour can be expected. Post-firing bending strength and porosity are extremely good. Kaolin-type china-clays: behaviour is generally refractory; the limited plasticity is basically due to the intrinsic particle size distribution of the material. Carbonitic and/or vitrifiable clays are evidently used to achieve a beige-orange biscuit, whereas ball-clays and china-clays are used for white biscuits.

261


1

3 4

1

2

3

4

2

3

1 2

3

1 2

1

3

Key: A - semi-plastic clays, B plastic clays, C plastic clays, D kaolinitic clays, E feldspathic sands, F quartz, G calcium carbonate.

2 3

PHYSICAL FEATURES

4 3

1

2 4

1 2 12 3

A

B

C

D

6/8

8/10

10/15

4/6

BREAKING LOAD BEFORE DRYING

kg/cm2

BREAKING LOAD AFTER DRYING

kg/cm2 15/20

FIRING AT 1100 °C

20/30 25/40

10/15

A

B

C

D

POROSITY %

8/12

2/6

5/10

15/20

SHRINKAGE %

2/3

4/6

3/5

2/4

180/200

200/300

200/300

80/120

BREAKING LOAD AFTER FIRING kg/cm2

Fig. 139. Chemical and physical characteristics of different raw materials suitable for the production of porous products, especially monoporosa for indoor tiles.

262

Wall tiles

Complementary raw materials Feldspathic sands, feldspars and quartz: introduced into the composition as fillers and/or sandy non-plastic materials to facilitate effusion of the volatile compounds which escape during firing. Potassium feldspars are preferred as they are less reactive than the sodium kind. Feldspars also help lower the overall expansion coefficient of the ceramic body. Quartz also plays a key role in adjusting the expansion coefficient (the latter increases proportionally to the quantity present). Given the rather coarse particle size distribution and the fast firing cycle, reactivity of free quartz against alkaline-earth oxides (CaO and MgO) is thought to be limited. Calcite and/or dolomite: fundamental materials for wall tile bodies. In-body percentages range from 8% to 15%. Their natural and post-grinding particle size distribution is particularly important. In fact, very fine particle size distributions favour both decarbonation reactions and, at a later stage, synthesis reactions with residuals of the clayey materials (especially amorphous silica), thus allowing formation of transparent neo-formed compounds (over 900 °C). Of special importance are the kinetics of decarbonation and thus the kinetics of gas (CO2) emission before softening of the surface glass (glaze) takes place. The evolution and completion of synthesis reactions between silica, alkalineearth oxides and alumina play a key role in defining the physical and mechanical features of the ceramic piece after firing (bending strength, expansion coefficient, etc.). It is evident that the quantitative ratio of clayey materials, calcite, feldspars, quartz etc. depend on the intrinsic mineralogical nature and particle size distribution of the clays. As a rough guide, Fig. 140 illustrates both a monoporosa (red and white fired colour) and a fast double firing composition. Body composition As stated above, wall tiles can be obtained by double or single firing, and always using fast cycles. Furthermore, body composition may aim to produce white or red firing products. The latter mainly consist of varyingly carbonatic clays with a high iron content. Other components may include feldspathic sands, feldspars, quartzites and if necessary calcites and/or dolomites. White firing bodies, on the other hand, use more balanced mixes of light coloured clays, calcite, feldspathic sands and quartz. The most significant difference between white and red compositions lies in the quantity and typology of clay used, whereas the most significant percentage difference between double and single-fired bodies is that the former may have high per-

263


Fig. 140. Possible compositions for porous single firing (red and white) and double firing bodies.

centages of calcite and/or dolomite, even as high as 15-18% and, in certain cases, raw materials with a slightly lower degree of purity may be used. Figure 141 shows examples of different body compositions for porous single firing and fast double firing, red and white firing respectively. In addition to compound formulations (i.e. formulations obtained by mixing raw materials), the chart also shows a clay-only body formulation. This kind of composition is used more frequently in the case of fast red double firing and is quite widespread in zones which have access to mixed clayey materials which already include the various complementary minerals. Figure 141 also shows the (CaO-MgO) - SiO2 - Al2O3 tertiary diagram for the formulations in the previous figure. Three specific body compositions are highlighted; two of them are 264

Wall tiles

TIPOLOGIA - TYPOLOGY

TiO2 Fe2O3

CaO

MgO

K2O Na2O

Monoporosa

65.38 12.89

SiO2 Al2O3

0.29

0.53

7.00

1.26

0.72

0.19 11.69

P.F.

Monoporosa

61.42 14.08

1.18

1.22

7.40

0.99

2.06

0.17 10.96

Monoporosa rossa - Red monoporosa

56.10 15.12

0.91

3.79

7.80

1.40

1.25

0.90 12.89


54.76 15.98

0.90

4.88

8.10

0.77

2.37

0.99 11.31


65.99

18.9

0.70

3.04

0.80

0.56

1.72

4.67

Monoporosa

60.26 16.34

0.76

1.16

7.60

0.42

2.23

0.59 10.78

Monoporosa

69.36 12.30

0.50

0.59

5.70

0.29

3.47

0.23

Bicottura - double fired

60.31 13.40

0.57

1.95

9.35

0.31

1.17

0.25 12.82


61.29 11.45

0.68

2.85

5.92

2.61

2.39

2.03 10.76


59.37 11.58

0.63

4.56

5.78

3.13

2.49

2.04 10.44

3.62 7.37


63.12 12.68

0.52

4.51

6.06

0.79

2.71

0.42

8.51


65.80 13.77

0.89

1.56

5.82

0.6

1.41

1.86

8.29

CaO+MgO

Bicottura - Fast firing Fast firing Monoporosa Monoporosa Monoporosa rossa - Red monoporosa Red monoporosa

SiO2

Al2O3 CD.0110

Fig. 141. Composition areas in the (CaO+MgO) - SiO2 - Al2O3 tertiary system relevant to the abovedescribed body composition for monoporosa and fast double firing.

for white and red monoporosa, whereas the third is a special combination obtained only with double firing clayey material. Product features The final features of a ceramic wall tile are heavily influenced by newly formed compounds which develop during firing. These are created by the reactivity of calcium and magnesium oxides which originate, respectively, from the destruction of the calcite and/or dolomite lattices during firing, with the silica and alumina originating from the destruction of the 265


clayey lattice. These newly formed compounds are gehlenite, diopside, anorthite and wollastonite. Features such as bending strength, expansion coefficient, moisture expansion etc. depend on the quantity and quality of these compounds. In fast firing cycles the tendency towards synthesis of new compounds, starting with phases of varying reactivity, largely depends on the natural particle size distribution of the materials and the degree of grinding that follows. The pattern of decomposition and synthesis reactions which take place during fast firing of a monoporosa body is shown in Figure 142: this diagram shows the dynamics of original mineral decomposition and the formation of new compounds. Figure 143 shows X-ray diffraction patterns for different firing temperatures. Note that at 900 °C dolomite has almost entirely disappeared, yet the first crystallisation embryos attributable to diopside and gehlenite have appeared. The latter tend to increase up to a temperature of 1140 °C. Raw materials for glazes The most frequently used glazes are glossy, either transparent (crystalline) or opaque (basically white). These glazes use different frits depending on how they are to be used (i.e. traditional or fast double firing or monoporosa). Frits generally account for 90-95% of the glaze. Small quantities (5-10%) of plastic materials such as china and ball clay, as well as organic additives which adjust rheological features and adhesion to the biscuit are used too. Traditional double firing Initially (in the 60 and 70s), these glazes used frits with a high lead content, such as lead silicates, and boron. Gradually, these fluxes were replaced by others of alkali and alkaline-boron type. Opaque glazes had the same glassy matrix as transparent ones, plus a zirconium silicate content of between 8% and 12%. The softening point of these frits was very low (750-850 °C), whereas the firing temperature was around 950 °C. Fast double firing Glazes suitable for the 30-50 minute firing cycles used in this technology have very short glass maturing times (2-4 minutes). Therefore the frits have moderate softening points and low melting viscosity at maximum firing temperature (about 1050-1100 °C).

266

Wall tiles

Fig. 142. Firing decomposition/synthesis reactions for a monoporosa body.

*103 1.50

1.50

1140°C

1100°C

0.75

0.75 900°C

*103 3.50

650°C

0.00 22.00

32.00

1.75

42.00

1140°C 1100°C 900°C

0.00

650°C

2.00

22.00

42.00

62.00

2θ

Fig. 143. Monoporosa body XRD analysis at different firing temperatures.

267

0.00 52.00

2θ


To achieve these goals, alkaline-boron fluxes are used in most cases. To further optimise viscosity at peak firing temperature, the latest frits include high quantities of CaO and ZnO in their formulation, mainly at the expense of Na oxide. Monoporosa The need for carbonate minerals in wall tile body compounds (useful for introducing calcium and magnesium oxide) has resulted in considerable technological problems in finding appropriate glazes for porous single firing. Gas (CO2) emissions from the calcite and/or dolomite within a temperature range of 750-950 °C (the maturing zone of traditional glazes) was one of the main problems in perfecting this process. The need for glaze softening temperatures higher than 950 °C made it necessary to use new high-temperature fluxing formulations, based on eutectic melting compounds. This was made possible by reducing the quantity of oxides such as B2O3 and Na2O and introducing CaO, MgO, ZnO and K2O as active elements in order to bring about high temperature eutectic fusion. To clarify this behaviour, Figure 144 shows heating microscope softening graphs for two frit samples A and B. Sample A is a glaze for traditional double firing which softens about 60-70 °C earlier than B. Sample A is clearly unsuited to porous single firing because of its tendency to seal the surface too soon, trapping gases in the body. Figure 144 also shows the different chemical compositions for frits A and B, used in traditional double firing and in porous single firing respectively. Besides softening points, other important factors useful in defining the firing behaviour of a monoporosa frit include: surface tension and hot viscosity of the glaze. Low surface tension helps to eliminate any gas bubbles in the glaze during firing. Low hot viscosity values, on the other hand, favour better glaze application and also improve the smooth covering of the surfaces it comes into contact with (engobe and/or body). In addition to reactivity with the engobe and/or the body, glaze expansion coefficient is also of great importance. A good expansion match with body and engobe is essential to control tile curvature. Engobes An engobe is a partially glassy composition usually applied on the body. It is virtually indispensable on monoporosa products and certainly advisable on double firing ones and serves the following purposes: inhibits any reactions of the glaze with chromophore impurities in the body

268

Wall tiles

A

A

B

55/56

53/55

Al2O3

7/8

8/9

B 2O 3

12/13

8/9

CaO

2/3

7/9

MgO

0.5/1.5

2/4

Na2O

6.5/7.5

-

K2 O

2/3

3/5

ZnO

1/2

9/10

ZrO2

8/9

5/6

SiO2 20°C

840°C

960°C

1050°C

1260°C

B 20°C

940°C

1020°C

1060°C

1210°C

Fig. 144. Softening graphs under Leitz heating microscope: two frit samples of different softening points.

improves the expansion match between body and glaze cuts applied glaze costs, as its use makes it possible to apply lower glaze weights. The engobes normally used for monoporosa and double fast firing consist of a percentage of frits (30-40%), ball-clays, zirconium silicate and sometimes feldspar and quartz. The specific functions of each component may be summarised as follows: the frits help to make the glassy matrix the ball-clays provide the required plasticity zirconium silicate improves the degree of whiteness quartz and feldspar help control not only the fusibility of the mix, but also its expansion coefficient. The final features of a monoporosa engobe are: no appreciable glassy phase is developed until 1000 °C: chemical inertia prevails before such temperature high degree of whiteness (covering power) waterproofing against any coloured solutions which may enter from the back of the tile when in service (no staining) adhesion to the fired body and formation of an important intermediate layer between body and glaze. Figure 145 shows two macro photographs which highlight how the engobe reduces contamination by any impurities present in the biscuit.

269


Fig. 145. Macro photographs showing how the engobe reduces contamination by any impurities inside the body.

Basic technological parameters The final properties and appearance of the product not only depend on the chemical and mineralogical nature of its raw material components but also on the practical as well as the technical requirements demanded by the various processes in the production cycle. a - Grinding b - Spray drying c - Pressing d - Drying e - Biscuit firing f - Glazing g - Monoporosa firing h - Glaze firing a - Grinding The purpose of grinding is to reduce the size and homogenise the incoming raw materials until a final, constant particle size distribution is attained. With porous single firing bodies, the extent of raw material grinding, together with other chemical and physical factors, may affect the extent to which carbonates break down during firing and hence appreciably influence the temperature at which gas (CO2) is emitted. The degree of grinding can also influence the reactivity of the components being fired, and the degree to which newly formed compounds are created. Considerable reactivity favours the formation of these compounds and helps improve the mechanical properties of the fired material. The graph in Figure 146

270

Wall tiles

Grinding residue _____ Breaking load after firing (kg/cm2)

.......... Calcite precentage in the grinding residue (%)

Fig. 146. Changes in fired breaking load (left scale) relative to the percentage of calcite in the grinding residue (right scale).

illustrates variations in fired breaking load (kg/cm2) in fired materials made from compositions with different residues after grinding. The same graph also shows the percentage of CaCO3 in the residue, again for the same compositions: post-grinding residue, in the case of compound bodies for either monoporosa or double firing, is generally around 4-6% (at 63 microns - 230 mesh). In the case of red firing bodies, which mainly consist of clayey materials, this value can be as low as 3-4%. b - Spray drying The purpose of this process is to evaporate a part of the water contained in the slip while forming spheroid particles. The particle size distribution of a wall tile body is not so different from that of any other spray-dried body (i.e. single-fired floor tiles, porcelain etc.). The graph in Figure 147 illustrates the similarity of the typical particle size bands for monoporosa and double firing bodies. Note that there is a concentration in both of around 70 to 80% in the 425-180 micron range. c - Pressing The purpose of pressing is to obtain the greatest possible green tile powder density that is compatible with the black core or degassing problems which may arise during firing. Applying different pressing forces will, of course, result in pressed pieces of different bulk densities and thus different shrinkage and porosity values too. 271


Double firing Single-fired wall tiles Microns

Microns

Single-fired

Fast-double

wall tiles

firing

> 600

2%

2%

600-425

13%

13%

425-300

13%

14%

300-250

44%

50%

250-180

14%

13%

180-125

9%

7%

< 125

6%

2%

Fig. 147. Estimated variations in particle size distributions of monoporosa and fast double firing bodies (spray-dried).

In the case of porous products, green bulk density variations do not involve substantial shrinkage changes because shrinkage values for these products are in any case very low (less than 1%). High bulk density values (more than 2.1-2.2 g/cm3) can make the escape of gas from the body more difficult during firing, and may cause boiling problems on the glaze (monoporosa). Bulk density differences within an individual pressed piece (e.g. powder loading errors) can give rise to areas of porosity values that cause glaze application and absorption problems. The graph in Figure 148 shows the range of bulk dry density within which frits having different softening points should not suffer from de-gassing problems. The graphs also show the generally accepted production threshold values (shaded area). Specific moulding pressures for compound wall tile bodies are generally around 200-250 kg/cm2; for red firing bodies, which tend to be more plastic, pressure may be less, even as low as 150 kg/cm2. 272

Moulding (kg/cm2)

Frit softening points (°C)

Wall tiles

1.83

1.91

1.93

1.97

1.99

Dried apparent density (gr/cm3) Fig. 148. Relationship between dry bulk density ranges within which frits with different softening points should not have degassing problems (shaded area).

d - Drying This is an apparently simple stage, since the physical phenomena that accompany the evaporation of residual humidity in the body (4-7%) are sufficiently understood and easily controlled. During this phase, as residual moisture evaporates, the bending strength of the ceramic piece increases because the particles move closer together and develop stronger bonds. In single firing products bending strength must be high: to withstand the mechanical stress of silk-screen printing. MOR must be no less than 25 kg/cm2. With modern day drying cycles it is good practice to keep dimensional shifts during drying within a shrinkage range of 0-0.3% so as to prevent cracking on the faces or edges of the tiles. e - Biscuit firing Firing curves and kiln temperatures must permit and aid the development of reactions between the various components so that the desired fired properties (porosity, bending strength, expansion coefficient etc.) can be obtained. It should be evident at this point that the sintering reactions of a ceramic mass depend on the chemical and physical nature of the body, the degree of grinding, the bulk density of the pressed material, and, finally, on the firing temperature (the graph in Fig. 149 shows how unfired bulk density affects porosity and shrinkage values. 273


The graph in Figure 150 highlights the thermal expansion coefficient of samples fired at different temperatures. In both cases the data refers to a red firing body composition. f - Glazing With wall tiles, the most commonly marketed products belong to the: transparent crystalline and opacified glossy white glazes. Todays market demands a wide range of finishes, including rustic and antique effects. This trend, deeply rooted in the floor tile industry, is also beginning to take hold in the wall tile market, though to a far lesser extent. The most common glaze application devices are the bell units. These allow producers to achieve smooth, mirror-like surfaces by passing the unfired or fired tile through a continuous curtain of glaze of constant thickness and flow rate. In order to optimise application and prevent problems related to the glazing of unfired tiles (monoporosa) there is now a tendency to reduce the quantity of water in glaze slips as far as possible. The rheological behaviour of such suspensions is often far from ideal and takes on characteristics which tend towards a plastic and often thixotropic condition. In the case of bell applications, optimum rheological performance in a traditional standard glaze will be provided by high density, very low yield point, constant viscosity and very low thixotropy values.

Bulkdensity densityafter after drying Bulk drying 33 (gr/cm )) (gr/cm 1,91

1,93

1,97

1,99 20

Contrazione Shrinkage Shrinkage

0,20 15 0,40 0,60

10

0,80 5 1,00 1,20

0

Assorbimento d'acqua Water absorption absorption Water

1,83 0,00

Fig. 149. Influence of post-drying bulk density on water absorption (porosity) and fired contraction.

274

400 350 300 250 200 150 100 50 0

CD.0106

3 Alpha (30+400°C)

Wall tiles

650 750 850 950 1050 1100 1150 - Firing (°C) temperature (°C) Temperatura di cottura Firing temperature Fig. 150. Thermal expansion coefficients of body samples fired at different temperatures.

