Examples | Thermal shock resistance of refractories | Dehydration and sintering of refractory castables

Thermal shock resistance of refractories

 Use of SonicByte^TM for non destructive characterization of refractories damage caused by thermal shock 

Thermal shock

One of the main causes of refractories deterioration in service is thermal shock. The later is produced by rapid temperature changes causing the appearance of thermal stresses in the refractory and therefore the propagation of its microcracks.

Thermal shock resistance of refractories

There are several methods to characterize the thermal shock resistance of refractories. The example presented in the following refers to a thermal fatigue test method where two 1 x 1 x 6 inches (2.54 x 2.54 x 15 24 cm) samples are simultaneously subjected to 100 "heating-cooling" cycles. The heating of the samples is done using an horizontal electric furnace. Their cooling is achieved on a water cooled copper plate kept at 20°C using a thermostatic bath (see Figs. 1 and 2).

To quantify the cumulative damage of the samples during thermal cycling, the set-up used was equiped with a system allowing to impact them under flexural conditions during the half-cycles cooling. The SonicByte^TM, laboratory version, was used to collect the acoustic signals generated after impact to determine the samples residual elastic modulus after each cycle.

The tests were conducted on an alumino-silicate refractory castable previously fired to 815°C. The temperature variations imposed on samples during the tests were varied between 200 and 900°C. The results obtained are shown on Figure 3.

 Fig. 1: Global view of the thermal shock set-up and the SonicByte^TM (Laboratory version).

 Fig. 2: View of the samples during one of the half-cycles cooling by means of a water cooled copper plate.

Fig. 3: Residual elastic modulus (E/E₀) in function of the imposed temperature variation and number of cycles (N).  

In reference to these results, quantification of the effect of thermal shock on the extent of damage of the tested material could have been achieved through SonicByte^TM. The latter has been able to provide accurate elastic modulus values regardless of the damage level achieved in the samples during testing. Note that a small increase of damage in an heterogeneous material greatly increases acoustic waves attenuation. Nevertheless, the SoniByte^TM succeeded in detecting the very low acoustic signals generated by the samples which reached up to 80% of damage (i.e., a residual elastic modulus of 0,2) during the tests (see Fig. 3).