Refractories are complex materials with a microstructure which is generally complex and heterogeneous (grains with a wide range of size and with a multiplicity of morphologies, porosity and cracks, etc). At room temperature, refractories generally exhibit microscopic cracks due to thermal expansion coefficient mismatches of their various phases caused by thermal stresses at the cooling stage of their fabrication. The analysis of their thermo-mechanical properties is crucial to improve their high temperature performance.
At room temperature, the thermo-mechanical properties of refractory castables for applications in the metallurgical and steel industry are continuously investigated in order to improve their performance against thermal shocks. In order to optimize the service life of refractories, the thermal shock resistance is currently determined based on the analysis of the Young’s modulus after different thermal shock cycles with the help of the non-destructive impulse excitation technique (IET).
At elevated temperatures, the Young’s modulus is a mechanical property relatively sensitive to the evolution of the microstructure and can be characterized by the impulse excitation technique (IET) as function of the temperature.
The impulse excitation technique is a non-destructive technique used to measure material properties. After the sample is mechanically excited, the resonance frequencies and internal friction are recorded. These material properties are a fingerprint of the sample and monitoring them in time gives valuable information regarding failure mechanisms. Typically, measurements are performed at room temperature after the samples are exposed to different cyclic loading schemes or specific heat treatments.
Traon et al. studied the thermal shock behavior of high aluminia castables composed of tabular alumina and partially stabilized zirconia (PSZ). They were pre-fired at 1300 °C for 6 hours at a heating and cooling rate of 1K/min. The grains of the sample had a maximum size of 3mm and the sample had a rectangular shape with a length of 160 mm, a width of 40 mm and a thickness of 40 mm.
The samples were subjected to thermal shocks in according to DIN EN993-11 at four different temperatures (750 °C, 850 °C, 950 °C and 1050 °C). After every thermal shock, the Young’s modulus and internal friction were measured according to the ASTM E1876-07 standard with the RFDA professional.
Fig. 1 shows the retained Young’s modulus values of the PSZ samples as a function of the number of thermal shock cycles at different tested temperatures (750, 850, 950 and 1050 °C). During the first thermal shock cycles, the Young’s modulus is degrading significantly. This can be explained by the crack nucleation in the grain boundaries which becomes more noticeable as the temperature changes increases.
As shown in Fig. 2, the internal friction data is highly correlated with the Young’s modulus data as the lower the Young’s modulus, the higher the internal friction due to the formation of cracks and microcracks. Damping values do not change as a two-step process. Indeed, damping increases proportionally with the damage increase. Based on the conclusions of Traon et al., the experiments conducted at a thermal shock temperature change of 1050 °C leads to the conclusion of a four-step evolution of the microstructural changes. Up to and including the second cycle, crack nucleation at the level of the grain boundaries causes the first increase of the damping measurements up to 1000 %. The second increase reaching 2000 % of the initial value occurring up to and including the fifth cycle can be explained by the propagation of the cracks within the matrix and throughout the Al2O3 aggregates. This step is followed by a steady state up to and including the eighth cycle while the cracks have to consume a lot of energy in order to bypass the ZrO2 aggregates.
Finally, the unification of cracks occurs during the ninth cycle, this network of cracks explains the depletion of the elastic properties and the further difficulties in achieving realistic damping values over 2500 % of the initial value.
Based on the obtained results, it is clear that the combination of elastic properties and damping data obtained by the impulse excitation technique leads to a better understanding of the microstructural changes within the refractory castables after progressive thermal shocks.
- N. Traon et al., Estimation of Damage in Refractory Materials after Progressive Thermal Shocks with Resonant Frequency Damping Analysis. J. Ceram. Sci. Tech., 07 (2016) 165-172
Pabst et al. studied the temperature dependence of the Young’s modulus of silica brick materials. They consist almost entirely of tridymite (64 %) and cristobalite (36 %). The rectangular shaped samples had a length of 160 mm, a width of 20 mm and a thickness of 10 mm. The Young’s modulus at elevated temperatures have been measured by the IET according to the ASTM E1876 standard in the IMCE RFDA HT1600 furnace. Resonant frequencies have been measured continuously from room temperature to 1200 °C (first heating cycle), down to room temperature (first cooling cycle) and followed by two similar heating and cooling cycles with a controlled heating and cooling rate of 5 °C/min.
Fig. 1 and 2 show the temperature dependence of the Young’s modulus of a silica brick material with 19,4 % porosity from room temperature to 1200 °C/300 °C measured with the impulse excitation technique. Based on the conclusions of Pabst et al., it is clear that the temperature dependence of the Young’s modulus during the first heating cycle is different from the subsequent cycles. In the intermediate temperature range between approximately 50 and 250 °C, these materials become very flexible with stiffness minima of around 60 % of the room temperature values. The sudden changes in this temperature regime are clearly related to phase transitions between subpolymorphs of tridymite and cristobalite.
Also a clear hysteresis effect i.e. shifts of the phase transition temperature (of around 30 °C) is shown in this temperature regime. The volume contraction opens existing microcracks and lowers the E-modulus of the material. Therefore the abrupt volume changes during heating and cooling over phase transitions are responsible for microcrack opening and closing and lead to an indirect enhancing effect. Furthermore, it is shown that the Young’s modulus at around 800 °C can be more than three times as high as the room temperature. Finally, after a complete heating-cooling cycle, the initial room temperature Young’s modulus is obtained again which means that a temperature of 1200 °C does not induce any irreversible changes in the microstructure. In other words, the elastic modulus of these materials is not a unique function of the temperature but due to the thermal memory, it is dependent on the thermal history.
From the results, it is clear that the hysteresis loops can be reproduced many times and that the microstructure affects the absolute values of the Young’s moduli, while the transition temperatures are generic features of the solid phases, essentially unaffected by microstructural details. Therefore impulse excitation measurements can be considered as a highly sensitive tool for the investigation of phase transistions.
- W. Pabst et al., Elastic anomalies in tridymite- and cristobalitebased silica materials. Ceramics International 40 (2014) 4207-4211