Figure 151 shows two rheograms for slips of different density. The diagram shows the variation of the speed gradient D (1/s) as a function of shear stress τ (Pa). Table 13 shows rheological parameters such as viscosity, thixotropy and yield point in different glaze suspensions. In order to minimise problems associated with the use of very high viscosity glazes, the bell unit has recently undergone significant changes: elimination of the vibrating sieve replacement of trapezoid tank with a double chamber funnel elimination of the adjustment valve in the final part of the overflow device. The purpose of these modifications is to minimize the development of air bubbles and scratches (bell lines). Glaze water is absorbed into the biscuit during glazing depending on the capillarity of the body. This facilitates adhesion to the biscuit and influences glaze spread and glaze drying time. There is thus a need to control the rate of absorption of the biscuit, a property which depends on its basic composition, the extent to which it is fired and the density and viscosity of the glaze. In monoporosa glazing the process variables to be controlled are many, and tolerances for the following are narrower: rheological characteristics of the glaze application equipment performance post-drying tile features (bending strength, residual moisture and tile temperature).

275


Tau τ [Pa] 300

NR 1 H2O 27.0% 200

100

NR 2 H2O 28.5% 0 0

300 D [1 / s]

200

100

Figure 151. Variations in the speed gradient D (1/s) as a function of shear stress t (Pa) for glaze suspensions of different densities.

Another important aspect is screening and magneting of the glazes. Highly efficient circular vibrating screens are now used almost universally. There are three types of magnet but their efficiency depends on the flow rate and density of the glaze: cylindrical metal bars beehive-shaped rollers. Glaze suspension

Viscosity

Yield point

Thixotropy

(cPoise)

(mPa sec)

(Pa/sec cm3)

Engobe

198

4.5

92.3

Engobe

175

3.0

145.7

White glaze

367

2.5

109.4

Transparent

327

2.5

44.0

Tab. 13. Rheological characteristics such as viscosity, yield point and thixotropy for different glaze suspensions (values indicative only).

276

Wall tiles

g - Monoporosa firing This is a very important stage of production, since all the reactions which determine the features of the final product take place here. Monoporosa firing dynamics are special, since the body mass contains carbonates and must be compatible with glaze characteristics. The glaze must behave eutecticly, have certain refractory characteristics and be fairly gas-permeable up to temperatures in the 950-1030 °C range after which it must melt suddenly. The critical parts of the firing curve and the technical explanations of them are highlighted in Fig. 152. The initial part of the curve up to 800 °C (A) coincides with the pre-heating phase of the material and the destruction of the clayey materials. In part (B), between 800 and 900 °C, the carbonates start to decompose and CO2 escapes. It is essential that the glaze maintain a certain porosity during this stage to aid gas emission. In part (C), between 900 and 1100 °C, synthesis reactions between alkalineearth oxides (CaO, MgO), generated by decomposition of carbonates, take place and residual amorphous phases appear due to the disintegration of the clays. The development of these neo-formed compounds is essential in adjusting and defining the physical and mechanical features of the product. Temp.

Temperature °C

Monoporosa Upper zone

......

Tile Upper zone Lower zone

Tile

Lower zone

Firing time min.

Fig. 152. Simplified firing temperature graph illustrating basic firing aspects of monoporosa materials: A: destruction of clayish minerals B: total expulsion of gas (CO2) C: neo-formed compounds begin to appear D: stabilization of neo-formed crystalline compounds and complete melting of glaze E: fast tile cooling.

277


Sintering of the body is completed in the maximum temperature zone (D): simultaneously, spreading and melting of the glaze are optimised. Zone (E) coincides with the fast cooling phase of the glass down to 600 °C. A slow cooling phase follows, in order to limit the stress generated by transformation of the free quartz, still present in the fired body. The three lines of the graph illustrate tile temperature and in-kiln temperatures above and below the rollers. Different temperature gradients between the upper and lower surfaces of the tile can, within certain limits, condition and control its curvature. This is done by exploiting the expansion/shrinkage features of the material to induce stresses on the upper and lower tile surfaces. h - Glaze (or glost) firing In double firing two kinds of frits, which might be classified as traditional and eutectic composition, are generally used. Firing of both these frits will require curves and temperatures different from those for monoporosa in order to allow the glass to melt properly. The traditional frits will need temperatures of around 1020-1050 °C and the eutectic or monoporosa type frits around 1080-1120 °C. In both cases, maximum temperature is maintained for a few minutes so as to aid glaze spread and enhance glossiness. Whatever kind of frit is used, firing is followed by fast cooling down to 600 °C. Once this temperature has been reached, the subsequent drop to 500 °C must be very slow in order to prevent stress induced by transformation of the quartz (cooling cracks). Firing cycles last from 30' to 50' depending on tile size. Figure 153 shows firing curves for biscuit and glaze. Finished product features The graph in Fig. 154 aims to provide a clearer picture of overall wall tile characteristics, illustrating variations in bending strength, shrinkage and water absorption over various temperature ranges. The same graph also shows the final parameters at the optimum firing temperature.

278

Wall tiles

Glaze firing curve

Temperature °C

Biscuit firing curve

DOUBLE FIRING

Firing time min.

Fig. 153. Glazed tile and biscuit firing curves (double firing).

CD.0112

>

?

Shrinkage (%)

=

Water Abs. (%) Porosity

MOR modulus of rupture (kg/cm2)

<

>

=

<

Firing temp. (°C) Fig. 154. Variations in MOR, shrinkage and porosity for bodies fired at different temperatures. The significance of the different shaded areas is number-coded (4 refers to the most frequently used industrial firing temperatures for these materials).

279


Plant engineering solutions The dynamism of the ceramic industry has been decisive in the ongoing evolution of production processes. The 70s saw significant technological-engineering developments, with a changeover from traditional tunnel kilns to modern processes with roller kilns and fast firing cycles. Adoption of such technologies involved completely new layouts and the huge increase in automation led to significant cuts in production costs. Significant developments in each stage of the production process continued to be made throughout the 80s too. These included: widespread use of wet discontinuous grinding introduction of spray drying increased use of hydraulic presses adoption of fast driers increased use of roller kilns. The introduction of fast firing processes for wall tiles was neither as immediate nor as far-reaching as it was for floor tiles. Traditional firing technology, with the plant engineering and automation solutions available at that time, not to mention the aesthetic quality of the then-produced tiles, was still to some extent irreplaceable. However, the advent of so-called high-temperature fluxing eutectic glazes accelerated technological developments, since the surfaces attainable with these glazes were and still are of high aesthetic quality and comparable to, if not better than, traditional ones. The 90s thus saw a boom in monoporosa and fast double firing processes, resulting in manufacturing improvements and lower industrial costs. Todays production plants offer various levels of sophistication, depending on their degree of automation and type of control systems. The factors influencing the degree of sophistication of a new plant are: local levels of technology and general education plant size required plant flexibility. Figure 155 shows flow diagrams for monoporosa and fast double firing, and highlights the various processing phases (assuming use of continuous grinding, increasingly popular even in small and medium-size plants). Note that with monoporosa the material can be stored after glazing and firing and can also be sent directly from the kiln outlet to the sorting area. In the case of double firing, storage of the biscuit and the glazed tiles takes place in specially designated areas. 280

Wall tiles

Fig. 155. Flow diagram for monoporosa and double firing processes with modern plants featuring continuous grinding.

281


Tiles are moulded in hydraulic presses, dried in fast automatic driers, put through fast firing cycles in roller kilns and sorted using almost entirely automated machinery. Modern electronics has made it possible to automate the movement of green and fired tile cars completely using wire and laser-guided systems. The effects of such rapid technological-engineering developments are evident and may be summarised as follows: increased operator productivity decrease in surface area occupied by plant lower unit industrial costs. Figure 156 shows the lay-out for a fast double firing wall tile plant. 1

2

4

6

5

3a

3b CD.0109

Fig. 156. General plant lay-out for the fast double firing manufacture of about 8000 m3/day of wall tiles. Key: 1. Body batching and preparation - 2. Pressing - 3a. Body firing area 3b. Glazed tile firing area - 4. Glazed and fired product storage 5. Glazing department - 6. Sorting.

Machines A brief description of the main machines used in a typical wall tile production plant follows. Other machines and production processes involving dry grinding and powder re-granulation, which can also be used to produce wall tiles, will not be dealt with here. Dry preparation can be used in double firing where body compositions rich in clayey materials (or materials with very similar morphological features) are available. Production of monoporosa using dry grinding is less widespread and is generally more difficult.

282

Wall tiles

Weighing systems Batching is performed using machinery and equipment of varying levels of automation. With discontinuous grinding, batching of raw materials may be carried out using traditional mechanical leverage systems, or more sophisticated systems involving individual weighing hoppers mounted on load cells. Continuous grinding generally features a microprocessor-controlled on-belt continuous weighing system. The batched mix is temporarily stored in a pre-loading silo, from where it is fed into the continuous mill. In certain cases, where the clay has appropriate features, pre-dispersion of a part of the clayey materials can be effected. In this event, the clay suspension, which already includes the fluidizer (totally or in part), can either be sent, after proper batching, to the mill inlet or can be added to the rest of the body after it leaves the mill. Mills Wet grinding of the body may be performed using either a continuous or discontinuous process: selecting the most appropriate one depends on a number of factors, such as: plant size professional qualifications of work force characteristics of raw materials. As mentioned above, continuous grinding is becoming increasingly widespread. The advantages it provides may be categorised as: technological technical-managerial economic. From a technological viewpoint, the most immediate advantages are: greater consistency of slip characteristics increased density improved rheological features. From a technical-organizational point of view, plant management becomes more rational, while from an economic standpoint continuous grinding implies not only direct labour savings but indirect energy savings too as less water evaporation is

283


required in the spray drier: this is because it is possible to work with slips having a greater solid content than would be possible with discontinuous grinding. To clarify, a brief description of the constructive and functional differences follows. Discontinuous mills are cylinder-shaped and made of very thick sheet metal. They have material loading-unloading ports and their interior is lined with various materials, such as silica, alubit® (sintered alumina) or rubber; the grinding media may be made of silica or alubit®. The drive unit may be fitted with a static frequency converter and PLC control logic so as to allow, where necessary, variations in mill rotation speed even during the grinding cycle itself in order to optimise power consumption and productivity. The continuous mill also consists of a cylindrical steel structure. The cylinder interior is divided into two or three grinding chambers, separated by one or two bulkheads. Hatches that open outwards are provided for inspection and maintenance purposes. The inside of the mill is lined with specially shaped wear-resistant rubber designed to provide maximum grinding efficiency. Once again, grinding media may be silica or alubit®. In discontinuous grinding the properly batched load is introduced into the mill together with the preset quantity of water and fluidizer. Once grinding is over and the residue has been checked, the slip is screened and unloaded into storage tanks. In continuous grinding, instead, the mix is introduced into the mill without interruption by various loading systems. The raw materials and the deflocculant are pre-mixed with the watery suspension containing the screening residues. At the mill outlet, the slip is classified first with a wide-mesh screen, then with a battery of finer screens to sort the material more efficiently. The screen residue is returned to the mill together with the pre-batched water, while the slip is delivered to a holding tank equipped with an agitator from where it can be sent directly to the spray drier. Should the physical characteristics of the materials be satisfactory (naturally low residue values), a part of the clays (and, if needed, the green scrap) can be dissolved in a turbine mixer without being introduced into the mill. Alternatively, the clayey raw materials may be put through a preliminary disintegrator after which they are ground in the mill together with other body components. This option is especially appropriate where bodies have a high proportion of very plastic clays with relatively high natural moisture content (above 15-20%).

284

Wall tiles

Spray driers Spray drying evaporates nearly all the water contained in the slip. Water content is reduced to percentages of around 4 to 7% by spray drying. The slip is injected upwards from the bottom of the drier through variously sized nozzles situated on a ring mounted at the centre of the cylinder: these nozzles are fed by high pressure pumps (25-30 bar). Simultaneously, hot air flows downwards from the top of the cylinder and is distributed tangentially, the counter-flow with the slip resulting in heat exchange. The spray drier may be seen as the final digester of dirty wash water, sludge and all the powders generated during the production process as it batches them into the slip systematically. To ensure compliance with environmental standards, the spray drier can also be provided with dry or wet filters to capture particularly fine dust which would otherwise be dispersed with the steam coming out of the chimney. Where appropriate, the spray drier can be powered by co-generation systems so as to reduce running costs. The spray drier, then, transforms the slip into a powder of controlled particle size and humidity which is then conveyed, via conveyor belts, to the storage silos. Presses This stage of the process is a very important aspect of ceramic production technology. Normally, the aim is to obtain the highest green tile density that is compatible with degassing and black core problems. The pressures generally used for these products are in the 250 ±50 Kg/cm2 range. The currently available hydraulic presses are especially suitable for large tiles. They provide high mechanical/ceramic reliability and cut power consumption to a minimum. In addition to the actual press, key accessories such as the powder filler box and the mould need to be taken into account. The former feed the powders into the moulds and ensure that their cavities are loaded homogeneously. The dies generally used to form the tile are of the entering-punch kind and are normally rubber-lined to cut down the frequency with which they need to be cleaned. The punch may be of the traditional (rigid) type or the so-called isostatic kind, which allows the user to optimise pressing homogeneity and hence obtain uniform bulk density values over the entire tile.

285


Hydraulic press.

This last requirement is met not only by the above-mentioned isostatic punches: special feeding devices which ensure uniform powder distribution in the die are also used. Driers Drying, which eliminates the residual moisture from the just-pressed tiles, is performed on either vertical or horizontal driers. The tiles exiting the presses are transferred on roller conveyors. Vertical driers consist of a load-bearing structure made of steel sections with insulation panelling. The tiles are laid out on racks (consisting of revolving roller shelves) which are hinged on the links of a chain system that supports them and moves them through the drying channels. Internal ducts feed hot air to the drying area and cold air to the cooling area via a series of manually adjusted dampers. Heat is provided by air-flow (in-vein) burners which can run on liquid and/or gas fuels. Horizontal driers consist of metal modules with insulating panels and insulated exterior piping for air re-circulation.

286

Wall tiles

Vertical rapid drier.

The tiles run through the whole length of the drier on a speed-adjustable roller conveyor. The machine consists of a series of identical modules, each one with its own independent thermo-hygrometric and ventilation air flow characteristics. Each zone has its own hot air generators too. Normally, the final section of the machine is designed to stabilise the temperature of the outgoing pieces. With vertical driers, the drying cycle lasts 35-70 minutes; with horizontal driers, cycles can be speeded up to 6-20 minutes. In both cases, however, the duration of the cycle depends on the type of body and the size and thickness of the pieces. With monoporosa, a uniform tile temperature at the drier outlet is essential. Since vertical driers have high thermal inertia, they keep the surface temperature of the outgoing tiles within minimal tolerances even in the event of frequent production line (glazing) stoppages. Glazing machines The glazing lines used for monoporosa and double firing products are essentially the same, except for a few complementary devices such as the fettlers (the type depends on whether the tiles are fired or dried). Monoporosa glazing lines are 287


generally longer. As seen, the majority of wall tiles have smooth, mirror-like surfaces. The most suitable engobe/glaze application devices for this sort of finish are bell units. Special airless atomisation systems have also become increasingly widespread: these provide even distribution of water, fixing agents and engobe. Where antique or rustic effects are desired, repeated application tasks such as flashing, brushing, dry application, etc. will be required in addition to the traditional disc or bell curtain glazing systems. Glazing machine manufacturers have responded to such requirements with machinery that provides high-quality aesthetics. Note also that in order to obtain a proper degree of flexibility on glazing lines there should be numerous application stations and all the relevant equipment useful for rapid product change-over. This inevitably implies very long glazing lines which have, over the years, maintained the same basic structural and functional features. Moving on to decoration, the wide range of print types and patterns currently in demand means that both flat and rotary screen machines are used. Rotary machines are increasingly used because they are fast and require little monitoring and maintenance. There are several kinds of rotary silk-screen printing machines, the main differences lying in the printing technique. Those deserving of special mention are the tampo-print and lithographic decor machines and others employing more sophisticated decor techniques, such as the incave-graphic method which uses hollowed silicon rollers carrying the decor image. Glazing line automation concepts are evolving steadily. This has led to the availability of: devices for dynamic applied glaze weight control, PLC units to manage drive speed via inverters, etc. However, the extreme automation now seen at just about every stage of the production process has yet to appear in the glazing department as personnel still play an essential, priority role. Kilns The 70s saw the industry start switching from slow to fast firing: hence all stages of the firing process have been reviewed accordingly, as fast firing has created new technological opportunities. Today, an extensive knowledge of raw materials, the nature of the bodies, the composition of engobes and glazes as well as the wealth of acquired experience have all resulted in the simplification and standardisation of this phase too. Ever-more complex, sophisticated kilns and equipment now make it possible to achieve a degree of temperature adjustment and control which, even just a few years ago, would have been unthinkable.

288

Wall tiles

Single layer kiln.

Kiln cross-sections have increased and combustion systems have been improved, providing constant, uniform temperature over the entire kiln cross-section. Burner lay-out and number, of course, differ in the double firing and monoporosa kilns. The burners and their combustion chambers are normally sized so as to provide high combustion flame speeds and low per-unit thermal potential. In addition to high speed burners, todays technology provides other burner types which allow the user to fluctuate the flame over the kiln cross-section and adjust flame power, cyclical flow and duration. This optimises the way in which the heat is distributed over the tiles, especially at critical points of the firing curve. The burner unit generally constituting the combustion chamber may be made either of refractory rammed material or silicon carbide. The latter gives better heat resistance at maximum firing temperatures and also provides greater resistance to thermal shock. Tile conveyance rollers are a key feature on modern kilns. These may be made of ceramic or metal; each kiln zone is fitted with specific roller types suited to that particular stage of the firing cycle. Adoption of rapid single or multi-layer firing has facilitated tile handling, favoured production of large tiles and increased production flexibility. A high degree of flexibility in determining and setting firing curves, together with maintenance of firing conditions, is ensured by microprocessor control systems which keep temperatures within very narrow tolerances. Computers provide high levels of kiln automation and also allow the user to monitor and store the process data.

