One of the many virtues of engineered ceramics is their ability to withstand extremely high service temperatures-even in the neighborhood of 3,000°F (1,650°C), Applications that have received wide attention include both moving parts -for such things as heat engines-and stationary products such as recuperators for high-temperature operating equipment. The temperature resistance of ceramics is determined by several thermal properties including thermal conductivity, thermal expansion, thermal shock resistance, and creep resistance.
Creep resistance
This property must be taken into consideration when designing with moving components because high creep rates can lead to excessive deformation and uncontrolled stress. Creep is a thermally activated process and can be divided into three types of behavior. Primary creep starts at the instant of loading and continually decreases with time. Secondary or steady-state creep remains constant with increasing strain and time. Finally, tertiary creep rapidly increases just prior to fracture.
Steady-state creep is most important in structural applications. This rate, E, is based on the following relation.
where is the stress, T is the temperature, N is the stress coefficient, E is the active energy and A is a constant. Though such measurements can give some indication of creep rates, other rate-controlling mechanisms must be considered. These include grain boundary sliding, nucleation of microcracks, porosity in grain boundaries, dislocations, diffusion within grains, and diffusion along grain boundaries.
For structural ceramics, especially hot-pressed silicon nitride, creep rate can be controlled by adjusting the composition and thus properties of the grain boundary phases, An amorphous intergranular phase, in MgO-dopedSiN, can cause grain boundary plasticity, cavity nucleation and growth, and hence high creep rates. By controlled crystallization of the intergranular phase the creep resistance is improved. Stress-dependence can be reduced by doping SiN with Y0, and Zr0, which produce a deformation-resistant grain-boundary phase. Reaction-sintered SiN, on the other hand, has good creep resistance due to lack of oxide in purity phases. The rate-controlling process has been attributed to both grain boundary sliding and microcracking.
Similarly, silicon carbides’ creep resistance is affected by intergranular phases. In general, creep rates are extremely low with a linear stress dependence. However, siliconized SiC have higher rates due to the presence of the continuous Si phase, which is near its melting points under test conditions. The rate-controlling process may be due to a carbon-vacancy diffusion mechanism.
Polycrystalline ceramics generally have high activation energies that weassociate with non-viscous creep or dislocationo movements. Glasses and glass-ceramics, on the other hand, hava low activation energies that are associated with viscous flow mechanisms due to the amorphous phases present. Consequently, glass ceramics are rate-controlled by the glass phase, and like ceramics, the creep rate is influenced by the structural morphology and composition of the crystalline phase. Refractory phases reduce deformation or creep. By further heat-treatment of glass-ceramics, most of the glass phase is removed and the material acts more like a conventional ceramic.
Thermal expansion
Another critical property that affects temperature resistance. Compared to oxides, silicon-based ceramics have a low coefficient of expansion, which helps to increase thermal shock resistance. However, when used in conjunction with metal components or as coatings on metals, large thermal-expansion differences between the two different materials may be a problem.
Like creep resistance, thermal expansion for silicon ceramics is a function of the solid phase and is not much affected by porosity and minor impurities. SiC materials have 50% higher thermal expansion than SiN, materials. Both materials are affected by additives. For instance, Si-coated.
SiC has a lower expansion than Al0 doped SiC because the latter has a higher expansion aluminosilicate grain boundary phase; the silicon metal phase has lower expansion.
Thermal expansion for SiN materials is a function of the amount of intergranular oxide phase present, Minimizing the phase, as for pure reaction sintered SiN, produces Low thermal expansion. Hot-pressed SiN, materials, on the other hand, contain higher amounts of high-expansion oxide phases. The difference in expansion between a reactionsintered-SiN, and a hot-pressed-SiN, can be as much as 30% indicating the influence of processing methods. SiC materials are not as affected, with maximum variation of only 5%.
With glasses, the expansion with temperature is also important because thermally-induced stresses are a function of the rate of expansion. Both the linear dimensions and the volume of glass change with temperature; the rate of increase of volume is three times the linear rate of expansion. Up to 572°F (300°C) the total expansion is basically linear with increases occurring with annealing. The addition of oxides to conventional glasses usually increases the coefficient of expansion significantly, except for certain borosilicate glasses where a change in coordination number between the B and 0 does occur.
Glass-ceramics have similar thermal expansion requirements to ceramics. For high thermal shock resistance, the coefficient of thermal expansion must be kept low, except for sealing or joining to metals, it also must closely match that of the other material. As with structural ceramics, the crystalline phase or phases (both type and volume) present will influence the final expansion. Depending on the phase, the expansion can range from a negative coefficient to a very high positive coefficient. However, the composition of the residual glass phase is important – if changed it can cause excessive thermal-expansion mismatching.
