Industrial Ceramics

TECHNICAL NOTE
gave excellent results for many years, but now a day the problem is showing up again in different circumstances. Fast firing is now the dominating technology and the time available to reach a optimal stabilisation of the body is reduced to few minutes. The size of the tiles is growing quickly: now the market is asking for wall tiles up to one meter in size and above. Thickness is down to the minimum, to save in cost and transportation. The value of the product lays in the richness of the surface and the thickness of the glazes is rising. Geometrical perfection is a must: tiles are mechanically squared to give a marble-like look to the wall. The big problem now is planarity. The old trick of increasing the state of compression of the glaze onto the body to avoid delayed crazing was working well on the old thick and small tiles. Now the same level of compression on the very same body and glaze but with a much larger size, has a nasty side effect: bending. To reduce the bending there is only one solution: reduce to the minimum the amount of compression. A second deflect could appear in case of too high compressive tension in glaze: peeling. In some cases the tendency of the glaze to peel off the body appear in the sharp corner of three-dimensional products, such as kitchenware. The problem, now, is to know how much it is the minimum amount of compression. Since all glazes, like glasses, are characterized by a very low resistance to tractions stress, a little value of traction may causes their rupture. If the body expands, even if in a small percentage, the glaze may easily crack, unless it is in a state of compression with respect to the body. In this case, a little expansion of the body reduces the state of compression, without generating a dangerous traction. This tendency to react with water in the course of time cannot be completely eliminated: all glazed bodies show this problem. The only type of body that can be considered stable during the time is the completely sintered body, characterized by a lack in porosities through which the humidity would penetrate into the body. A body formulation and a firing cycle intended to avoid delayed crazing do not exist. The problem, however, can be successfully tackled by studying the state of tension between glaze and body, by means of thermal expansion and bending tests. The state of tension between glaze and body depends essentially on two factors: the relation between their coefficients of thermal expansion and their coupling temperature. Generally, a common error consists of performing a simple comparison between coefficients of thermal expansion in order to prevent the defect. However, the fact that the glaze coefficient of thermal expansion (CTE) at 400C is lower

THE OPTICAL FLEXIMETER TO STUDY DEFORMATIONS ON CERAMICS

M. Paganelli, D. Sighinolfi
Expert System Solutions, Modena, Italy

Generally, the deformations of ceramic materials during and after firing may have a complex origin. If the products are made up of a single material, such deformations are mainly due to pyroplastic phenomena. In the case of glazed materials, two further factors must be considered: the state of tension established between glaze and body, and their differences in behaviour during sintering. 1. New techniques, old problems During firing, a glaze applied on a ceramic body undergoes some transformation which can be simply listed as: losing of the forming water of the clay components, glass transition and softening. In correspondence of this point, the glass part of the glaze start to melt, giving as a result a continuous liquid layer. During cooling, the glaze viscosity increases until the glaze becomes rigid and it starts to contract simultaneously with the ceramic body. Delayed crazing is due to the volume increase of the body, because of absorption of humidity after firing. The body, in fact, can show a slow but inexorable tendency to react with the water present as humidity in the air, which changes permanently its properties. This dimensional increasing of the ceramic bodies occurs in a period of time which can vary from some days to some years. In the past, this phenomena was not understood and it was indeed the biggest problem of porous ceramic bodies. During the last 50 years it was deeply investigated and it is now completely understood. The solution looks easy at first sight: minimise the moisture expansion and increase the compression of the glaze on the body. This approach indeed

Industrial Ceramics Vol. 29 1/2009

 Industrial Ceramics

than the body CTE at the same temperature cannot assure that the glaze will be in a state of compression with respect to the body after firing. Also a glaze with a CTE lower than the body one can give rise to delayed crazing, as it will be shown later. In fact, another important parameter must be considered: it is the coupling temperature between glaze and body. Until now, the instrument used to measure the coupling temperature was the Steger tensiometer and the analysis was quite problematical. The test was performed with very long (30 cm) specimen which was not placed completely inside the kiln, with consequent lack in homogeneity in sample temperature. 2. A novel optical technique Now, thanks to MISURA FLEX-ODLT (shown in Figure 1), the new instrument developed by Expert System Solutions,