289


Sorting Although sorting does not affect product features, it is a delicate stage in the production cycle. Here too, automation is pervasive, covering everything up to boxing and palletizing. Tile geometry (flatness) can be checked by electronic and/or camera systems which then send the product to specific outlet points. The number of outlets depends on classification logic and product flatness/shading criteria. Normally, there are seven outlet positions (two shades and three sorting classes). In highly automated installations the operator need only analyse aesthetic defects and code the tiles to define their class. Handling and storage systems In addition to the traditional metal trolley tile handling and storage systems where the trolleys are pushed along tracks by rigid push and transfer bars, handling and storage installations with automatic drive vehicles are becoming increasingly widespread. In double firing, biscuit storage systems consist of wide racks. The tiles are then picked up by devices provided with suction cups which transfer them to the glazing line. Other solutions feature tile stacking platforms. Tiles are then sent on to the glazing lines via trolleys which move along flat trackless surfaces, guided by a floorembedded wire or laser beam system. The same AGV (wire-guided) or LGV (laser-guided) systems can be used in single firing to move the trolleys carrying both the green and fired tiles. In both cases, a computerised control station monitors all transport system handling, effecting real-time monitoring of production flows to and from the tile storage area. Conclusions Porous single firing and fast double firing, following initial teething problems, have undoubtedly resulted in technical and aesthetic achievements which are comparable and certainly superior to those of traditional double firing. Today, monoporosa and fast double firing may be seen as two complementary technologies, and only an in-depth assessment of the technical, economical, cultural and environmental aspects facing the manufacturer will allow selection of the most appropriate plant. 290

Wall tiles

Although the decorative features achieved to date have been excellent (they would have been unimaginable just 10 years ago), research both technical and graphic is likely to result in even more exciting results. Larger tiles and surface effects which realistically resemble marble and breccia will allow ceramic products to compete with natural materials on an almost even footing, as many now have technical properties superior to their natural equivalents.

291


292


Chapter X FLOOR TILES

Foreword Simple brick-like cotto tiles have been used as paving for so many centuries that it is impossible to establish exactly when they first appeared. Painted and glazed natural-coloured terracotta, and decorated paving tiles are, however, thought to have appeared in the 11th century AD. From then on, up to the 20th century, floor tile production in Italy was very limited, largely being undertaken by craftsmen working for wealthy clients. The acceleration of development that was to transform Italian ceramics into a true industry took place at the beginning of the 1950s, when cottoforte first came on the market; this classically Italian product was made using traditional double firing techniques, and had features which made it ideal for paving. Two other key technological developments were made during the 50s: the automatic press and the tunnel kiln. Increasing market demand combined with these two innovations transformed craftsmens workshops into true industrial companies. During the 60s fast firing plants appeared. These introduced single firing technology, thus paving the way for the use of highly automated continuous production systems. In simple terms single firing means a production system in which the tile body and the on-tile glazes are fired together in a single operation. It became widely established around the mid-70s. The ceramic industry of the 70s and 80s was dominated by three highly important factors: Automation Fast firing The energy crisis. The introduction of single firing, with the use of fast firing cycles, was to lead to substantial changes, especially as regards firing technology and kiln construction. Innovation was certainly not limited to firing: changes occurred throughout the production cycle, especially in tile handling. Tunnel kilns, which conveyed stacked tiles on wheeled cars, were abandoned in favour of roller kilns where the tiles were supported by refractory plates. The latter developed into present-day kiln versions where tiles are moved by rollers directly without the need for an intervening support. 293


The same period also saw modernisation of the kiln structure and improved reliability, with greatly improved firing performance and much enhanced control of thermodynamic parameters. Thermodynamic notions vis-à-vis the new roller kilns were entirely different from those applied to tunnel kilns. New ideas concerning fast tile firing were based on the theoretical models developed by Prof. Korac, and with roller kilns these ideas were soon adopted on an industrial scale. These models identified the single layer kiln as the one with the lowest specific consumption too. In the context of the 70s energy crisis, then, this new technology was also to spread fast on account of its economic benefits. Yet the success of vitrified single firing products was not just related to energy savings: there were other plus factors, such as simplification of the glazing system, the opportunity to produce large tiles, the adoption of advanced automation systems. The introduction onto the market of vitrified single firing products meant that consumers could enjoy greatly improved physical and mechanical features but without having to compromise on aesthetic content. This combination of technical quality and aesthetic content plus lower manufacturing costs has, then, made floor tiles extremely popular. The market Figure 157 shows the evolution of vitrified single firing tile output in Italy from 1990 to 1998. This data is compared with porcelain tile output over the same period. Note that the slight fall off in output in 1991 and 1992 was followed by a constantly upward trend that lasted until 1996. Over the same period, porcelain tiles gained ground constantly. The following graph (Fig. 158) shows changes in the percentages of output accounted for by vitrified single firing tiles and porcelain tiles as from 1990. Fig. 159, however, shows annual output increases for these two products.

294

Floor tiles

Fig. 157. Comparison of Italian vitrified single firing tile output and porcelain tile output from 1991 to 1998 (the porcelain tile figures for 1997-1998 include 27 and 70 million m2 of glazed porcelain tiles respectively). Fig. 158. Percentages of total floor tile output accounted for by vitrified single firing and porcelain tiles, 1991-1998.

Fig. 159. Annual output variations for vitrified single firing tiles and porcelain tiles (variations on 1991 output levels).

295


Product classification The term vitrified single fired tile is self-explanatory. It refers to a tile body with a compact mass (grès in Italian), although international standards define different degrees of porosity, expressed as water absorption, for different classes of materials (see Figure 160). The classification below is based on ISO 13006 standards which distinguish products according to their degree of porosity: Note that some products included in the BIb family are totally frost-proof. Water absorption

Shaping

extrusion

dry pressing

GROUP I ≤ 3%

Old EN

GROUP IIA 3% - 6%

Old EN

GROUP IIB 6% - 10%

Old EN

GROUP III > 10%

Old EN

GROUP A1

EN 121

GROUP AIIa1

EN 186/1

GROUP AIIb1

EN 187/1

GROUP AIII

EN 188

GROUP AIIa2

EN 186/2

GROUP AIIb2

EN 187/2

GROUP BIIa

EN 177

GROUP BIIb

EN 178

GROUP BIII

EN 159

GROUP BIa ≤ 0.5%

EN 176

GROUP BIb Bib 0.5% - 3%

Fig. 160. Classification of floor tiles according to ISO-EN 13006 standards.

Technical features Table 14 summarises the main features of the materials belonging to the various families (specifications required by UNI EN ISO 10545-1/17 standards).

296

Tab. 14. Floor tile technical characteristics according to ISO standards.

297 ISO 10545.16 ISO 10545.17

COLOUR DIFFERENCE

FRICTION COEFFICIENT**

> 22 N/mm (min. 20)


TEST AVAILABLE

CLASS 3 min ****

TEST AVAILABLE

CLASS GB min - UB min ***

REQUIRED FOR SUB-ZERO EXTERNAL USE

REQUIRED **

TEST AVAILABLE

TEST AVAILABLE

TEST AVAILABLE


a) max. 175 mm 3

min 1100 - min 700 N


TEST AVAILABLE

CLASS 3 min ****




REQUIRED **

TEST AVAILABLE

TEST AVAILABLE

TEST AVAILABLE



TEST AVAILABLE

CLASS 3 min ****




REQUIRED **

TEST AVAILABLE

TEST AVAILABLE

TEST AVAILABLE


a) 540 mm3

a) 345 mm 3

min 800 - min 500 N

> 18 N/mm2 (min. 16)

> 6, < 10 % (max. 11 %)

STANDARD-PRESCRIBED VALUE

B II b ( H2O = 6 - 10 % )

min 1000 - min 600 N

2

30 N/mm (min 27) 2

> 3 , < 6 % (max. 6.5%)


> 0.5, < 3% (max. 3.3 %)


B II a ( H2O = 3 - 6 %)

Standards 10545.1 and 10545.2 regard sample acceptance methods and size and surface quality characteristics. * Depending on thickness (≥ 7.5 mm ≥ greater than 7.5 mm) ** To be effected only where used as paving *** Unglazed products **** Glazed products For more detailed information see UNI publications (Italy).

I SO 10545.14


ISO 10545.10

MOISTURE EXPANSION

ISO 10545.13

ISO 10545.9



ISO 10545.8


ISO 10545.13

ISO 10545.7

b) SURFACE ABRASION RESISTANCE**


ISO 10545.6

a) DEEP ABRASION RESISTANCE**

ISO 10545.12

ISO 10545.4

BREAKING STRENGTH*

FROST RESISTANCE

ISO 10545.4


ISO 10545.11

ISO 10545.3

WATER ABSORPTION

CRAZING RESISTANCE

STANDARD

FEATURES

B I b ( H2O = 0.5 - 3 %)

Floor tiles


The table clearly shows why these materials, capable of withstanding heavy wear and mechanical stress, are especially suited for use as floor tiles. In the context of single firing floor tiles, glazes assume enormous importance and should be seen as an integral part of the tile itself. Since the advent of vitrified single firing, glazes have gone through several stages of development. The first glazes repeated the features of products obtained by traditional firing. Later, during the 80s and early 90s, marked changes in the technical characteristics of the body led to the development of new glazes (grains, sintered and devitrified materials). These were generally applied using a mixture of dry and wet techniques. The resulting products were of a rather simple appearance based on combination of colours and sizes; technical aspects took precedence over aesthetic ones. The trend over the last few years has been towards glazed surfaces with a natural stone look. Especially popular are surfaces with an aged (or antique) appearance. Summing up, it can be said that the evolution and optimisation of technical features has made excellent aesthetic results available to all. In terms of abrasion resistance most tiles of this kind are Class 4, with some Class 5, as per test methods defined in ISO 10545-7. With regard to body features, there is an increasing, though not generalised, tendency for porosimetrical parameters to approach those of porcelain tiles (i.e. water absorption values around 1%). This trend towards emulation of porcelain tiles is slowly shifting single firing products out of their traditional water absorption range (3% < WA < 6%) into the BI UNI EN 176 class. Aesthetic features Size Much of the production over the last few years has focussed on square, average-size tiles. In 1996, Italian production of the 30 × 30 cm size accounted for about 40-45% of total output. Rectangular sizes took a far less significant share of the market. Surface appearance Recently, there has been an increasing trend, especially in Italy, towards tiles with an antique or stone-like appearance (matt satined surfaces). Such finishes are achieved through use of specifically developed glazes and structural-type morphological operations on the tile surface and perimeter. Attaining such results requires quite complicated glazing lines consisting of 298

Floor tiles

various application devices/systems such as discs, spray-guns, screen printing machines, brushes, etc. Where glossy glazes are used they sometimes have a printed matt relief effect to increase their abrasion resistance. Recently, innovative new tile types with a very high aesthetic content have appeared: these require especially sophisticated and complex production technology. For example, excellent special effects can be obtained by applying powder or grained glazes (generally dry-applied) and then polishing after firing. These techniques provide highly attractive marble or breccia-like finishes. Raw materials for bodies General features According to one kind of commercial classification, bodies suitable for the manufacture of vitrified single firing floor tiles can be divided into red and white. In both cases, raw materials are of two basic types: clayey materials complementary materials (feldspars, feldspathic sands, quartzes, calcites). In red bodies, the clayey materials include illitic-chloritic and illitic-kaolinitic clays. In white bodies, the ball-clay or china-clay types, of illitic-kaolinitic or kaolinitic nature respectively, predominate. Complementary raw materials include minerals with fluxing and/or inert features: such as feldspathoids, feldspars, feldspathic sands and quartz. The quantitative ratio between clayey, fluxing and quartz materials will, of course, depend on the intrinsic mineralogical nature of the clays, on particle size distribution and the characteristics of the fluxes. White bodies are themselves subdivided into potassium and sodium types according to the prevalence of these oxides in their formulation. Figure 161 shows the chemical and physical characteristics of different raw materials suitable for white and red single firing tiles. Summing up the functions of the components: the ball and red clays provide plasticity and bending strength to the unfired tiles. The china-clay types are of secondary importance in determining green tile characteristics, yet play a key role in controlling vitrification on fired tiles on account of their high alumina content. Feldspars, feldspathoids, feldspathic sands are fluxing elements; their melting characteristics help make the materials denser during firing. Quartz balances the above, since it adjusts their viscosity and conditions shrinkage during firing. 299


2

2

3

3

3 2

1

1

1

1 3

2

1 3

2

3

2

1

Key: A - plastic clay; B - semi-plastic white clays (china clay); C - red vitrifiable clays; D - potassium feldspars; E - sodium feldspar; F - feldspathic sands; G - quartz.

PHYSICAL FEATURES

A

B

C

BREAKING LOAD BEFORE DRYING

kg/cm2

8/10

6/8

10/25

BREAKING LOAD AFTER DRYING

kg/cm2

20/30

15/20

30/40

A

B

C

POROSITY %

5/10

15/20

2/6

SHRINKAGE %

3/5

2/4

4/6

80/120

200/300

3 2

1

FIRING AT 1100 °C

3

2 1

BREAKING LOAD AFTER FIRING

kg/cm2 180/250

Fig. 161. Vitrified single fired tiles: chemical-physical characteristics of various raw materials (single firing technology). The lowest percentages (0.5-1.0%) are shown in the table only as they will not show up on the graph.

Body composition Unlike white, red bodies mainly consist of clays with a high iron content and small percentages of complementary materials such as feldspars and quartzites. White-firing bodies, on the other hand, are more diversified and feature a better balance of raw materials that includes iron-free clays, feldspars, feldspathoids, quartz. 300

Floor tiles

Figure 162 shows, as an example, two compositions for white and red bodies respectively. White bodies, due to the inherent nature of their raw materials, have wider and more stable size and absorption ranges than red ones because of their higher alumina and lower ferric oxide content, and have very different vitrification characteristics (see Fig. 163). The increase in linear shrinkage and the decrease in water absorption develop less gradually and at lower temperatures in red gres compared to white gres. The red develops more liquid phase of lower viscosity from the illitic materials. This is particularly true in the case of red bodies consisting mainly of clayey minerals (7580%). Where red clays are mixed with feldspars and quartz, firing behaviour (variations in shrinkage and porosity) is closer to that of light-coloured bodies. Figure 164 shows composition areas occupied by various formulations inside the Al2O3 - SiO2 - Alkali tertiary diagram, while Figure 165 shows the areas which can be derived from the same compositions (A = Feldspathoids, B = Quartzes, C = Clay minerals). In particular, the diagrams show six light-coloured body compositions of which three have predominantly sodium fluxing agents (Ws) and three predominantly potassium fluxing agents (Wp), together with four red vitrified single firing compositions.

Argille rosse greificabili Red vitrifiable clays Feldspatoidi Feldspathoids

70 - 80 % 15 - 20 %

Quarzo

10 - 20 %

Quartz

Fig. 162. Body compositions for red and white vitrified tiles.

301


<

<

=

=

<

=

Fig. 163. Vitrification curves for white and red-firing bodies. TYPE

SiO2

Al2O3

TiO2

Fe2O3

CaO

MgO

K2O

Na2O

P.F.

Ws Ws Ws R R R R Wp Wp Wp Wp

67.93 66.80 66.35 68.00 66.37 68.30 65.35 66.75 67.03 66.18 67.54

18.70 18.09 21.33 15.85 16.00 17.15 15.60 19.95 18.00 18.80 18.76

1.15 1.50 0.76 0.53 1.07 0.53 0.76 0.48 0.54 0.54 0.71

1.51 1.96 1.22 4.56 4.79 3.04 4.25 1.71 1.52 1.52 1.06

1.35 1.42 0.68 0.36 0.98 0.54 1.15 0.57 0.67 0.67 0.57

0.35 0.38 0.40 0.46 0.96 0.54 1.55 0.51 0.95 0.95 0.48

1.20 1.50 1.76 3.00 2.91 2.80 3.33 5.00 4.00 4.37 4.37

3.44 3.36 3.18 1.91 1.52 1.73 1.63 1.80 1.55 1.30 1.30

4.49 5.46 4.50 4.81 5.70 5.65 6.15 3.66 5.78 5.78 5.50

SiO2

Alcali Alkali

2

1 2 3

Wpp GB Wss GB R Gr

Ws = white with sodium = gres grosso GB sbody

2

R = red body grosso Gr = gres 1 3 Wp =p white bodygrosso GB = gres with potassium

CD.0064

3 1

Al2O3

Al2O3

Na2O

CD.0063

K2O

Fig. 164. Composition areas for various body formulations within the Al2O3-SiO2-Alkali tertiary system. 302

Floor tiles

Fig. 165. Composition areas of red and white single firing bodies related to mineralogical components.

Raw materials for glazes The physical properties of glazes, such as resistance to wear, acids and alkalis, ultimately determine where such products can be used. This is why research into new kinds of glass with innovative physical-mechanical and chemical properties is ongoing. Glaze properties basically depend on the raw materials used in their formulation and on firing temperature. Formulation is based on the use of different kinds of frit (medium and high viscosity) in varying proportions; nowadays eutectic-melting frits (calcium and zinc) are used too. Other natural and synthetic materials are introduced into the glazes together with the frits, depending on the exact texture and glaze properties the manufacturer is aiming at. Some of these non-fritted components, such as feldspars, nepheline and zinc oxide form, together with the frits, the vitreous matrix of the glaze. Others, such as zirconium sand and corundum, are partly dissolved inside the glass and help improve abrasion resistance, or else work as opacifiers (zirconium silicate) or matting agents (alumina oxide). Other glaze components include: Anatase (titanium dioxide). Not only a matting agent: it also has positive effects on the mechanical and chemical properties of the glass. Wollastonite, calcium and magnesium carbonate. Minerals with an alkaline-earth base which act as matting agents and help form the vitreous matrix. One of the most important features of a glaze is its expansion coefficient, which must match that of the body so as to control the final tile curvature.

303


Basic technological parameters The final features of a product depend not only on the chemical and mineralogical nature of the raw materials, but on the technical conditions during the production process and these two factors will be inter-related particularly in the following areas: a - Grinding b - Spray drying c - Pressing d - Drying e - Glazing f - Firing a - Grinding The purpose of grinding is to reduce the size of the raw materials to a final, constant required particle size. With vitrified bodies, the degree of grinding of the raw material, together with other chemical and physical factors, may affect the degree of vitrification and thus influence shrinkage and porosity values. The graph in Figure 166 shows variations in shrinkage of a given composition against grinding residues over a range of temperatures. Generally speaking, post-grinding residue, for compound vitrified bodies, is around 7-10% (63 microns-230 mesh). In the case of red bodies, mainly consisting of clayey materials, this value may drop to 4-6%. b - Spray drying The purpose of this process is to evaporate a part of the water contained in the slip while forming spherical particles. Such particles are of a similar size in all types of ceramic tile body. Fig. 167 illustrates particle size distribution for a single-fired body. Note that some 70-80% of particles are grouped in the 180-425 micron range.