Heat-treatment of glass-ceramics (or glasses), which determines the volume fraction and nature of the crystal phases present, also can change thermal-expansion characteristics. For instance, if crystalline forms of silicon are present, different treatments can result in different coefficients. This occurs because the material undergoes structural changes at different temperatures. Quartz is one type of silica structure that has a high coefficient of expansion.
Heat capacity and thermal conductivity
These also must be considered for ceramic candidates for high-temperature applications, since both determine the rate of temperature change in a ceramic during fabrication and use. Both properties determine the thermal stress resistance as well as the operating temperatures and temperature gradients. For instance, a low thermal conductivity is required for materials used as thermal insulators.
For most oxides and carbides, the heat capacity increases from a low value to near 6 cal/g atm °/C at about 1,832°F (1,000°C). Further increases in temperature have little effect on heat capacity and neither does the crystal structure. However, heat capacity does rapidly increase for order-disorder transformation over limited temperature ranges. Similar changes in heat capacity occur at magnetic and ferroelectric transformations. The change in molar heat capacity is not large for a polymorphic transformation, though the volume heat capacity is structure-dependent, specifically when porosity is present. Less-dense materials have lower heat capacities.
For glass, the heat capacity increases with temperature and approaches zero at absolute zero. The only abrupt changes that occur are when the glasses pass through the glass transition to the liquid state. The heat capacity can increase by a factor of 1.3 to 3-a condition that can be reduced by annealing. Glass-ceramics also show an increase in heat capacity with an increase in temperature and have similar values to glasses and ceramics.
Thermal conductivity in glasses differs greatly from that of crystalline ceramics and is difficult to measure accurately, In glasses, the conductivity decreases with temperature and reaches very-low values near absolute zero. In crystals, conductivity increases as temperature drops to very low values. Glass-ceramics are similar to conventional crystalline ceramics, though actual values of thermal conductivities are lower than for oxide ceramics and higher than for glasses.
The thermal conductivity of single-phase ceramic materials, which is based on phonon scattering, is influenced by the structure and composition. Complex structures with wide ranges of atomic size variations have lower thermal conductivity caused by more scattering. Impurity atoms in solid solution also decrease conductivity, though scattering is greatest in simple lattices and at low temperatures. The conductivity of mukiphase ceramics becomes even more complicated as the amounts and the arrangement of each phase, which includes porosity, must be taken into consideration.
Thermal shock resistance
Basically, this property reflects the ability to withstand thermal stresses generated during large temperature differences. Such stresses can lead to “thermal-shock” damage, producing fractures that result in catastrophic failure, or causing existing flaws to grow, reducing strength and component integrity.
Thermal shock resistance usually is determined by quenching, samples from various known temperatures into a fast-flowing water bath having a standardized temperature. The initiation of thermal shock damage is determined by internal-friction measurements. Conditions such as shape of sample and cooling mechanism must be controlled carefully since both affect the final results. The shape can change thermal-shock resistance and the medium can influence the rate of surface-heat removal.
Most glasses, porcelain, white wares, and special electronic and magnetic ceramics fall into the category of elastic materials, which fracture when the surface stress reaches a certain level. Temperature conditions for fracture thus can be easily calculated. However, conditions of surface heat transfer must be considered since these parameters will change the advantages of certain materials over others. For example, MgO is better for slow rates than porcelain. The latter has better stress resistance at rapid rates of heat transfer. Other condition are important – is the heat transfer steady state?
For structural ceramics, thermal shock resistance is dependent on both material type and processing method. For instance, hot-pressed silicon nitride usually has higher resistance than reaction-sintered silicon nitride. Furthermore, the dopants have an effect- Y0-doped SiN, is better than MgO-doped SiN. However, silicon carbide materials have lower resistance than the nitrides because of their high thermal expansions and elastic moduli. Processing methods can not influence these properties though higher thermal conductivity can help in certain applications. For use as combusters, where the temperature remains static, the high thermal conductivity helps to minimize hot spots.
Similar to ceramics, glass-ceramics must have high strength, low thermal expansion, and low elastic modulus to have high thermal shock resistance. Studies have shown that the linear coefficient of expansion is the most critical factor for determining thermal shock resistance. In fact, glass-ceramics have such low coefficients of expansion that in some cases they have better thermal shock resistance than ceramics. Again, the major crystal phase present controls the overall properties of the material.
Oxidation resistance is important for certain high-temperature applications. Silicon-carbide materials, due to their higher purity and higher density, are generally more stable in long-term applications than silicon nitride because they are more oxidation resistant. Hot-pressed silicon nitride is affected by alkali impurities that segregate in grain boundaries which increase the oxidation rate. This rate can be reduced by adding Y0 and Si0. Reaction-sintered silicon nitride is usually accompanied by higher porosity which also lowers the oxidation resistance.