it is possible to measure the state of tension of ceramic glazed bodies. With this instrument (which combines together Optical Fleximeter and Optical Dilatometer), it is possible to perform both thermal expansion measurements and Steger analysis, which allows to identify the coupling temperature. The same instrument can be used to study the problem related to the opposite situation, i.e. when there is a too high compressive state of the glaze with respect to the body. This instrument works with no contact with the samples tested: thus, their behaviour is never modified by the measuring system. The specimens can be cut directly from an industrial tile; then there are placed into the kiln, which is equipped with an automatic temperature controller and can reach a heating rate of 30 C per minute and a maximum temperature of 1600 C. All the experimental part of this work was performed by using this optical instrument.

Figure 1. The combined instrument MISURA FLEX-ODLT of Expert System Solutions

Industrial Ceramics Vol. 29 1/2009

Figure 2. Thermal expansion curves of body and glaze

TECHNICAL NOTE
3. Delayed crazing Starting from a glazed product, we want to study the state of tension between glaze and ceramic body. The first experimental step consists of two thermal expansion tests, performed on fired glaze and body (see Figure 2). Two beams of lights illuminate both the ends of a specimen 50 mm long placed horizontally into the furnace and two digital cameras capture the images of the last two hundreds microns of each tip. The specimen, completely free to expand or contract, is measured by the image that it projects on an image sensor. With a wavelength of 478 nanometres, a resolution of 0,5 microns can be obtained. The second experimental phase consists of a flexion analysis performed on a fired glazed body (Figure 3). Two holding rods, 70 mm spaced, support an 85 mm long sample. A camera frames the centre of the sample, which moves downward or upward during the test. The test is performed heating the sample up to a temperature high enough to cause the glaze softening and following the cooling below to about 300 C. In Figure 3, the flexion curve of a glazed tile is expressed as a function of the temperature. At the beginning of the cooling phase, the glaze is liquid and follows the body contraction without developing tensions. At a certain moment, a rapid variation of inclination occurs: this mean the glaze has become rigid enough to build up tensions, causing the flexion of the sample. In correspondence of the coupling temperature, the glaze softens during heating (absorbing tensions) and solidifies during cooling (building up tensions). In this case the coupling temperature is 720 C. In order to obtain the final result, its necessary to translate the glaze thermal expansion curve so that it coincides with the body thermal expansion curve in correspondence of the coupling temperature. The result of the translation is reported in Figure 4. The two curves, after the translation, do not coincide anymore at the origin (room temperature). This difference is indeed the of traction or compression which has established between glaze and body immediately after firing. From this moment, the body start to react with the air humidity , increasing its volume. This is a hydrothermal phenomenon, which is favoured by the presence of humidity and high temperatures. This process of rehydration, in the case of tiles already placed on the floor, may require some years, but can be also artificially accelerate by increasing the water vapour pressure and the temperature. Performing an autoclave test at high temperature and pressure, the process can be completed in few hours. The measurement of the percentage of expansion due to the adsorption of water should be compared with the measurement of the percentage of compression of the glaze layer with respect to the body. If the value of expansion after the autoclave test is lower than the level of compression established between glaze and body, then the product will not be at risk of delayed crazing. The following example (Figure 5) shows the thermal

Figure 3. Flexion curve of a fired glazed tile

Industrial Ceramics Vol. 29 1/2009

 Industrial Ceramics

Figure 4. Thermal expansion and flexion curves overlapped to identify the state of tension of the glaze with respect to the body after firing

Industrial Ceramics Vol. 29 1/2009

Figure 5. Although the glaze CTE is lower with respect to the body CTE, the glaze at room temperature will be in a state of traction

expansion and flexion curves of a product which is strongly at risk of crazing. A simple comparison between coefficients of thermal expansion is misleading: the glaze CTE at 400 C

is 6.3810-6 C-1, while the body at the same temperature is 6.5810-6 C-1. Although the glaze CTE is lower than the body CTE, the glaze is in a state of traction with respect to the body