304

Floor tiles

Fig. 166. Fired shrinkage as a function of grinding residue variations (checked with a 63-micron screen).

50% 40% 30% 20% 10% 0% 600

425

300

250

180

125

microns

Microns

Vitrified single fired tiles

> 600

1%

600-425

9%

425-300

10%

300-250

42%

250-180

17%

180-125

14%

< 125

7%

63

Fig. 167. Particle size distribution of a spray-dried vitrified single firing tile powder.

305


c - Pressing Pressing aims to achieve the greatest possible green tile powder density that is compatible with the black core or degassing problems which may arise during firing. Applying different pressing forces will, of course, result in pieces of different bulk densities and thus different shrinkage and porosity values too. The graphs in Figures 168-169 highlight variations in these two parameters as a function of moulding pressure and firing temperature for a white body. In white bodies specific moulding pressure is around 250-300 kg/cm2 (white). For red bodies, which tend to be more plastic, pressure may be lower, around 200 Kg/ cm2. d - Drying This is an apparently simple operation, since the physical phenomena which take place during the evaporation of residual body moisture (4-7%) are understood and easily controlled. Note, however, that extremely kaolinitic bodies do tend to expand after drying, resulting in low unfired bending strength. Where drying is too fast this may lead to cracks on the tiles. With todays cycles, post-drying expansion should be a maximum of 0.3% or 0.4% and may even be negative. Dried piece bending strength should be at least 25 kg/cm2. e - Glazing The catalogues of todays ceramic companies feature countless types of glazes. Generally, however, they all belong to one of the following families: glossy matt (devitrified) rustic. These products are obtained by layering different glazes using dry and wet application techniques. Wet techniques require special devices such as discs, spray-guns and silk-screen printing machines. With dry materials (grains and agglomerates) powder applicators are used. The use of so many kinds of glaze requires a very flexible production plant, especially on the glazing lines themselves. Furthermore, each application and the numerous kinds of glaze which overlap on the same product have very different features, thus necessitating different know-how and skills for each product. Some floor tiles require only simple disc applications where the glaze is applied in minimal quantities (up to 500 g/m2), while others require complex procedures (up to 20 different applications) involving applied glaze weights of up to 3 kg/m2. These glazes may be both wet (density of 1.2 - 2.0 kg/l) and dry applied (highly 306

Floor tiles

Fig. 168. Fired shrinkage and water absorption as a function of specific moulding pressure.

Fig. 169. How green bulk density influences post-firing shrinkage and water absorption. 307


variable particle size distribution). It is often necessary to apply binders and other organic products. The green body is therefore subject to considerable stress and must be able to absorb considerable quantities of water without bending beyond a certain limit, as this would compromise silk-screen printing operations. f - Firing Optimum firing curves allow fusion reactions and crystallisation to take place on the various body and glaze components. These reactions help produce the required absorption and shrinkage goals, while improving the technical features and the finish of the glazed surface. The vitrification conditions of a ceramic mass evidently depend not only on the chemical and physical nature of body but also on the extent of grinding, the bulk density of the pressed material and on maximum firing temperature. Fig. 169 shows the influence of green tile bulk density on fired shrinkage and porosity of a red vitreous body. To evaluate the characteristics of a body, it is very important to define its size stability and change in porosity (if any) where firing is performed using different temperature cycles and/or gradients. Figure 170 highlights variations in breaking load (MOR), shrinkage and water absorption at different temperature intervals for a typical vitrified single firing body. The same graph also identifies the physical-mechanical characteristics relative to optimum firing temperature. Plant engineering solutions The dynamism of the ceramic industry has been decisive in the ongoing evolution of the production process. The 70s saw significant technological-engineering developments, accompanied by a shift from double firing to single firing production processes. Adoption of such technology, which uses roller kilns and rapid firing cycles, involved completely new layouts. Shortened firing times and single/multi-layer roller kilns led to huge increases in automation while cutting production costs significantly. Important developments in each stage of the production process continued to be made throughout the 80s too. These included: widespread use of wet discontinuous grinding with Alsing mills introduction of spray drying increased use of hydraulic presses adoption of fast driers increased use of roller kilns. 308

Floor tiles

<

?

> >

=

=

<

Fig. 170. Variations in physical-mechanical characteristics at different temperatures. The significance of the different shaded areas is number-coded (4 refers to the most frequently used industrial firing temperatures for these materials).

The 90s witnessed further refinement of these production processes, leading to increased output capacity and lower maintenance costs thanks to technological innovation and massive use of automation. The same period also saw marked improvements in tile finish and decoration as a result of fast-changing market demand. Todays production plants offer various levels of sophistication, depending on their degree of automation and type of control systems (department supervisor). The main factors to assess in deciding the level of sophistication of a new plant are generally the following: local levels of technology and general education plant size required plant flexibility. The flow chart in Figure 171 illustrates the various stages of the vitrified single firing process. Continuous grinding now becoming more widespread even in small and medium-size plants is followed by pressing with high-tonnage hydraulic presses, 309


Fig. 171. Vitrified single firing process block diagram for a modern plant (with continuous or discontinuous grinding).

310

Floor tiles

drying with fast automatic (both vertical and horizontal) driers and fast firing using roller kilns. The sorting department uses almost entirely automated machinery. Electronic control systems now allow full automation of green and fired tile handling, with storage cars being wire or laser-guided. The effects of such rapid technological-engineering developments are evident and may be summarised as follows: increased operator productivity decrease in surface area occupied by plant lower unit costs. An analysis of the various stages of the production process shows that it is possible to store the material after glazing. The material can also be sent directly from the kiln outlet to the sorting line. Figure 172 shows the lay-out for of modern plant using continuous grinding technology. 1a

4

1b

6

2

3

Fig. 172. Lay-out for production of vitrified single firing floor tiles. 1a. Body preparation; 1b. Glaze preparation; 2. Pressing and drying dept.; 3. Glazing dept.; 4. Unfired/fired tile storage area; 5. Firing dept.; 6. Sorting dept.

Machines A brief description of the main machines such as wet grinding mills and spray driers used in a typical single firing production plant follows. Although other machines and production processes (involving dry grinding and powder re-granulation) can be used to produce floor tiles, they are much less commonplace and will not be dealt with here. Weighing systems Batching may be performed using machinery and equipment of varying levels of automation. With discontinuous grinding, batching of raw materials may be carried out using traditional mechanical leverage systems, or more sophisticated systems 311

CD.0075

5


involving individual weighing boxes with deformable load cells. Continuous grinding generally features a microprocessor-controlled on-belt continuous weighing system. The batched mix is temporarily stored in a pre-loading silo, from where it is fed into the continuous mill. In certain cases, where the clay has appropriate features, pre-dispersion of a part of the clayey materials can be effected. In this event, the clayey suspension, which already includes the fluidizer (totally or in part), can either be added, after proper batching, to the mill or can be placed in storage tanks for future mixing with the milled slip. Mills Selecting either a continuous or discontinuous process depends on a number of factors, such as: plant size professional qualifications of work force characteristics of raw materials. Continuous grinding is becoming increasingly widespread as it offers technological, technical-organizational and economic advantages. From a technological viewpoint, the most immediate advantages are: greater consistency of slip characteristics increased density improved rheological features. From a technical-organizational point of view, plant management becomes more rational, while from an economic standpoint continuous grinding implies not only direct labour savings but indirect energy savings too as less water needs to be evaporated in the spray drier from the higher density slip compared to that from discontinuous mills. To clarify these points, a brief description of the constructive and functional differences follows. Discontinuous mills are cylinder-shaped and made of very thick sheet metal. They have material loading-unloading ports and their interior is lined with various materials, such as alubit® (sintered alumina), rubber or rarely silica; the grinding media may be made of silica or alubit®. The drive unit may be fitted with a static frequency converter and PLC control logic so as to allow, where necessary, variations in mill rotation speed even during the grinding cycle itself in order to optimise power consumption and productivity. The continuous mill also consists of a cylindrical steel structure. The cylinder interior is divided into two or three grinding chambers, separated by one or two bulkheads (separating plates). 312

Floor tiles

Grinding department with continuous mill.

Even more modern continuous mills do not have dividing separation plates but control the flow of material through them with spiral lifters built into the rubber linings. Hatches that open outwards are provided for inspection and maintenance purposes and for replenishing the grinding media. The inside of the mill is lined with specially shaped wear-resistant rubber designed to provide maximum grinding efficiency. Once again, grinding media may be made of silica or alubit®. In discontinuous grinding the properly batched load is introduced into the mill together with the preset quantity of water and fluidizer. Once grinding is over and the residue has been checked, the slip is screened and unloaded into storage tanks. In continuous grinding, instead, the mix is introduced into the mill by various loading systems. The raw materials and the deflocculant pre-mixed with the watery suspension containing the screening residues are introduced together. At the mill outlet, the slip is sieved first through a wide-mesh screen, then with a battery of finer screens to classify the material more efficiently. The screening residue is returned to the mill together with the pre-batched water, while the slip is delivered to a holding tank equipped with an agitator from where it can be pumped to the spray drier. Should the physical characteristics of the materials be satisfactory (low residue values), a part of the clays (and, if need be, the green scrap) can be dispersed in a turbo-mixer without being introduced into the mill. Alternatively, the clayey raw materials may be pre-dispersed after which they are ground in the mill together with other body components. This option is particularly appropriate where bodies have a considerably high 313


proportion of very plastic clays with relatively high natural moisture content (above 15-20%). Spray drying Spray drying evaporates nearly all the water in the slip leaving a powder containing just 4-7% moisture. The slip is injected upwards from the bottom of the drier via nozzles of various size situated on a ring placed at the centre of the cylinder: these nozzles are fed by high pressure pumps (25-30 bar). Simultaneously, hot air flows downwards from the top of the cylinder and is distributed tangentially. The counter-flow of the hot air and the atomised droplets of slip results in heat exchange. The spray drier may also be seen as the final consumer of dirty wash water, sludge and all the powders generated during the production process. To ensure compliance with environmental standards, the spray drier can also be provided with dry or wet filters to separate and collect particularly fine dust which would otherwise be dispersed with the steam coming out of the chimney. Where appropriate, the spray drier can be powered by co-generation systems so as to reduce running costs. The spray drier, then, transforms the semi-finished slip into a powder of controlled particle size and humidity which is then conveyed, via conveyor belts, to the storage silos. Presses This stage of the process is a very important aspect of ceramic production technology. Normally, the aim is to obtain the highest green tile density that is compatible with degassing and black core problems. Moulding pressures used for these products generally fall within the 200 - 350 Kg/cm2 range or may be higher still. Todays hydraulic presses are highly developed and given the available pressing power, they are also ideal for the production of large tiles. High mechanical/ceramic reliability and greatly reduced power absorption provide further benefits. In addition to the actual press, accessories such as the die and powder filler box play a key role. Most of the dies used to form vitrified single fired tiles are of the entering punch type and are normally rubber-lined so as to cut down the frequency with which they need to be cleaned. Rubber is also used to obtain structured surface effects. The punch may be of the traditional rigid or more innovative isostatic type. The latter allows optimisation of pressing homogeneity and thus ensures uniform bulk density right across the tile. 314

Floor tiles

Hydraulic press.

This is an important concept because, with vitrified single firing floor tiles, not only pressing power but homogeneous loading of the powders inside the die cavity is essential. Proper loading is normally achieved by using alveolar grating (grid) as opposed to the more traditional slatted ones. Imperfect loading can lead to incorrect tile geometry (orthogonality and squareness). Such problems can be resolved using special powder feeders or, as mentioned above, isostatic punches, which have proved highly successful in reducing geometry defects of the finished tiles subject to shrinkage. Driers Drying, which eliminates the residual moisture from the just-pressed tiles, is performed on either vertical or horizontal driers. The tiles exiting the presses are transferred on roller conveyors to the driers. 315


Press cavity powder feeding device (note the alveolar grating).

Vertical driers consist of a load-bearing structure made of steel sections with insulated panelling. The tiles are laid out on racks (consisting of revolving roller shelves) which are hinged on the links of a chain system that supports them and moves them through the drying channels. Internal ducts feed hot air to the drying area and cold air to the cooling area via a series of manually adjusted dampers. Heat is provided by air-flow burners which can run on liquid and/or gas fuels. Horizontal driers consist of metal modules with insulating panels and insulated exterior piping for air re-circulation. The tiles run through the whole length of the drier on a speed-adjustable roller conveyor having one or more levels. The drier consists of a series of identical modules, each one with its own independent thermo-hygrometric and ventilation air flow characteristics. Each zone has its own hot air generators too. Normally, the final section of the machine is designed to stabilise the temperature of the outgoing pieces. With vertical driers, the drying cycle lasts 35-70 minutes; with horizontal driers, cycles can be speeded up to 6-20 minutes. In both cases, however, the duration of the cycle depends on the type of body and the size and thickness of the pieces.

316

Floor tiles

Three-layer horizontal roller drier.

Glazing machines The world tile market has become an increasingly sophisticated and diversified one, requiring machines and equipment capable of producing a wide range of glazing and decoration applications. Glazing machine manufacturers have responded to this demand with machinery that provides superb aesthetic effects yet still relies to a large extent on the manual skills inherent in all glazing operations. Note also that in order to obtain a proper degree of flexibility on glazing lines there should be numerous application stations and all the relevant equipment useful for rapid product change-over. This inevitably implies very long glazing lines which have, over the years, maintained the same basic structural and functional features. Where antique or rustic effects are desired, repeated application tasks such as flashing, brushing, dry application, etc. will be required. Where glossy finishes are required the engobe and glaze are best applied using bell units. There is also an increasing trend towards the use of spray-gun devices that evenly distribute water, fixing agents and (airless) engobe over the tile surface. Moving on to decoration, the wide range of print types and patterns currently in demand means that both flat and rotary screen machines are used. Rotary machines are increasingly popular because they are fast and require little monitoring and maintenance. 317


Several kinds of rotary silk-screen printing machines exist, the main differences lying in the printing technique. Those deserving of special mention are the tampoprint and lithographic decor machines and others employing more sophisticated techniques, such as the incave-graphic method which uses hollowed silicon rollers carrying the decor image. Glazing line automation concepts are also evolving steadily. This has led to the availability of: devices for the dynamic control of the applied glaze weight, PLC units to manage drive speed via inverters, etc. However, the extreme automation now seen at just about every stage of the production process has yet to appear in the glazing department as personnel still play an essential, priority role. Kilns The 70s saw the industry start switching from slow to fast firing: hence all stages of the firing process have been reviewed accordingly, as fast firing has led to the development of new technologies. Today, an extensive knowledge of raw materials, the nature of the bodies, the composition of engobes and glazes as well as the wealth of acquired experience have resulted in the simplification and standardisation of this phase too. Ever-more complex, sophisticated kilns and equipment now make it possible to achieve a degree of temperature adjustment and control which, even just a few years ago, would have been unthinkable. Kiln widths have increased and combustion systems have been improved, providing constant, uniform temperature over the entire kiln cross-section. The burners and their combustion chambers are normally sized so as to provide high combustion flame speeds and low per-unit thermal potential. In addition to high speed burners, todays technology provides other burners which allow the user to fluctuate the flame over the kiln cross-section and adjust flame power, cyclical flow and duration. This optimises the way in which the heat is distributed over the materials to be fired, especially at critical points of the firing curve. The burner unit generally constituting the combustion chamber may be made either of refractory rammed material or silicon carbide. The latter gives the unit better heat resistance at maximum firing temperatures and also provides greater resistance to thermal shock. Tile conveyance rollers are a key feature on modern kilns. These may be made of ceramic or metal; each kiln zone is fitted with specific roller types suited to that particular stage of the firing cycle. Adoption of rapid single or multi-layer firing has facilitated tile handling, favoured production of large tiles and increased production flexibility. Firing temperatures and cycles generally fall within the 1100-1200 °C and 3560' ranges respectively, depending on the nature of the ceramic materials and the size of the tiles. Fig. 173 shows two firing curves. The first is generally suitable for standard 318

Floor tiles

Single-layer high temperature roller kiln.

white bodies, while the second is designed for bodies sensitive to black core problems. A high degree of flexibility in determining and setting firing curves, not to mention maintenance of firing conditions, is ensured by microprocessor control systems which keep temperatures within very narrow tolerances. Computers provide high levels of kiln automation and also allow the user to monitor and store the process data.

Fig. 173. Typical firing curves for standard vitrified single firing floor tile bodies and those sensitive to black core.

319


Sorting Although sorting does not affect product features, it is an important stage in the production cycle. In this area too, automation is pervasive, covering everything up to boxing and palletizing. Tile geometry (size and flatness) can be checked by electronic and/or camera systems which then send the product to specific outlet points. In highly automated installations the operator need only analyse aesthetic defects and code the tiles to define their class. Handling and storage systems In addition to traditional metal trolley tile handling and storage systems where the trolleys are pushed along tracks by rigid push and transfer bars, handling and storage with automatically guided vehicles is becoming increasingly widespread. In these systems the trolleys are transported by the vehicles along flat trackless surfaces, guided by a floor-embedded wire or laser beam system. In both cases, a computerised control station monitors the entire handling system, effecting real-time monitoring of production flows to and from the tile storage area. Conclusions After twenty years of continuous technological development, todays single fired floor tiles are products of superb technical and aesthetic quality. However, research continues and there is still room for innovation and improvement. Technical and structural tile properties already go far beyond the standards required by the construction industry; tile finish and decor can also successfully compete with anything offered by other natural or artificial materials. Last but not least, ceramic floor tiles remain inexpensive and enjoy excellent market competitiveness.

