Abstract

In service, refractories endure steady and/or transient thermal loading, which induces thermal stresses that are able to damage refractory components, even leading to failure of refractory structures. The assessment of the thermal shock resistance (TSR) of refractories is therefore of central concern for both refractory manufacturers and users.

Traditional methods and techniques to investigate the TSR of refractory systems still strongly rely on rather unrealistic testing conditions (e.g. descending thermal shocks, low temperatures) when compared to their industrial applications.

When considering refractory castables, it is additionally important to keep in mind that this type of refractory system is highly heterogeneous over the thickness of the lining, where only the first few centimetres can develop a ceramic bond, followed by a dehydrated and thus mechanically weak transition zone. Finally at the cold face even hydrate phases may still be found. For reasons of practicality, classical approaches to investigate their TSR largely ignore this key feature. Test pieces made from castables are, on the contrary, usually completely and homogeneously pre-fired before testing for TSR with standardized methods.

Thanks to an innovative testing device, which enables thermal cycling at high temperatures, the TSR of high alumina refractory castables can be investigated under practice-oriented thermal shock conditions and compared to the behaviour of test pieces from high alumina bricks. The damaging was assessed quantitatively and with ultrasonic velocity measurements as well as optical observations.

Compared to the refractory bricks, sintering processes during the thermal shock strongly impact the TSR behaviour of the tested castables. While in test pieces made from fired bricks the damaging is more pronounced and concentrated near the hot face where thermal cycling takes place during testing, the damaging of test pieces made from unfired castables tend to occur at first in the weak bonded “cold part” of the test pieces whereas important sintering takes place near to the hot face. By increasing the number of thermal cycles, damaging near the hot face increases. Pre-fired castable test pieces basically behave like refractory bricks.

These results broaden the knowledge obtained from traditional methods and techniques to investigate the TSR of refractory castables and highlight the necessity to develop smart testing devices and enhanced castables with tailored properties in the transition zone/dehydrated zone.