Silicate Ceramics

In the production of silicate ceramics, the heat treatment determines product quality but also adds to manufacturing costs. It is therefore worthwhile to optimize the process parameters of the heat treatment step. For the drying process, including the binder burn-out, the quartz conversion and the dewatering, methods have to be identified which, on the one hand, enable a maximum throughput but on the other hand ensure that the component are not damaged. During firing, near-net-shape production with tight production tolerances and good material properties is aimed for. For example, new designs can be realized by introducing suitable firing conditions. For financial reasons, the output of the heat treatment has to be maximized but the energy consumption and maintenance effort has to be minimized. The rejection rate has to be minimized as well.

The methods used at Fraunhofer-Center HTL are suitable for all types of silicate ceramics: Porcelain, earthenware or stoneware as well as siliceous refractory materials. Small components can be optimized as well as multi-meter-long high-voltage insulators made of porcelain or pipes made of stoneware. For the recording of the necessary in-situ measured data, suitable ThermoOptical measurement systems (TOM) are available at HTL. These can be operated in air, but also with the atmosphere of gas-fired furnaces. All material properties necessary for the optimization of the heat treatment of silicate ceramics can be measured in situ. The in situ measurements are carried out on samples with a volume of up to 30 cm³. The upscaling to component scale is supported by finite elements (FE) methods. For all essential heat treatment steps on silicate ceramics, corresponding FE models are available. The process parameters are optimized using FE simulation and experimentally verified afterwards. In the last step, the optimized process parameters are transferred to the production furnace.

During the process of drying, the water in the green bodies has to be removed without damage. Due to the small pores, considerable drying stresses can occur. Too fast or uneven drying leads to cracking or warping. The local drying rates are strongly dependent on the relative humidity, the temperature and the air flow. The HTL is equipped with a special device that can be used to weigh bodies under controlled humidity, temperature and gas flow during drying. In this way, the drying rate can be determined during the different drying stages. In the critical stages, the drying rate can be lowered selectively. Warping of the green bodies due to non-uniform drying is simulated using FE models, in which the air flow and the local moisture gradient are taken into account. During the optimization process, moisture and flow measurements in the industrial drying units are taken into account, too.

If the green bodies contains organic additives, combustion heat is generated during burnout. Inside the green bodies, where the oxygen concentration is very small during the binder burnout, pyrolysis reactions can take place simultaneously, which consume heat. The resulting temperature differences cause thermal stresses, which in turn can lead to cracks or to the destruction of the green bodies. If the binder burns too fast, the resulting gases cannot move through the pore channels to the component surface fast enough. The resulting overpressure in the pore channels can also lead to component damage. Similar to drying, it is important to find the fastest, still reliable temperature profile with which the burnout of the binder can be performed without creating defects. In many cases, the flow velocity and the composition of the furnace gases can be varied. Analogous to drying, experiments with respect to the burnout of the binder are carried out in a controlled atmosphere and temperature. The degree of reaction is monitored by continuous measurement of the sample mass. Crack formation is detected by sound emission analysis. The optimization of the process parameters takes place analogously to drying by means of FE simulation, the verification of the optimized conditions is done by additional experiments on large components. A gas-heated measurement furnace, TOM_gas, is available for burn-out of binders on large components at the HTL. In this furnace, green parts with a mass of up to 15 kg can be weighed during the binder burnout.

Dewatering takes place in silicate ceramics, when the bodies are heated up to higher temperatures - approximately between 500 °C and 700 °C. The crystal water contained in the clay is removed from the bodies (dehydration). For example, kaolinite is converted into metakaolin by dehydration. If the heating is too rapid, the water vapor released in the bodies cannot escape quickly enough. There is an increase in pressure, which can lead to the destruction of the body. In the same temperature range at approx. 573 °C, the quartz contained in the mass is also converted. The quartz transforms into its high-temperature modification and expands by approx. 0.8% (quartz conversion). The volume increase generates mechanical stresses, which can also damage the component. The optimization of the process parameters of the dewatering step is carried out by means of in-situ measurements of the weight changes and, if applicable, scaling and subsequent FE simulation. The verification of the optimized process parameters can then be carried out on large components using TOM_gas. Similar optimization strategies as for dewatering are also used, when carbonates are contained in the raw materials. These decompose in the temperature range around 800 °C to 900 °C, whereby the released gases can also lead to a critical overpressure inside the bodies.

When firing silicate ceramics, densification with a defined residual porosity has to be achieved. The structure should be defect-free and homogeneous, and the component must not warp. If glazes are used, they must be free from defects such as pinholes or bubbles. Deviations from the desired geometry are a consequence of a non-uniform distribution of the porosity in the body as well as of the influence of friction, gravitation and temperature differences during the heat treatment:

• A non-uniform porosity distribution in the body can be caused during shaping, e.g., by sedimentation effects during slip casting. This can be detected at the HTL using very sensitive measuring methods.
  
• Friction and gravitation lead to the warping of the bodies if their stability is too low. In particular, the region above 1000 °C, where larger amounts of melting phase can be formed but stabilizing reactions like the formation of secondary mullite are still low, is critical
  
• Temperature differences within the components lead to warpage during firing. Higher temperature differences in the components can also cause cracks or lead to the destruction of the components. Temperature differences in the furnace lead to scattering in the residual porosity and the component dimensions.

If iron oxide is present in the raw materials, the conversion of the trivalent to the divalent iron must also be considered, which takes place in air at about 1200 °C. The oxygen released during this conversion must be able to leave the components through the pore channels. If the pores are already closed, due to strong sintering shrinkage, the bodies swell. As a countermeasure, reducing atmospheres that reduce the temperature of the iron oxide conversion must be used. Gas-phase processes often lead to glaze defects. By means of optical dilatometry and weighing, the sintering and degassing behavior of glazes can be analyzed and optimized at the HTL.

The sintering shrinkage of the bodies can be measured exactly using the TOM systems. The atmosphere of gas-fired ovens is precisely replicated. In addition, other important material properties with respect to heat transport and creep deformation during firing are recorded. By means of coupled FE models, the thermomechanical effects in the component are simulated, and the process parameters are optimized. In addition to the furnace atmosphere and the firing cycle, firing auxiliaries can also be considered which have a great influence on the temperature distribution in the body and may have to be used to support bodies with low stability. The optimization procedure can also be used for biscuit and glost firing processes as well as for fast firing. The verification of the optimized process parameters can be carried out on large silicate ceramic components using TOM_gas.