320


Chapter XI PORCELAIN TILES

Foreword Porcelain tiles (also known as fine porcelain tiles, fully vitrified stoneware or sometimes referred to by the Italian gres porcellanato) have excellent technical characteristics, featuring exceptionally good bending strength and water absorptions often lower than 0.1%. In the context of the ceramic tile industry, porcelain tiles have gained ground rapidly, from a very small share of the market and a limited range of applications to an ever-increasing level of demand and wider fields of use. While, in the past, this product was of interest largely on account of its technical characteristics, its refined aesthetic potential has now ensured its success in more sophisticated market segments, accompanied by a marked increase in output. Yet porcelain tiles are not exactly new. While their origins lie in now-obsolete production technologies, development can be attributed to the introduction of innovative chemical-mineralogical compositions suitable for use with modern technologies such as high pressure moulding, recently developed decoration techniques and, of course, fast firing - even on large tiles. This chapter provides a general overview of the porcelain tile market, technological aspects, the different stages of the production cycle and the outlook for the future. Market Note that in the early 80s just 8 million m2 of porcelain tiles were produced, accounting for just 2% of Italian ceramic tile output. By the year 2000 that figure had jumped to 300 million m2, some 40% of all Italian output. Fig. 174 illustrates this extraordinary escalation. The increase in output can be explained by the fact that these tiles are appreciated not only in Italy and Europe but all over the world (60% of Italian porcelain tiles are exported). Global porcelain tile production capacity is currently estimated at around 700 million m2/year (accurate figures are difficult to obtain as there is a lack of information with regard to China, which has a huge industry, and new producers are appearing all the time the world over). Production involves 265 companies, 100 of which are Italian. The key producer nations are China with about 80 plants, Spain with 22, Taiwan with 20, France and Germany with 7, Malaysia with 7, Thailand, Indonesia and India with 5. Bringing up the rear are Portugal, Poland, the Czech and Slovak Republics, Turkey, South Korea, Japan, the Philippines, Sri Lanka, Argentina, the USA, Venezuela and Morocco, each having just a few or only one production plant. 321


Italian sales

Export

Output

Output

Fig.174. Production and sales data for Italian porcelain tiles from 1991 to 1998 (In 1997 and 1998 glazed porcelain tiles accounted for 24,750,000 and 70,322,000 m2 respectively).

Technical characteristics The word porcelain is well known as a description of fine, prestigious ceramic ware. It is used here to emphasise tiles of high specification having an extremely compact mass of crystal phases in the vitreous matrix. Fig. 175 shows product classification according to ISO 13006. These standards place porcelain tiles in Group BIa, which incorporates vitrified materials characterized by porosity values lower than 0.5% (expressed as water absorption). Actually, industrially produced tiles have a porosity much lower than that required by the standards, generally less than 0.1%. Such tiles are ideal for flooring as they are extremely resistant to all kinds of wear on account of their hardness. 322

Porcelain tiles

Water absorption

Shaping

A extrusion

B dry pressing

GROUP I ≤ 3%

Old EN

GROUP IIA 3% - 6%

Old EN

GROUP IIB 6% - 10%

Old EN

GROUP III > 10%

Old EN

GROUP A1

EN 121

GROUP AIIa1

EN 186/1

GROUP AIIb1

EN 187/1

GROUP AIII

EN 188

GROUP AIIa2

EN 186/2

GROUP AIIb2

EN 187/2

GROUP BIIa

EN 177

GROUP BIIb

EN 178

GROUP BIII

EN 159

GROUP BIa ≤ 0.5%

EN 176

GROUP BIb Bib 0.5% - 3%

Fig. 175. Classification of porcelain tiles according to ISO 13006.

Moreover, they are highly resistant to frost, chemicals and stains and have exceptional bending and compression strength etc. Tab. 15 illustrates their main characteristics, comparing standard-permitted minimum values with real values observed on marketed products. Other aspects which should not be underrated are their antistatic and hygienic properties, making them ideal for computer rooms, hospitals and operating rooms. Commercial specifications Size Over the last ten years some 90% of Italian output has been accounted for by medium size tiles. More specifically, 60% can be attributed to the 30 × 30 cm and 33 × 33 cm sizes, with 40 × 40 cm tiles accounting for just 10%. In addition to these standard sizes, there are other products which range from the very large (e.g. 60 × 120 up to 120 × 180 cm) to the very small (accessories and trims). Moreover, porcelain tile producers have recently begun to offer product ranges which coordinate square and rectangular tiles, especially on rustic product lines. Categories of use The combination of technical and aesthetic aspects determines the suitability of the final product for a specific use. Fig. 176 illustrates different fields of application for different types of glazed 323


FEATURES

STANDARD

WATER ABSORPTION

ISO 10545.3

STANDARD-PRESCRIBED

ACTUAL PRODUCT VALUES

≤ 0,5 %

< 0,1 %


ISO 10545.4

> 35 N/mm (min 32)

> 50 N/mm2

BREAKING STRENGTH*

ISO 10545.4

min 1300 - min 700 N

min 1500 - min 2000 N

a) DEEP ABRASION RESISTANCE***

ISO 10545.6

< 175 mm

< 150 mm2

b) SURFACE ABRASION RESISTANCE**

ISO 10545.7




ISO 10545.8

TEST DISPONIBILE

~ 7 x 10 -6


ISO 10545.9

TEST AVAILABLE

NO CHANGES

MOISTURE EXPANSION

ISO 10545.10

TEST AVAILABLE

TEST AVAILABLE

CRAZING RESISTANCE

ISO 10545.11

REQUIRED

REQUIRED

FROST RESISTANCE

ISO 10545.12

REQUIRED

NO VISIBLE EFFECT


ISO 10545.13

Class GB min

Class GB min


ISO 10545.13

TEST AVAILABLE

SPECIFIED


ISO 10545.14

Class 3 min

Class 3 min

2

2

Pb AND Cd LEACHING

ISO 10545.15

TEST AVAILABLE


COLOUR DIFFERENCE

ISO 10545.16

TEST AVAILABLE


FRICTION COEFFICIENT

ISO 10545.17

TEST AVAILABLE


Standards 10545.1 and 10545.2 are not shown in the table owing to limited space. Nevertheless, the reader is reminded that they refer to sample acceptance methods and size and surface quality characteristics. * Depending on thickness (< 7.5 mm ≥ greater than 7.5 mm) ** To be effected only where used as paving *** Unglazed products For more detailed information see UNI publications (Italy).

Tab. 15. Porcelain tiles according to ISO BIa standards. Standard-required values and actual product values are shown.

Large porcelain tile.

324

Technical characteristics Caratteristiche tecniche

Aesthetical characteristics Requisiti estetici

Porcelain tiles

Stoneware Gres ceramico

Glossy Brillante

Light comm. Medium comm. Comm. residenziale Comm. leggero Ordinary mat. Special mat. mat

Heavy comm. Comm. intensivo

Industrial use Industriale

Mat speciale Klinker

Resistance Resistenza

Aesthetics Estetica

CD.0023

Fig. 176. Surface characteristics related to fields of application. Note that porcelain tiles, previously confined to industrial uses, are now used extensively in the heavy commercial field and have gained a foothold in the medium commercial field.

and unglazed tile, namely: light commercial, medium commercial, heavy commercial and industrial. The y-axis indicates technical and aesthetic values while the x-axis illustrates the increasing bending strength and resistance to wear corresponding to a decrease in aesthetic appearance. Note that porcelain tiles, originally confined to an industrial context, now occupy a significant share of the heavy commercial field and, thanks to their muchimproved aesthetic properties, the medium commercial field too. Research has led to the creation of new, innovative tile types which maintain the above technical characteristics. This has resulted in marked product differentiation and the spread of porcelain tiles into areas traditionally occupied by tiles of high aesthetic value. There follows a general overview of these relatively new commercial types, including those which have been on the market for some time. a) Plain tiles The simplest products from an aesthetic viewpoint. Pastel shades dominate. Polished plain tiles are often used in shopping malls. Obtained from spray-dried powders of uniform colour. b) Granito (salt and pepper) Obtained from a combination of differently coloured spray-dried powders, re325


sulting in the so-called salt and pepper effect. The base body colour is often white or another light colour. c) Marble effect (veined or streaked) tiles Produced using mixtures of coloured powders (spray-dried, or sometimes micronized powders obtained from further grinding of spray-dried powders). These are distributed at random inside the press mould cavity by special feed systems, giving rise to marble-like (or veined) or shaded finishes, often polished to improve the veined effect. d) Macro granito Obtained from spray-dried powder mixes with a 10-50% of a large grain content. The latter are produced via dry or wet regranulation of spray-dried or micronized powders of single or multiple colour. The tile surface background colour is similar to that of granito or marble effect products, thus highlighting the grains to give a natural stone-like appearance. These products normally feature a polished finish.

Example of marble effect porcelain tile.

e) Tiles decorated with soluble salts and glazes Obtained by silk-screen printing or spraying (disc, spray-gun) glazes or solutions containing chromophore salts (Fe, Cr, Co, V, Mn, etc.) onto white or superwhite unfired or biscuit fired bodies. High-grade, innovative aesthetics result. Usually polished. 326

Porcelain tiles

f) Glazed rustic tiles (flashed/screen-printed tiles) Produced with base or recycled bodies and pressed with dies having structured surfaces. Finishing involves the flashing or silk-screen printing of small amounts of glaze and/or soluble salts. Surface brushing may complete the process. Special surface effects mimic the natural ageing caused by time and wear, thus making these tiles competitive with more traditional products like cotto. Fig. 177 illustrates some of these tile types. Their aesthetic properties rustics excepted are generally enhanced by total or partial smoothing and polishing. Ongoing product differentiation has allowed and continues to allow porcelain tiles to penetrate new markets where they have become a viable substitute for natural materials and glazed ceramic tiles. Fig. 178 shows market shares for the above-cited tile types at the end of the 1990s and highlights their short and medium-term growth.

Fig. 177. Various kinds of porcelain tile.

Raw materials for bodies The raw materials used in porcelain tile body compositions can be divided into several mineral groups, each with its own specific function: the clayey raw materials confer plasticity, while complementary non-plastic materials include fluxing minerals or those with a structural function. The former include minerals of illitic-kaolinitic or montmorillonitic origin. These have plastic characteristics that vary as a function of mineralogical structure and particle size distribution. Fluxing minerals include feldspars and feldspathoids, talc, 327


Tinte Plain Unite colours Event Tint Granito Granito Granite Marble effect Variegato Shaded (veined)

Macrograniti Macro granito Macrogranite Rustici Rustics Rustics Decorati Decorated Decorated

0

5

10

15

20

25

30

35

40

% over total production

Fig. 178. Expected development trends for different types of porcelain tile.

eurites, pegmatites, while quartz and generally quartzites the most refractory ones have a structural function. All components must have low concentrations of colouring oxides (e.g. Fe2O3 and TiO2) to prevent contamination of natural body colour. Quantitative body component ratios depend on the mineralogical nature and particle size distribution of the clays and, finally, on the reactivity of the latter with the fluxing minerals. Fig. 179 provides an overview of the chemical and physical characteristics of porcelain tile compositions/products. The PLASTIC CLAY (B) provides the green tile plastic properties needed during pressing and confers the required post-drying bending strength. The CHINA CLAY (A) complements the unfired properties of the plastic clay, yet is also essential in increasing the alumina content of the body. Feldspar (or possibly a small quantity of talc) acts as a flux at standard firing temperatures (1200-1230 °C). Quartz, where it participates in the melting of the feldspars, helps balance viscosity and the vitreous flows; where it does not, it constitutes the base matrix of the crystal phase in the finished product, together with a small quantity of mullite, generated by decomposition of the china-clays.

328

Porcelain tiles

3

1

2 4

1

2

3

4

1 2

3 3

1 2

4 2

Key: A - Semi plastic white clay; B - Plastic white clay; C - Sodium feldspar; D Quartz; E - Talc.

3 1

PHISICAL FEATURES BREKING LOAD BEFORE DRYING

Kg/cm2

DRY BREAKING LOAD AFTER DRYING Kg/cm2

3

2

1

FIRING TO 1100 °C POROSITY % SHRINKAGE % BREAKING LOAD AFTER FIRING

Kg/cm2

A

B

4/6

8/10

10/15

20/30

A

B

10/12

3/6

5/7

4/6

120/150

250/350

Fig. 179 - Chemical and physical characteristics of raw materials used in porcelain tile production. The table indicates the Na feldspar, yet K feldspar or mixes of the two may also be used. The lowest percentages (0.5-1.0%) are shown in the table only as they will not show up on the graph.

Compositions Porcelain tiles have evolved from the material known as chemical stoneware. This was once used to produce small (5 × 5 cm, 10 × 10 cm) tiles, using nowobsolete technologies. The adoption of modern compositions, roller kilns and modern, high power, high precision hydraulic presses have resulted in progressive improvement of technological product characteristics and yielded the benefits of reliability provided by fast firing. Fig. 180 compares a traditional chemical stoneware composition (item 1: firing temperature about 1200-1220 °C and firing cycles of 30-50 hours) and a present329


day fast firing composition (item 2: temperature of 1200-1230 °C and cycles of 5070 min.). The structures of the fired materials are similar in both, but in the slow firing one there is high mullite content and no microporosities (closed or open pores), thus conferring very high stain resistance. Nearly zero porosity and the formation of mullite crystals can mainly be attributed to the extended firing time, which aids sintering and hardening. The porcelain tile composition shown in fig. 180 (item 2) is typical of a so-called base body. For tiles decorated with soluble salts a very white (so-called superwhite) body is preferred to bring out the intensity and nuances of the colours. Other raw materials such as zirconium silicate, anhydrous alumina etc. are thus used to augment whiteness; where these refractory raw materials are introduced they partially replace quartz. To clarify the above concepts Fig. 181 gives a chemical analysis of several base and superwhite bodies as well as a ceramic stoneware body once used with tunnel kilns. The SiO2-Al2O3-(K2O+Na2O) tertiary diagram for these compositions is also given. These values should only be taken as a rough guide as there is a tolerance for each in relation to the type of oxide in question.

D A

1

C B

E D

2

C

A

B

A = CHINA CLAY

35/45 %

B = PLASTIC CLAY

12/18 %

C = FELDSPAR

27/32 %

D = QUARTZ

12/18 %

A = CHINA CLAY

12/18 %

B = PLASTIC CLAY

27/32 %

C = FELDSPAR

42/48 %

D = QUARTZ

5/10 %

E = TALC

0/3 %

CD.0025

Fig. 180. Possible porcelain tile compositions using the raw materials described above. 1) Standard porcelain tile composition for traditional processing cycles (temperature of 1200-1220 °C and cycles of 30-50 hours). 2) Composition for fast cycles (temperature of 1200-1230 °C and cycles of 50-70 minutes). 330

Porcelain tiles

Al 2O

1 4

2 5 3

20

70

80

SiO2

a2O

10 90

10

30

-N

20

O K2

3

40 0

60

30

Body components SiO2 Al2O3 K 2O Na2O CaO + MgO Fe2O3 + TiO2 ZrO2 P.F.

1 65 24 1.5 3.0 0.1 0.3 6.1

2 67 21.0 1.7 4.5 0.8 0.8 4.2

3 71 18 1.8 4.0 0.9 0.9 3.4

4 68 18 1.4 3.5 0.7 0.6 4.4 3.4

5 64 21 2.9 3.9 0.7 1.0 3.1 3.4

CD.0027

Fig. 181. SiO2-Al2O3-(K2O + Na2O) tertiary diagram illustrating the following body compositions: 1. Chemical stoneware (tunnel kiln firing) 2-3. Porcelain tile (base bodies) 4-5. Porcelain tile (superwhite bodies).

a) Spray-dried powders Spray-dried powder particle size distribution is largely the same as with other tile types. The table below shows particle size distribution for a standard body. Size range (µm)

Sieve residue (%)

> 600 600 - 425 425 - 300 300 - 250 250 - 180 180 - 125 < 125

1% 9% 10% 42% 17% 14% 7%

b) Regranulated powders Regranulation consists of increasing grain sizes from 0.1-0.8 mm (spray-dried powders) to 2-8 mm, depending on the specific process: Dry regranulation system: Grains are very compact, they have the same moisture content as the spraydried powders they are made from and they consist of sharp-edged flakes. Grain size is around 2-6 mm and specific weight 1.2-1.4 g/cm3. Wet regranulation system: Grains are rounded and 1-8 mm in diameter and can be plain or multi-coloured depending on the number of coloured slips used. Under certain conditions the grains may have a high moisture content (>12%), making it advisable to dry them before screening and pressing; optimum grain mois331


ture content is near to that of spray-dried powder. Specific weight is usually higher than in spray-dried powders (1.15-1.30 g/cm3). Whether dry or wet compacted, the percentage of grains added to the basic spray-dried powders is generally lower than 25% to prevent the risk of subsequent pressing problems. The machines and techniques used to produce the grains will be described in Plant engineering solutions further on. c) Micronized powders For certain tile types (e.g. marble effect tiles) it may be necessary to use superfine (micronized) powders which generally have a particle size distribution of less than 200 µm. These are obtained by crushing the spray-dried powders even further. Despite the handling and storage problems associated with micronized powders, where mixed with spray-dried powders they give tile finishes of superb aesthetic quality.

Example of porcelain tiles obtained with micronized powder.

332

Porcelain tiles

Auxiliary materials for decoration Soluble salt decoration techniques provide an original finish by using colouring solutions (or silk-screen pastes) applied on both dried and biscuit-fired bodies (although the latter has largely been abandoned). The last few years have seen the ceramic industry channel significant resources into the development and use of new colouring salts. Consequently, know-how has been increased and aesthetic quality and product personalization potential have been significantly improved. These are not simple inorganic salts but organo-metallic complexes, each with a different chemical nature that depends on the relevant metal, ranging from tartrates and oxalates to gluconates. They provide colouring and brilliance on a par with colouring oxides normally used for glaze compositions. The salts are applied by silk-screen units, in disc booths or, more rarely, by spraygun. Combinations of soluble salt application techniques may also be used: the salts are first applied in a paste thickened with the appropriate additives and followed by a disc-applied liquid solution. Materials decorated with soluble salts are partially or fully polished to give shiny mirror-like surfaces. The last few years have also seen rapid diffusion of techniques in which semifinished products (dry-pelleted materials, glaze or body flakes, glaze powder) are used to decorate the product directly on the press. Such techniques are already widespread and are likely to become even more so as semi-finished materials become ever-more refined and the on-press equipment evermore sophisticated. Basic technological parameters Final product features depend not only on the chemical and mineralogical nature of the raw materials, but on the technological parameters used during the production process too. a) Degree of grinding: to aid vitrification and hardening during firing, the ground slip residue on 230 mesh must be very low, in the 0.5 - 1% range (average particle diameters of 15 - 20 microns). Such fineness helps to increase the specific surface area of the particles in the ceramic mass and therefore their firing reactivity. b) Unfired density: the aim of pressing is to achieve the greatest possible green tile density that is compatible with the black core or degassing problems which can arise during firing. Standard moulding pressures (350-450 kg/cm2) give pressed tile densities of 1.95-2.00 g/cm3.