TECHNICAL NOTE

Figure 6. Pyroplasticity test on two porcelain stoneware samples

during the whole cooling phase. This is due to the fact that the coupling temperature is considerably higher than the glass transition temperature. The glaze is in a state of traction which respect to the glaze that can provoke its rupture, even immediately after firing. Considering the various experimental results it is easy to understand that it very difficult to solve the various problems of deformations without the support of complete laboratory data. The combined instrument MISURA FLEX-ODLT was designed to meet this requirement. 4. Pyroplasticity Another parameter of great technological importance is pyroplasticity, in other words the tendency of the body to deform within a given temperature range. This parameter affects the productivity of a production plant because it determines the relationship between the thickness and the rate of deformation of the material as a function of firing temperature. It is now possible to monitor changes in pyroplasticity during the thermal treatment by measuring the bending of a body sample suspended between two supports 70 mm apart. This measurement is particularly important for bodies that have to be completely sintered because during the final stage of firing they develop an abundant vitreous phase with a sufficiently low viscosity to cause rapid deformation of the material. The stress suffered by the sample inside the instrument is certainly much greater

than that suffered by the material inside the roller kiln. As the material moves over the rollers, the supporting points shift continuously across the entire width of the piece, where as in the laboratory instrument the material is suspended on two static supports. However, unfired ceramic materials may behave in very different ways when subjected to the thermal firing treatment. In some cases the highest bending rate occurs when fusion of the feldspars commences; then at higher temperatures the tendency to bend falls as the vitreous phase dissolves other mineral components of the body, thus becoming more viscous. The addition of vitroceramic materials to bodies can reduce bending, indicating an increase in modulus of elasticity caused by the transition from the vitreous phase to the crystalline phase. Figure 6 illustrates the behaviour of two samples of body for glazed porcelain tile which have undergone a firing cycle with 10 minutes dwell at 1220 C. The black curve is the one in which the temperature rises as a function of time, while the red and green curves relate to the bending of two body samples. In order to be able to display the progression of the thermal treatment, we need to show on the graph the data regarding bending and temperature as a function of time, which is therefore shown on the x-axis. The differences in behaviour are very clear, sample 2 bends twice than sample 1, situation that can cause serious production problems, adversely affecting the productivity and quality yield of an entire factory. Of great importance is the measurement of the bending rate (first derivative of the

Industrial Ceramics Vol. 29 1/2009

 Industrial Ceramics

bending curve) during the maximum temperature dwell: while the sample 1 is decreasing it, sample 2 does not stop bending but exhibit an increase in bending rate. The measurements for monitoring the behaviour of the sample during the thermal treatment are all made using the new optical measuring systems, which allow for very precise measurements without any contact with the material under examination. Most of the properties that are displayed at high temperature would be substantially modified by a measuring system in contact with the material, especially in the sintering stage. Pressure of any kind applied to the sample alters its behaviour. Using optical systems it is now possible to perform surface tension measurements even on molten glaze and to monitor the behaviour of material as it undergoes successive stages of wetting and drying and its wet surface becomes completely plastic.

References 1. A.N. Smith, - Trans. Brit. Ceram. Soc. 54 300 (1955). 2. James E. Young e Wayne E. Brownell, Moisture Expansion of Clay Products - J. Am. Ceram. Soc. 42 (12) 571-81 (1959). 3. I.C. Hope, J. Gabriel e I.C. McDowall, Moisture Movement and Readsorption Phenomena in Dried Clay Articles J. Am. Ceram. Soc. 43(11) 553-60 (1960). 4. A.G. Verduch, Expansin por humedad de los productos cermicos - Bol. Soc. Esp. Cerm. 4 (3) 259-84 (1965). 5. Ravaglioli, C. Fiori e G. Vecchi, Espansione di piastrelle ceramiche per imbizione di acqua e vapore. Studio Statistico. - La Ceramica 17-26 (luglio-agosto 1976). 6. R. Bowman, Importanza della cinetica dellespansione in vapore - The importance of the kinetics of Moisture Expansion - Ceramica Acta 5 (4,5) 37-60 (1993).

Industrial Ceramics Vol. 29 1/2009