333


With macro granito higher pressing pressures (up to 600 kg/cm2) are needed to compensate for the unavoidable variations in the bulk density of the grains. c) Firing cycle and temperature: it is in this final phase that the results of grinding and pressing are seen. Here, the key parameters are temperature and cycle time: these must be carefully evaluated if the goal of a vitrified material of very low porosity is to be achieved. On average, fast firing of porcelain tiles requires cycles of 50-70 minutes and firing temperatures of about 1200-1230 °C. Note that the above parameters must be evaluated as a whole because the interaction and combined effects of simultaneous variations in one or more of them must be considered as illustrated in Fig. 182. With coloured bodies it is very important to obtain a good vitrification curve balance vis-à-vis the various compositions; added pigments, in fact, can modify the tendency of the body to vitrify and shift the optimum firing point (see Fig. 183), sometimes making it necessary to modify the composition. Fig. 184, instead, shows changes in bending strength, firing shrinkage and water absorption as a function of firing temperature. Finished product properties (virtually zero water absorption, very high bending strength, exceptional deep scratch resistance and excellent resistance to staining) are influenced both by the choice of raw materials and the conditions prevailing at the batching, grinding, pressing, drying and firing stages.

Pr es td

tem

uc

pe

rod

rat

dp

ure

se yg

Fir

sit

ing

en r/c 3

m

Residue % Fig. 182. Diagram showing how the degree of grinding, unfired bulk density and firing temperature influence tile vitrification.

334

Porcelain tiles

Shrinkage (%)

Fig. 183. Vitrification curves for bodies coloured with different pigments or colouring oxides. In this example the balance of the different curves may be regarded as satisfactory.

< >

=

= >

<

Fig. 184. Diagram illustrating changes in bending strength, shrinkage and water absorption for different firing temperatures.

335


Production technology Compared to other kinds of ceramic tile, porcelain tiles involve some technological and engineering aspects that are unique: the slip colouring process, common to all porcelain tile output batching and mixing of the coloured powders needed for the granitos production of grains and their combination with powders of different density for the manufacture of macro granitos distribution of the coloured powders directly in the press cavity (veined products) formulation, application and drying of colouring solutions for decorated products. The block diagram in Fig. 185 gives a general overview of the different stages of the porcelain tile production process, while Fig. 186 illustrates how flows relate to the different tile types. Manufacturers of plain tiles will, of course, have the most straightforward production plant set-ups. Where granitos are manufactured the plant becomes more complex on account of the batching and mixing of differently coloured and/or treated powders. Macrogranitos require further changes to the powder granulation process and also involve handling of separation phenomena which can appear when the grains are mixed with the spray-dried powders. Whilst the diagram shows the application of colour onto unfired tiles it could also be applied to prefired or inertised tiles. Fig. 187 gives an example of a versatile production plant lay out, arranged for all the above products, glazed or decorated; each area identifies a department. The production process Batching Depending on the degree of plant complexity and the type of grinding process (discontinuous or continuous), batching may feature various levels of automation. With discontinuous grinding, batching of raw materials may be carried out using traditional mechanical leverage systems, or more sophisticated systems involving individual weighing hoppers with load cells. Continuous grinding generally features a microprocessor-controlled continuous weighing system. The batched mix is stored in a pre-loading silo, from where it is fed into the mill continuously after mixing of the raw materials and the deflocculant with the aqueous suspension containing re-circulated sieve residue. 336

Porcelain tiles

POWDER STORAGE

UNFIRED STORAGE

Fig. 185. Block diagram showing flows related to the production of different product types.

337


Plain colour

Granito

Veined-shaded

Decorated

Rustics

Macrogranito

Fig. 186. Block diagram showing possible production flows for various porcelain tile types.

In certain cases, producers pre disperse part of the clayey materials (usually the finer, more plastic part) and sometimes the unfired base body scrap too. This gives a clayey suspension that can be added to the mill or sent further downstream where it is mixed with the slip produced by the mill itself. Grinding Continuous mills, generally used to obtain the base slip (white body), are becoming more and more common in porcelain tile production plants too. A series of sieves are installed at the mill outlet. The first of these is designed to separate any coarse solid particles (e.g. grinding media residue) while the others have a more technological function, screening the slip to be spray-dried. The sieving residue is recycled into the mill. Body colouring Base bodies can be coloured by the discontinuous addition of concentrated syrups (the term syrup refers to a coloured slip with a high concentration of pigments). These suspensions are ground in task-specific mills or by continuously batching and mixing the syrups themselves using devices for the measurement and regulation of flow (by volume or mass). 338

Porcelain tiles

2 4 1

6

5 3 CD.0032

Fig. 187. Lay-out for manufacture of porcelain granito, macro granito and decorated tiles. Key: 1. - Body preparation area; 2. - Pressing and drying dept.; 3. - Glazing dept.; 4. - Unfired/fired tile storage; 5. - Firing dept.; 6. - Sorting dept.

Volumetric delivery systems batch concentrated syrups held in tanks (each containing just one colour) and allow producers to adjust the delivery rate according to the desired degree of concentration. Massic batching units, instead, use density measurements to dose liquid semifinished products. Electronic batching control systems ensure accurate setting and monitoring of the mixing process. Spray drying The size of a spray drying department largely depends on the tile type and the number of spray-dried powder colours required. Although spray drying is theoretically the same for all ceramic products, the complexity of the powder mix (dependent on aesthetic and commercial goals) determines the required number of tanks and silos and the number and production potential of the spray driers. Furthermore, where regranulation processes are required there will be a need for extra grain storage silos. Leaving aside micronized powders and grains (used in veined and macro granito respectively), optimum particle size distribution for base bodies and colours used to manufacture granitos is actually very similar to that seen in the manufacture of other types of tile. Regranulation of spray-dried powders In most cases, the semi-finished product used to obtain the agglomerate is the spray-dried powder. The main industrial methods for the production of the granulate are: dry regranulation The spray-dried powders are dry compacted by pelletizers which produce agglomerates that can then be crushed and screened to meet particle size distribution requirements. 339


Storage silos.

wet regranulation Agglomeration is effected by spraying the slips (which may be differently coloured) onto a mass of powder agitated by rotary stirrers. The process simultaneously gives rise to agglomeration and colouring. Batching and mixing of semi-finished products (spray-dried powders - large grains) A key aspect of macro granito production is batching of the powders and grains; batching is effected by belt feeders placed at the outlet of the respective storage silos. As with the raw materials, these devices are equipped with load-cells and encoders to monitor conveyor speed. An electronic batching control system stores the formulas and controls and adjusts delivery rate settings. The batched and mixed powders are then conveyed to the mixed product silos or directly into the press feed hoppers. Pressing Extreme compaction gives rise to a high density tile structure which limits shrinkage and greatly reduces the porosity of the fired product. However, the specific moulding pressure (normally in the order of 350-450 kg/cm2) must be such that 340

Porcelain tiles

the resulting pressed tile density still allows oxidation of organic substances and escape of the gases produced during firing. Another essential factor is homogeneity of powder loading inside the press cavities. The physical nature of some mixtures, such as those consisting of both spraydried powders and grains, can vary substantially, thus making pressing somewhat problematic. To ensure optimum cavity filling press fillers can be fitted with an alveolar floating grating. These devices allow manufacturers to limit in-cavity filling variations, resulting in fewer surface defects (shaded lines, differences in colour tone etc.) and dimensional variations. In recent years the resolution of such geometrical defects has been significantly aided by the introduction of isostatic moulds.

Hydraulic press.

Drying The drying of porcelain tiles does not involve any particular difficulties as drier operating conditions and drying cycles are very similar to those used in the manufacture of other sorts of tile. Where tiles are to be decorated with soluble salts it is essential that they have bending strengths that support the pressure of silk-screen printing (25-30 kg/ cm2). Driers may be vertical or horizontal: from a technological point of view these are essentially the same. Drying cycles in vertical driers generally range from 45 to 65 minutes, depending on tile size and thickness (the bigger and thicker the tile, the longer the cycle). 341


On horizontal driers cycles fall within the 15-30 minute range, once again depending on tile dimensions. Where soluble salts are applied on glazing lines it is essential that the temperature at the drier exit be stabilised so that the salt spreads uniformly through the tile. If this is not so problems of heterogeneous colour tone will probably arise, becoming especially evident after the tile has been polished. Decoration The technology concerning the use of soluble salts is relatively new and sometimes complex. Steady product evolution and increasing demand for ever-more sophisticated tiles has seen the number of special glazing lines equipped for this purpose grow rapidly. Such lines vary in length from 30 to 80 m depending on the complexity of the product, which in the simplest cases need just 1 or 2 silk-screen printings compared to 3 or 4 silk-screen printings plus disc or spray-gun applications in the more complex ones. The key factors affecting the outcome of glazing are: body temperature: affects the diffusion of the saline solution through the tile application type and method: the quantities of colouring salts or glazes to be used largely depend on the application method (disc, silk-screen printing, spray-gun), as each one has its specific characteristics and properties. Note that with silk-screen paste decoration the salt takes much longer to penetrate the body applied quantity: the amount of colouring salt determines the intensity of the colour, as does the quantity of solvent, which affects the depth of impregnation. Drying and stabilization of the impregnated tiles Once the salt has been applied on the unfired or biscuit-fired body, it can be stabilized in a number of ways: by special driers on the production line or naturally in a dedicated area of the department. Firing As already said, firing aims to achieve vitrification (nearly zero water absorption) of the mass and dimensional stability. The most important factors are: reactivity among the body components the degree of slip grinding moulding pressure firing temperature and cycle time. Optimum sintering in completely vitrified tiles depends on the reactivity of the clayey components and the synergic vitrification action developed by the feldspars and the gradually forming vitreous phase. 342

Porcelain tiles

The complex destruction reaction of the clayey lattices, which starts with the vitreous flows and continues until a compact structure is formed, is controlled by the heat energy supplied during firing. Other factors, such as slip grinding and pressed tile density, affect the kinetics of the reaction (i.e. its rate). The heat energy responsible for the inter-component reactions is defined by the firing curve (fig. 188); this is designed to supply a quantity of energy that allows reactions to take place gradually without affecting tile geometry. Firing cycles currently used for porcelain tiles range from just 45 minutes for the smallest, thinnest tiles (e.g. 20 × 20 cm, 7 mm thick) to 90 minutes for the largest, thickest ones (e.g. 60 × 60 cm, 12 mm). Cycles may stretch to 120-160 minutes where outsize tiles in the order of 100 × 200 cm are produced. Maximum firing temperatures range from 1180 to 1240 °C. The exact temperature will depend on body composition, the extent of slip grinding, the compactness of the pressed powders etc. Recently developed roller kilns specially designed for high temperatures employ sophisticated microprocessor control systems which allow manufacturers to maintain accurate control over firing conditions and thus keep temperatures within very narrow tolerances. High power computers and dedicated software let the user display real-time production data, firing conditions and work parameters: where required, these can be printed.

Fig. 188. Example of firing curve for porcelain tiles.

343


Single-layer high temperature roller kiln.

Sorting Although sorting selection and sub-division of tiles according to the markets quality requirements does not affect the characteristics of the finished product it is an important part of the production process. Tile sorting is an essential tool in defining company image: hence sorting criteria must be evaluated with great care. Polishing and squaring Polishing of extremely hard porcelain tile surfaces results in mirror-like finishes of high aesthetic quality. The various stages of industrial polishing may be summed up as follows: a) polishing Tiles are generally wet polished after being sorted and selected by flatness and straightness of edge. The machines consist of a series of stations with oscillating rotary heads containing blocks of abrasive material. The polishing phase is divided into: surface levelling: using diamond-tipped tools, the purpose of this operation is to eliminate any small differences in flatness and even out tile thickness surface grinding: effected with coarse silicon carbide based abrasives. Similar to 344

Porcelain tiles

Tile flatness control device. Finished product sorting line.

surface levelling, this operation makes the tile as flat as possible and eliminates any scratches or surface imperfections polishing: this represents the first stage of finishing and involves progressive reduction of surface roughness through the use of a series of abrasives of decreasing coarseness finishing: the surface is treated with finer and finer abrasives to produce the final mirror-like finish. Once the above process has been completed, squaring and bevelling machines can be used to finish the edges of the tiles. Theoretical polishing line feed rates may be as high as 8 m/min, but in practice work rates are conditioned by tile size and sorting methods: where sorting is manual, speeds are unlikely to exceed 4 m/min. b) squaring and bevelling Specially designed wet milling machines fitted with squaring and bevelling chucks holding diamond-tipped heads are used: up to 1-2 mm of material may be removed. Since edge squaring allows manufacturers to produce tiles of a uniform size it is sometimes done on all the tiles exiting the kiln whether they are to be polished or not. Technical outlook The technical characteristics of porcelain tiles consisting of very hard crystal phases and sintered at high temperature are undoubtedly the very best that ceramics has to offer. Further technical development aimed at improving performance may seem, from a commercial viewpoint, superfluous. From a technical viewpoint, though, it should 345


be pointed out that even a multiphase ceramic material has intrinsic microporosity owing to the impossibility of filling the empty gaps between the solid crystal particles and the surrounding glass phase, even after a long firing cycle. Even porcelain tiles feature such microporosity, albeit at minimum levels: unless controlled carefully, this could lead to problems of surface cleanliness. In general, open porosity values are very low (about 0.1% in terms of water absorption and 0.5% in terms of mercury porosimetry) and so the tile surface tends to resist staining well. Internal closed porosity, instead, is estimated at 6%, with pore sizes ranging from 1 to 10 microns; this porosity is exposed during polishing where about 0.5-1 mm of surface material is removed. There are several ways (some of which have already been widely adopted) to reduce the effects, or the incidence of residual microporosity itself, on product performance. On the question of microporosity, it has been demonstrated that penetration of staining agents and their removal are linked to pore size and diameter. Relatively large open pores will obviously make the product easier to clean, but to the detriment of the visual aspect of the surface. Vice versa, very small pores solve the problem in advance, as they block any intrusion of foreign matter; as part of their product enhancement efforts, manufacturers thus prefer this small pore approach. Appropriate adjustments to work parameters can also provide significant increases in tile density. For example: by increasing the specific surface area of body components, particularly the harder ones (quartz and feldspars). Controlled via the milling process by increasing powder compaction with higher moulding pressure by using more reactive fluxes compatible with the chemical-physical stability of the bodies and their tendency towards pyroplastic deformation. Through application of these and other solutions manufacturers are steadily succeeding in raising porcelain tile quality even higher. Aesthetic outlook In the light of current (2000) output trends, porcelain tile aesthetics will inevitably continue to evolve, widening the range of decorative effects a key factor in the success of the product even further. New products playing a key role in this aesthetic diversification are the alreadycited macro granitos, tiles decorated with soluble salts, veined tiles, structured rustic products decorated with soluble salt flashing and other techniques such as brushing. Then there are those obtained using mixed application techniques (silk-screen printing, pelletized glazes and the use of soluble salts on both structured and flat surfaces). 346

Porcelain tiles

With some product types, polishing provides surfaces that closely resemble marble and natural granite. This ongoing aesthetic progress can only raise the added value of porcelain tiles even further, and is likely to lead to their use on the residential market and as outdoor wall cladding through the application of tiles as large as or even larger than 100 × 200 cm. Conclusions It is logical to assume that advances in decoration will also allow porcelain tiles to capture a significant share of the high quality wall tile market (indoor and outdoor tiles for public buildings, shopping malls etc.). From a technical viewpoint the results already obtained are excellent. Now, together with the evolution of decoration devices (on the press itself or along the production line), the industry is steadily creating a well balanced product capable of competing with natural materials. Further improvements to productivity are likely to be achieved by making organisational changes to body preparation departments, where the production of high-quality, versatile base bodies will probably become more common, allowing manufacturers to make modifications, using additives, to the bodies in the slip tanks or the mills themselves.

347


348


Chapter XII ACCESSORIES AND TRIMS

Accessories and Trims form a vast high added value product category that consists of tiles decorated in second (single firing tiles), third and fourth firing, mosaics, strip and border tiles and other elements that can be matched with both wall and floor tiles. The accessories and trims industry is, then, closely tied to standard tile manufacturing. It consists of companies that actually produce these complementary products and others that perform specific tasks separate from the actual ceramic side of manufacturing (cutting, polishing, on-mesh gluing etc.). Even today, accessories and trims are often referred to as third fired products. This is because, back in the mid-70s when the industry largely consisted of small family concerns, finished tiles were usually hand decorated (mostly with floral motifs); since the bulk of output consisted of double firing wall tiles this hand-painted decoration was subsequently third fired. Over the last twenty years things have changed somewhat: aesthetics have improved immeasurably and output has rocketed. These changes have affected not only decoration (trims included) but also accessories such as strip and skirting tiles.

Some examples of accessories and trims.

349


A continuously expanding market for wall and floor tiles now key elements in furnishing has significantly accelerated growth in the accessories and trims industry, to the extent that some of the original third firing workshops are now modern industrial concerns of key importance. This evolution has also stimulated the development of complementary production processes, which tile manufacturers often prefer to outsource; this has resulted in the appearance of a whole host of small companies specialising in tasks such as porcelain tile polishing, border and insert cutting and mosaic mesh mounting. This new production scenario has not only allowed manufacturers to respond quickly to fast-changing aesthetic demand it also allows them to forecast that demand and even influence future trends. Such results have largely been achieved thanks to effective teamwork between the R&D divisions of the ceramic accessories and trims companies themselves and ceramic colour producers, the former focussing on aesthetics and graphics, the latter on glazes and coloured pigments. A further boost has come from the producers of ceramic manufacturing machinery. Recent developments have allowed technology and automation already tested with standard tiles to be transferred to the accessories and trims sector. The market The accessories and trims market is closely tied to the wall and floor tile market: tile companies, in fact, usually offer customers a line of decorated items (borders, strips, inserts and trims) specially made by external accessory manufacturers to match their own plain tiles. In recent years the main accessory producers have activated direct-to-market distribution, providing the final user with a greatly increased selection of plain and decorated items and thus allowing him to engage in true product personalisation. Market potential is continually on the increase, as accessories and trims consumption is to some extent dictated by fashion a concept once confined to the clothing market. In Italy alone some 120 companies employ over 3000 workers (late 90s), generating total sales of over e 250 million; 9% of companies (11) employ over 100 people and may thus be considered small-medium plants rather than just workshops. Yet attaining reliable output figures for individual items is virtually impossible owing to the vastness of the product range. From a quantitative viewpoint, the accessories and trims market is inevitably linked not only to tile consumption but demand in the construction industry.

350

Accessories and trims

Different ceramic accessory and trim types (from Fashion tile, Gruppo Editoriale Faenza Editrice).

351


A brief summary of the various types of accessory might be: Materials obtained by subsequent working of the finished tile and a 2nd, 3rd or 4th firing. Elements or pieces obtained by pressing them while plastic and/or by extrusion. Elements or trims obtained by dry pressing. Elements or pieces complementary to wall tiling (borders, bullnose tiles, strips etc.). Composites of small pieces sometimes of different materials such as marble or travertine etc. glued onto paper or mesh in the form of particular designs for use as borders or friezes. Trims assembled from several components (e.g. steps, edging). Trims obtained from on-mesh assembly of elements cut with a water-jet (e.g. round or elliptical multi-piece mosaic patterns). Trims obtained with special techniques such as laser engraving and sculpting, chemical attack, lapping, partial polishing etc. At least 50% of output is accounted for by the first item on the list (i.e. tiles produced in 2nd, 3rd and 4th firing). Technology In the mid 70s the accessories and trims industry essentially consisted of small workshops: decoration in the form of hand painting of simple motifs or application of decals was largely manual, with second-grade tiles often being recovered for these purposes. The decorated tiles were then fired in muffle kilns, which had limited holding capacity and relatively long firing cycles. Since then two key innovations have been introduced into the decorating department, both regarding the initial phase of operations: automated silk-screen printers have replaced the hand-held brush while roller kilns have greatly modified factory layouts and increased cycle flexibility. The whole decorating process has thus become very similar to the glazing/firing sequence used to produce the tiles themselves, although manual ability is still important in ensuring accurate application. Tile decorators may no longer need any particular drawing or painting skills, but they must be able to position the printing screens, batch the colours and monitor the process in general. Moreover, the introduction of computerised graphics has revolutionised decoration by increasing the opportunities for artistic expression while drastically reducing the time needed for new product introduction. No important innovations are forecast for the immediate future. 352


Instruments for chromatic graphic settings (from Fashion Tile, Gruppo Editoriale Faenza Editrice).

Today, accessories and trims still involve a degree of craftsmanship, thus precluding the introduction of massive automation (which is not justified by the small production runs). Nevertheless, given the keenly competitive nature of the accessory and trim market and the fact that both machinery and glaze/colour manufacturers dedicate significant resources to it, it seems reasonable to expect further improvements in the future. Technologically, then, the industry has moved on from manual or carbon-dusting techniques followed by hand painting, to the application of decals, ever-more numerous and complex silk-screen printings and the dry application of pigment grains, glazes or glasses. Recent years have also seen the perfection of cutting and assembly techniques for composite trims and huge improvements in polishing. The current trend is towards personalised relief decoration: this effect, attainable via the employment of complex on-press systems, acquires particular aesthetic value in 3rd firing by combining 2-3 silk-screen colour coatings with the application of highly transparent grains (vetrosa).

353


The latter are mixed with glues and starches, the low expansion coefficient and chemical nature of which allow application of very thick layers, thus providing double firing, single firing and porcelain tiles with a substantial degree of relief. Vetrosa are generally layered: a first pressed layer is compacted onto the tile and a second silk-screen layer and a third layer are sprinkled over the entire tile so as to consolidate the decoration. Second firing decoration is also increasingly used on porcelain tiles: silk-screen machines apply powders on the fired tile which then has a fast re-firing cycle (960980 °C). These decorations, often polished, are specially designed to provide not only excellent aesthetic properties but also a high degree of hardness and abrasion resistance. This is achieved by mixing compositions that, when combined with industrial frits, lead to the formation of glass ceramics (partially recrystallised glasses). At this point a few words should be said about lapping, a technology often used to enhance aesthetics. Lapping involves a technique that, even where glazes of a particular hardness and gloss are applied on perfectly flat tiles, allows the user (through use of carefully selected abrasives) to partially polish the glaze without

Products featuring polished relief glaze decoration (from Fashion tile, Gruppo Editoriale Faenza Editrice).

354


Lapped and semi-lapped ceramic accessories (from Fashion tile, Gruppo Editoriale Faenza Editrice).

actually removing any surface layers and therefore without exposing any closed pores in the glaze itself. The above techniques, then, allow manufacturers to produce a vast range of effects that realistically simulate marble and other natural stones using waxed or antique effects. Moreover, aesthetic potential can be heightened by matching differently shaped pieces (that have been cut by water jet) to form intricate compositions. The water jet technique, used extensively in the porcelain tile sector, allows manufacturers to produce myriad geometric forms, from the very simple to the very complex: very high pressure water jets carrying abrasive sands have proved to be highly effective in cutting the hard, compact ceramic while numerical control devices provide pinpoint accuracy. Attractive, highly detailed compositions, pre-assembled on paper or mesh, sometimes using marble or cotto inserts shaped by diamond-tipped tools, can thus be created. The resulting design opportunities are innumerable. Another key development in the decoration of accessories and trims is the Four Colour (or Quadro) system. This is essentially an evolution of silk-screen printing in which four prime colours are overlapped to obtain any other desired colour: used in conjunction with computerised graphics, manufacturers can produce high-defini355


Mesh-mounted trims produced using the water jet technique (from Fashion tile, Gruppo Editoriale Faenza Editrice).

tion photographic quality images with excellent colour separation. Four colour printing requires more precise alignment at the moment of application than with standard silk-screen printing: consequently, the latest machines feature electronic piece alignment. Materials

The decorative materials used with accessories and trims can be split into: Ceramic colours and pigments. Precious metals and lustres. Ceramic glazes (relief and non-relief). Vetrosa grains or glass pastes. Metallic-like colours. Solvents, suspending agents, printing and painting media.

The ceramic colours used to decorate accessories are similar to those used on tiles: they largely consist of frits, ceramic pigments and various inorganic raw materials (mainly oxides) which are mixed and ground to obtain a fine powder (generally < 30 µm). Depending on the decoration technique and the specific requisites of the final, 356


decorated, fired product, one of the following three kinds of ceramic colour will be used: On-glaze colours: applied on the already-glazed, fired tile; this is then re-fired between 700 and 950 °C, the exact temperature depending on the firing cycle which causes the colours to melt and combine with the glaze surface underneath. In-glaze colours: these penetrate the glaze during firing. Used extensively in tableware and porcelain decoration, highly resistant to chemical aggression and abrasion. Under-glaze colours: normally applied beneath a layer of transparent or semitransparent glaze, diffusing into the latter during firing. Ceramic colours are usually supplied in powder form and thus need to be suspended in appropriate liquids so that they can be applied on the tile; such vehicles generally consist of glycols or organic polyglycols that burn away during firing. For porcelain tiles a decoration technique has been developed in which the applied metal organic pigments penetrate the dried unfired or fired tile. Their effects can be further enhanced after firing by polishing. Colours for third firing applications are normally pre-mixed by the supplier, who may also provide the correct fluxes to suit the firing cycles. Where the supplier provides the bases small quantities of pigments can be added to adjust their shade. Third firing decoration often makes use of precious metals and lustres. Lustres are semi-finished items similar to lacquers, made up of organo-metallic compounds with a precious metal base combined with resins that form a film once fired on glass, majolica or porcelain. Lustres produce very thin, intensely coloured films, often no thicker than the wavelength of the incident light. The resulting luminous interference and reflection phenomena produce attractive, highly glossy, iridescent effects. Lustres are classified according to tone, which depends on precious metal content, and the type of effect (e.g. crackled, veined, iridescent, pearl). Metal preparations involve the use of organic compounds with a precious metal base; these are combined with other organo-metallic fluxes to aid adhesion and with resins to produce films. When applied on a smooth substrate they form, after firing, a very thin yet highly reflective metallic film. The resulting colours depend on the combination of precious metals: bright golds (red-yellows consisting of Au), yellow golds (yellow-green, consisting of an Au/Ag alloy), platinums (white gold, Au/Pt alloy) or palladiums (white gold, Au/Pd alloy). Introducing small quantities of matting agent produces a similar effect but with a matt finish. A whole series of intermediate burnished finishes can be obtained by introducing special additives. 357


There are also other products which can be sprinkled onto the piece: these are obtained from gold leaf and precipitated gold powder. They are extensively used in the production of high quality decals. Organisation Accessory and trim production enterprises must be flexible and capable of responding to the needs of their client companies (i.e. tile manufacturers) quickly, as the market often forces the latter to operate on a just-in-time basis. That flexibility has largely been attained thanks to the advent of roller kilns, which, because they are much more compact and have much faster firing cycles than traditional systems, are far better suited to the production of small lots. Flexibility has also been enhanced by the increased number of small sub-suppliers performing so-called intermediate processes (silk-screen printing) or complementary tasks (cutting, mesh mounting, decal application etc.). In certain manufacturing areas these enterprises form an important part of the local economy. Another, equally important factor is logistics, an aspect that has changed considerably over recent years: in an industry in which the handling of materials is crucial, the refinement of logistics management techniques and the appearance of logistics specialists virtually unheard of in ceramics until just a few years ago has proved inevitable. It seems reasonable to assume that, in the near future, increasing market pressures will make logistics a key aspect of every stage of the product life-cycle, from the supply of raw materials to post-sales assistance. Providing a realistic Italy-wide picture of the accessory-trim output situation is virtually impossible. Accessories, trims etc. play an increasingly important auxiliary role in top-of-the-range tiles, and represent a fast-expanding area of strategic importance (the last few years have seen consistent annual growth rates of around 12%). Whats more, there is a whole host of small and very small businesses which are difficult to categorise as they are often small workshops or even family concerns focussing on tasks such as hand painting, silk-screen decoration, cutting, bevelling, gluing and polishing. Accessory manufacturers (who are in direct contact with big industrial clients and thus play a key role in this market) now outsource an increasing amount of work to these enterprises. Trims Trims are ceramic products generally used to enclose or finish a geometrically and spatially complex ceramic lay out. The number of pieces that can be utilised is limited only by the producers im358


agination, as todays highly personalised tile layouts clearly demonstrate; however, producers do need to offer a minimum number of different pieces so as to cope with the majority of laying situations. There are four main types of trim, each reflecting a particular geometric configuration in the laying area and each including a wide range of products specially designed to line such configurations, taking into account the need to cover both internal and external corners. The following table offers only a brief overview of the most common trims (a complete list would run into the hundreds). As a rule factories need to be organised so they can produce about 30 different trim types per tile size.

Several types of trim.

Use of trims is especially widespread in the United States and other AngloSaxon countries; the relative narrowness of the overall tile market dictates rather scant total output. In quantitative terms output can generally be calculated as being 15-18% of total matchable product output. Most of these (50-70% approx.) are flat, rectangular or square, with one rounded 359


MAIN TRIM FAMILIES 1) Skirting complements both terminal : and consecutives:

cove base normal cove

2) 90° angles between two surfaces, EXTERNAL

3) 90° angles between three surfaces, INTERNAL AND EXTERNAL

4) End pieces and countertops

Some commonly used trims.

edge. This explains why, despite the fact that there are over 160 different types of trim, so few of them are produced in numbers that justify the installation of automated machinery. For further information on the terminology, nomenclature and classification of these products the reader should consult the relevant trim production standards (see bibliography). Production line An explanation of how a trims production line should be organised is best begun with the following observations:

360


a) single and double firing trims All trims can be produced using single firing techniques: the choice of technology largely depends on logistical concerns. Note that pieces of a size or geometry that require manual fettling (edge scraping) are easier to handle if fired. Another point in favour of double firing is that it is more suitable for small lots: a large amount of fired material can be produced and then glazed and re-fired when required. Where instead, production is more continuous and product size/geometry is more standardised, single firing is a more logical choice. b) piece handling Another key factor is the conveying system: while most pieces can be conveyed normally along a standard production line there are many others which require the use of refractory supports, racks and plates. However, as the latter generally involve only a relatively small portion of output, an alternative to manual handling especially on the glazing line is difficult to justify. c) porous or vitrified material Beyond the obvious technological aspects, the need to produce porous or vitrified pieces also has important plant engineering implications: for example, a porous trim glazing line will require bell or waterfall systems while vitrified products will require the use of disc booths. In firing increased vitrification will require the use of refractory supports, items that are unnecessary where the same product has higher absorption. The main features of a trim production line are illustrated in the flow sheet in Fig. 189 and the plant layout diagram in Fig. 190. A brief description of the various stages of the production process follows. Body preparation Generally the body is the same as that used for standard tile production (both porous and vitrified products). This does not, however, mean that bodies are always spray dried the same way: with trims spray-drying should aim to produce coarser grains with as low a fine content as possible (maximum 3% at < 200 micron). Moisture content will generally be higher too (up to 6.5-7%). This coarser, wetter spray-dried product makes the powders more plastic, thus reducing the density differences created in the piece during pressing. Pressing Regularly shaped pieces can be produced using the standard single firing process. In this instance the presses are those used for the traditional formats, equipped with normally-raised moulds or shaped dies. 361


More geometrically complex pieces must be produced outside the main line using manual-ejection presses. Keeping the moulds clean can sometimes be a problem on account of the higher moisture content of the powders and the frequent use of fluidisers and release agents (special sprays or even diesel fuel). Moulds are generally lined with rubber.

PRESS e.g. PH 690 whit automatic collection, fettling, brushing

PRESSES e.g. PH 120

Manual fettling and rack loading on refractory plate

Drier

Glazing Manual glazing

Storage on racks

Biscuit storage

Support change and clean

Firing

Firing

Single firing cycle Double firing cycle

Fig. 189. Trim production flow sheet.

362

Chamber drier


1 Feeder 2 - Brusher 3 - Blower 4 - Water spray 5 - Drive unit 6 - Silk-screen printing 7 - Double disc booth 8 - Conveyor 9 - Bell unit 10 - Glaze stirrer 11 - Spray gun 12 - Scraper unit 13 - Turntable 14 - Single disc booth 15 - Aligner + straightener 16 - Row former 17 - Aut. KPT4 line unloader 18 - Aut. KPT5 line unloader 19 - Press 20 - Aut. KPT2 line loader 21 - Tile aligner 22 - Roller pre-kiln drier 23 - Automatic roller unit 24 - FL 29.4/1600 gas kiln 25 - Bar cooler 26 - Modular roller unit 27 - Tunnel drier 28 - Glazing station (optional) 29 - FL 14.7 kiln (optional) 30 - Press with auto. unload. (optional) 31 - Complete cutting line (optional)

FUME TREATMENT UNIT

STORAGE AREA

Fig. 190. Trim production plant layout.

363


Fettling (or edge scraping) Post-pressing brushing and fettling is a delicate stage of the production process. It must be suitable for all types of piece and should be automated where pieces are flat. Scrapers, grinding wheels or shaped brushes that can be changed quickly are normally used. Drying A standard fast single layer drier is generally used. Where geometry allows, the pieces are run through the machine directly on the rollers; where more irregular they are conveyed in special containers or on refractory or metal plates. Where piece thickness varies considerably (with the accompanying risk of excessive stress) a shuttle drier may be more appropriate. Glazing Together with fettling this is the most delicate stage of the process and the list of potential problems is a long one. Most difficulties stem from the need to apply a similar thickness of glaze on differently oriented surfaces. Glazing may be single or double-row and use bell, waterfall or disc application systems. The piece is sometimes tilted (vertically and transversely) when under the applicator to optimise glazing of curved surfaces. As at the press outlet, piece-specific scraping and deburring systems are installed. Where pieces are of particularly complex geometry they are glazed manually on lines characterised by versatility and flexibility. The trims are normally placed on trays or directly on the conveyor belts. In any case they must arrive at the glazing zone under pre set temperature conditions so as to limit variations in shade. Glazes should generally be of high thixotropy so that they freeze as soon as they come into contact with the piece. This reduces the risk of bare patches or glaze build-up on non-horizontal surfaces: it also helps stop lumps of glaze forming at the edges of inclined surfaces, an inconvenience that requires fettling. Storage Given the highly discontinuous nature of production a loading-unloading storage system should be installed at both the drier and glazing line outlets.

364


Firing Regularly shaped pieces may be fired directly on the rollers using the same firing cycles employed for standard production. Others of more complex geometry will require separate piece-specific firing sessions. Kiln unloading and sorting are normally done manually, with a first quality yield comparable to that of standard production for simple pieces and about 1015% less for more complex items. Machines While the manufacture of accessories and trims is still, to some extent, a craftlike operation, industrial development has been rapid over recent years. Although the introduction of modern technologies and materials have yielded products with innovative aesthetic characteristics, this situation has obviously had (and has) an effect on the originality and hand made nature of the individual pieces. The range of accessory and trim products is vast: some may be porous, others partially or totally vitrified (e.g. porcelain tile accessories) and they may be produced using different technologies (e.g. double or single firing). A perhaps even more important aspect concerns the configuration of the piece (i.e. tile, border, bullnose tiles, pencil tiles etc.). Such a heterogeneous range of products, then, clearly entails a wide range of plant engineering solutions, an exhaustive description of which is beyond the scope of this book. The very nature of the materials (e.g. pieces pressed in a semi-plastic state to obtain relief effects) calls for a high percentage of manual work and working conditions are very different from those adopted for other, more standardised ceramic accessories. However, increased automation and, consequently, improved handling, has paved the way for innovative plant layouts that reduce the need for manual labour. The end result has been a huge fall in production costs. The advent of fast firing has radically modified third firing (i.e. ceramic accessory) production criteria, yet it should be pointed out that its adoption was neither immediate nor widespread and there are still items for which it cannot be used (the appropriateness of fast firing ultimately depends on the configuration and nature of the body). The degree of sophistication of a third firing plant, then, essentially depends on the characteristics and the individuality of the pieces. Some manufacturing operations require hybrid layouts that feature a combination of both industrial and craft-like aspects (i.e. a high degree of automation and a high degree of manual work). 365


The flow diagram in Fig. 191 shows a typical production sequence for third firing products. It is only intended as a rough guide in that it refers to one specific product. For example, flow charts illustrating the production scenario for bullnose tiles, pencil tiles or pieces pressed in a semi-plastic state would be very different. As illustrated in Fig. 190, the tiles, after being fired and unloaded from the kiln, are stored on special racks (should further silk-screen applications or manual decoration be required) and then re-introduced into the production cycle and refired according to parameters compatible with the new decoration. Fig. 192 shows a typical layout for the manufacture of tiles with third firing technology. The production process The following section only deals with those processes peculiar to second, third and fourth firing technology, leaving aside more global matters such as body grinding, pressing, drying glazing and firing. Decoration While tile decoration in second, third and fourth firing processes can be done manually, producers generally use, where possible, silk-screen printing machines: their level of sophistication depends on the configuration of the piece and the overall number of pieces to be decorated. Manual decoration and silk-screen printing may be complemented by the application of decals, for which a wide range of automatic machines is available. Whatever the decorative task, it is preferable to brush the piece before starting. The market offers a wide range of silk-screen machines, from traditional models to more modern ones that allow printing to be effected parallel or perpendicular to the conveyor. Together, the above machines form decorating lines of varying complexity; furthermore, the application of electronic logic can provide producers with accurate line feed systems. Relief effects can be obtained on tiles and multi-tile compositions with dryground grains (Fig. 193) for which the market offers a wide range of application machines. Production lines may also need to be equipped with compensers. Depending on how much space is available for the various silk-screen printing/glazing stages, the compenser may be equipped with an autonomous forceddraught heating system. Decorated pieces are generally stored on automatic rack systems which can 366


THIRD FIRING FLOW CHART Fired product storage

Raw material storage

Line loader

Glaze formulation

Brushers and blowers

Glaze weighing and mixing

Water sprays

Grinding

Disc-type glaze applicator

Refining and screening

Fixative spray

Lab tests

1st silk-screen printing

Finished glaze storage

Spray gun fixing 2nd silk-screen printing

Storage and control of finished screens

Fixative spray

Screen preparation

3rd silk-screen printing Mobile spray gun Line unloader Glazed product storage Manual finish (where required)

Manual silk-screen printing

Kiln loading

Manual sponge decoration

Firing

Manual brush decoration

Sorting Quality control Packaging and palletization Finished product shipment

Fig. 191. Flow diagram illustrating tile production with third firing technology (standard production).

367


Fig. 192. Generalised layout for production of tiles with third firing technology.

368


have drive systems consisting of two inverter-controlled mechanical transfer arms. The production line may be made up of a complete plant for closed-circuit silk-screen printing (Fig. 194), itself consisting of a loader-unloader and racks connected to the decorating machines by a conveyor. Firing Except for very thick pieces and trims with variable thickness, tiles decorated in second, third and fourth firing are fired in single-layer kilns with cycles appropriate to the particular decoration (i.e. gold, platinum or glass grains). A complex piece could therefore require several different cycles. These successive firing cycles, even at lower temperatures, may alter the product shade. It is therefore essential to inspect the tile for any such changes before each successive firing: should this problem arise testing will be required to find more suitable firing temperatures and cycles.

Fig. 193. Screen printing machines for the application of dry ground grains (Kemac).

369


Fig. 194. Closed-circuit silk-screen printing machine for the production of trims in third firing (Kemac).

Sorting Tiles are normally sorted at the kiln exit or stored on racks and sorted later. These products are normally sorted manually. Recently, semi-automatic sorting systems have been introduced. The choice of manual or semi-automatic sorting system will depend on the type and concept of production plant logic and, above all, on the type of material being produced. Cutting the material Given the vast range of ceramic accessories and the relative production difficulties, the spread of cutting technology already used extensively with stone products has been rapid. Cutting machines allow manufacturers to reduce standard tiles to a whole range of sub-sizes which can then be used as borders or strips, cornices, mosaic elements etc. 370


These machines cut the ceramic material continuously and generally have tile conveyance mechanisms featuring variable-speed drives. They also feature vertical and transverse adjustment and allow fast change-over of the diamond-edged cutting wheels. Ceramic accessories and trims and the machines used to produce them play a key role in todays ceramic industry. An exhaustive analysis of such a vast argument is beyond the scope of this book and this chapter, like the preceding ones on wall, floor and porcelain tiles, is intended only as a basic introduction.

371


372

Floor tiles

Appendix 1 STANDARDS

In July 2000 Italy adopted unified international ceramic tile standards, as drawn up by ISO between August 1985 and July 1992. To a large extent, these reflect the contents of the widely applied UNI-EN standards, yet it should be observed that substantial modifications have been made to ceramic product classification and test methods. Furthermore, new technological parameters have been introduced. With regard to tile classification, the most important innovation is the sub-division of the first water absorption class (≤ 3 %) for pressed products into two subgroups: the first of these is BIa, which includes tiles with water absorption of ≤ 0.5%. This new system allows for easier, more accurate product classification.

Technological tests have been modified as follows: NEW TESTS: Friction or attrition coefficient Impact resistance (resilience) Breaking load Colour tone Pb and Cd leaching on glazed products.

VARIATIONS: Moisture expansion and stain resistance (glazed and unglazed products) Agents for resistance to chemical attacks at low and high concentration Elimination of the Mohs hardness test (ex EN 101).

A complete version of the new ISO standards would amount to more than 100 pages. Owing to limited space and commercial copyright reasons there follows just a brief summary of the various tile classes, the test methods and a table comparing the old UNI-EN standards.

373


ISO 13006 PRODUCTS and main characteristics GROUP A I (ex EN 121) Extruded tiles with WA ≤ 3 % MOR increased from > 20 to 23 N/mm2 Deep abrasion reduced from 300 to 275 mm3 max. GROUP A IIa part 1 (ex EN 186-1) Extruded tiles with 3 < WA ≤ 6 % GROUP A IIa part 2 (ex EN 186-2) Extruded tiles with 3 < WA ≤ 6 % MOR increased from > 10 to 13 N/mm2 GROUP A IIb part 1 (ex EN 187-1) Extruded tiles with 6 < WA ≤ 10 % GROUP A IIb part 2 (ex EN 187-2) Extruded tiles with 6 < WA ≤ 10 % MOR increased from < 8 to 9 N/mm2 GROUP A III (ex EN 188) Extruded tiles with WA > 10 % ______________________________________________ GROUP B Ia (ex EN 176) Pressed tiles with WA ≤ 0.5 % MOR increased from 27 to 35 N/mm2 Deep abrasion reduced from 205 to 175 mm3 max. GROUP B Ib (ex EN 176) Pressed tiles with 0.5 < WA ≤ 3 % MOR increased from <27 to 30 N/mm2 Deep abrasion reduced from 205 to 175 mm3 max. GROUP B IIa (ex EN 177) Pressed tiles with 3 < WA ≤ 6 % GROUP B IIb (ex EN 178) Pressed tiles with 6 < WA ≤ 10 % GROUP B III (ex EN 159) Pressed tiles with WA > 10 %

374

Standards

ISO 10545 TEST METHODS ISO 10545.1 Sampling and acceptance criteria

ex EN 163

ISO 10545.2 Surface size and quality

ex EN 98

ISO 10545.3 Water absorption and apparent porosity, relative apparent density

ex EN 99

ISO 10545.4 MOR and breaking load

New

ISO 10545.5 Impact resistance via measurement of coefficient of restitution

New

ISO 10545.6 Deep abrasion resistance (unglazed tiles)

ex EN 102

ISO 10545.7 Surface abrasion resistance (glazed tiles)

ex EN 154

ISO 10545.8 Coefficient of linear thermal expansion

ex EN 103

ISO 10545.9 Resistance to thermal shock

ex EN 104

ISO 10545.10 Moisture expansion

ex EN 155

ISO 10545.11 Crazing resistance (glazed tiles)

ex EN 105

ISO 10545.12 Frost resistance

ex EN 202

ISO 10545.13 Resistance to household chemicals

ex EN 106,122

ISO 10545.14 Stain resistance

ex EN 106

ISO 10545.15 Leaching of Cadmium and Lead from glazed tiles

New

ISO 10545.16 Colour differences

New

ISO 10545.17 Friction (attrition) coefficient

New

375


Brief comparison of test methods TEST

Number (EN)

EN STANDARD

Part (ISO)

ISO 10545 standard

Dimensions and surface quality

98

same as ISO

2

same as EN

Water absorption

99

boiling method only

3

uses boiling method to assign group (as EN) but also measurement of porosities (this is carried out under vacuum)

Bending strength

100

measurement of F and σ

4

measurement of F and (the s in EN) and measurement of a load that takes into account surface area of piece (defined as Σ).

Mohs hardness

101

1 – 10 on Mohs scale

inexistent

inexistent

Deep abrasion resistance

102

Meas. of lengths and volumes erased

6

same as EN

Linear thermal expansion

103

same as ISO

8

same as EN

Thermal shock

104

test with and without immersion in water (temp. 15 and 105 °C)

9

test with and without immersion in water, heating temp. changed to 145 °C.

Crazing resistance

105

1 hour to reach 5 Atm and 1 hour holding

11

1 hour to reach 5 Atm and 2 hours holding

Chemical resistance (unglazed)

106

immersion solutions and times different from ISO

13

immersion solutions and test times different from EN

Chemical resistance (unglazed)

122

chemical solutions and times different from ISO

13

Harshness of attack depends on utilisation and degree of resistance offered by material

Abrasion resistance (glazed)

154

classes from 1 to 4

7

classes from 1 to 5 with stain test compulsory for class 5

Moisture expansion with boiling water

155

Same as ISO

10

same as EN

Frost resistance

202

50 cycles from +15 °C to –15 °C

12

100 cycles from +5 °C to –5 °C. cycles are faster and tile wetting method is different

Impact resistance

-

-

5

coefficient of restitution of the material is measured and impact damage is evaluated by visual inspection

Stain resistance

122

K permanganate and methylene blue for 24 hours on glazed products: cleaning with water and neutral detergent

14

chrome green in light oil, iodine in alcohol solution, olive oil for 24 hrs on glazed and unglazed products: gradually more vigorous 5-stage cleaning.

376

Standards

Pb and Cd leaching

-

-

15

As per loss test for tableware

Measuring colour differences

-

-

16

effected with colorimeter (as per ASTM method)

Measuring friction coefficient

-

-

17

- measurement of angle of slippage (as per DIN 51130) - dynamic attrition measurement (as per BCRA) - measurement of static attrition (as per ASTM but with different weights and sliding elements)

Material classification

87

- 4 categories in Group B (dust pressed tiles) - 6 categories in Group A (extruded tiles)

13006

- 5 categories in Group B (dust pressed tiles) (group BIa with AA >0.5% in introduced) - 6 categories in Group A (extruded tiles)

377


378


Appendix 2

Table 1. CLASSIFICATION OF PRESSED MANUFACTS

TABLES AND FIGURES (source: Italy)

379

Table 2. PERIODIC TABLE

Atomic weight

Chemical symbol

Atomic number

INERT GASES


380

Tables and figures

Table 3. ATOMIC WEIGHTS TABLE Based on carbon -12

381


Table 4. WRITING AND PRONUNCIATION

382

Tables and figures

Table 5. MATHEMATICAL TABLES Squares, cubes, roots, inverses, logarithms, perimeter and circles area

383


Table 6. MATHEMATICAL TABLES Numerical common factors

384

Tables and figures

Table 7. MATHEMATICAL TABLES Plane figures area

385


Table 8. MATHEMATICAL TABLES Conversion factors

386

Tables and figures

Table 9. MATHEMATICAL TABLES Conversion factors

387


Table 10. MATHEMATICAL TABLES Physical quantities and units of measure

388

Tables and figures

Table 11. MATHEMATICAL TABLES Si-Units of the international system

389


Table 12. MATHEMATICAL TABLES Metering units conversion factors

390

Table 13/1. TABLE OF MATERIALS MOST COMMONLY EMPLOYED IN CERAMIC INDUSTRY

Tables and figures

391

Table 13/2.


392

Table 13/3.

Tables and figures

393

Table 13/4.


394

Table 13/5.

Tables and figures

395

Table 13/6.


396

Table 13/7.

Tables and figures

397

Table 13/8.


398

Table 13/9.

Tables and figures

399

Table 13/10.


400

Tables and figures

Table 14. SOLID BODIES SPECIFIC WEIGHTS (in kg/dm3)

401


Table 15. SPECIFIC WEIGHTS, ANGLES AND FRICTION COEFFICIENT OF SOME INCOHERENT MATERIALS

402

Tables and figures

Table 16. LINEAR EXPANSION ON 1 M-LENGTH FOR HEATING FROM 0° TO 100 °C

403


Table 17. COEFFICIENT OF CUBIC EXPANSION RELEVANT TO DIFFERENT SOLIDS

404

Tables and figures

Table 18. MELTING POINTS OF SOME COMPOUNDS AND MINERALS

Table 19. HARDNESS COMPARATION TABLE

405


Table 20. CHEMICAL-PHISICAL PROPERTIES OF CERAMIC PRODUCTS SUITABLE FOR MILLS (LININGS AND GRINDING MEDIA)

optimum optimum

optimum optimum

406

optimum optimum

optimum optimum

Tables and figures

Table 21. LOAD FORMULAE OF SACMI MADE MILLS

A) BODY 1) For table 22A and 22D use the following formula: Kg dry material to charge = 0.55 × Vu × d × y/100 2) For tables 22B and 22C use the following: Kg dry material to charge = 0.67 × Vu × d × y/100 where: Vu = working volume of the mill (with lining) in litres d = slip density y = % dry matter in the slip 0.55 – 0.67 = mill load coefficient with silica or alubit grinding media. B) GLAZE 1) For porcelain grinding media use the following formula: Kg dry glaze to charge = 0.345 × Vu × d × y/100 2) For table 22C use instead: Kg dry glaze to charge = 0.455 × Vu × d × y/100 where: Vu = working volume of the mill (with lining) in litres d = glaze density at the outlet (generally l.7) in kg/l y = % dry in the glaze (generally 67%) 0.455 = mill load coefficient with alubit grinding media.

407

Table 22b. MILL LOADING Lining: Rubber - Grinding media: Alubit

Table 22a. MILL LOADING Lining: Silica - Grinding media: Silica


408

Table 22d. MILL LOADING Lining: Rubber - Grinding media: Silica

Table 22c. MILL LOADING Lining: Alubit - Grinding media: Alubit

Tables and figures

409

1.5-2.5

0.6-1

2.5-3.5

B) Red body single firing vitrified tile

C) Porcelain tile

-

0.8-1.6

2-4

Lin. ALUBIT Pebbles SILICE

2.5-3.5

0.6-1.2

1.5-3

Lin. RUBBER Pebbles ALUBIT

3-4

1.2-1.6

3-4

Lin. RUBBER Pebbles SILICE

Specific consumption in dry kg/ton of ground body Lin. ALUBIT Pebbles ALUBIT

A) Monoporosa, double firing, white body single fired vitrified tile

Product type

Table 22e.


410

Tables and figures

Table 23. VISCOSITY CONVERSION TABLE

411


Table 24. TABLE OF COMPARISON BETWEEN THE INTERNATIONAL STANDARD OF THE SCREENS

412

Tables and figures

Table 25. SPECIFIC WEIGHT - CONTENT OF WATER - CONTENT OF SOLID MATTER

413


Table 26. EQUIVALENCES BETWEEN BAUMÈ DEGREES AND SPECIFIC WEIGHTS OF HEAVIER THAN WATER LIQUIDS

414

Tables and figures

Table 27. MAIN COMBUSTION REACTIONS

415


Table 28. SEGER CONES

416

Tables and figures

Table 29. WATER HARDNESS

417


Table 30. SURFACE TENSION OF FRITS AND GLAZES

418

Tables and figures

Table 31. COEFFICIENTS OF VISCOSITY

419


Table 32. OPACIFIERS AND SUBSTANCES WICH CONCUR TO OPACIFICATION

420

Tables and figures

Table 33. NATURAL DYES: METAL SALTS OR OXIDES

421


Table 34. ARTIFICIAL AND SYNTHETIC PIGMENTS

Red Si-Zr-Fe

422

Tables and figures

Table 35a. SILK SCREEN PRINTING FABRIC SPECIFICATIONS Nylon fabric

423


Segue Table 35a.

424

Tables and figures

Table 35b. SILK SCREEN PRINTING FABRIC SPECIFICATIONS Polyester fabric

425


Segue Table 35b.

426

Tables and figures

Table 35c. SILK SCREEN PRINTING FABRIC SPECIFICATIONS Steel fabrics

427


The tertiary diagrams and graphs on the following pages illustrate the reactions between the various oxides and compounds that take place during firing. The graphs are numbered 1-19 to make consultation easier.

Fig. 1.

Fig. 2.

428

Tables and figures

Fig. 3.

429


Fig. 4.

430

Tables and figures

Fig. 5.

Fig. 6.

431


Fig. 7.

432

Tables and figures

Fig. 8.

Fig. 9.

433


Fig. 10.

Fig. 11.

434

Tables and figures

Fig. 12.

435


Fig. 13.

436

Tables and figures

Fig. 14.

Fig. 15.

437


Fig. 16.

438

Tables and figures

Fig. 17.

439


Fig. 18.

440

Tables and figures

Fig. 19.

441


442

Standards

BIBLIOGRAPHY

For a greater understanding of ceramics and more detailed information the reader is referred to the following works: Kingery, Bowen e Uhlmann Introduction to Ceramics Wiley-Interscience (1975) Grimshaw The Chemistry and Physics of Clays and Other related Materials Benn Ltd (1981) Singer, Singer Industrial Ceramics Chapmann & Hall (1975) Jouenne Traité de Céramiques et Matériaux Minéraux Editions Septima (1975) Sigg Les Produits de Terre Cuite Editions Septima (1991) Handle Brick and Tile Making Bauverlag (1992) Parmelee Ceramic Glazes Cahners Book (1978) Hench Characterization of Ceramics Dekker Publ. (1985) Materie prime per fritte e smalti Centro Ceramico Bologna Tozzi Smalti Ceramici Gruppo Editoriale Faenza Editrice (1992) Fiori, Fabbri, Ravaglioli Materie prime ceramiche Gruppo Editoriale Faenza Editrice (1989) 3 voll. Peco I Prodotti Ceramici Marzorati (1991) 2 voll. 443


Emiliani Tecnologia Ceramica Gruppo Editoriale Faenza Editrice (1998) Galassi e Pozzi La Reologia dei Materiali Ceramici Tradizionali Gruppo Editoriale Faenza Editrice (1994) Soc. Ceramica Italiana Reologia Ceramica Applicata Distr. da Faenza Ed. (1984) Soc. Ceramica Italiana Il reparto Controllo e Sviluppo Distr. da Faenza Ed. (1982) S. Caillerre S. Henin Minéralogie des argiles Masson & Cie

444

Introduction

445

Printed September 2002 by Tipografia Moderna of Ravenna for Editrice La Mandragora of Imola

Sacmi Vol 1 Inglese - II Edizione

Recommend Documents