Effects of restraint and test configuration on thermo-mechanical stress states in high temperature laboratory tests for spalling in concrete

This paper presents a numerical investigation of the role of restraint applied to high temperature laboratory tests for spalling of concrete. Specifically, the study focused on the effect of restraint on the stress state in the concrete at the time of spalling. The investigation used 3D numerical modelling and parametric studies to explore a consistent set of experimental tests reported in the literature and make a comparative study of the influence of the different factors at work. Different test configurations were considered with different shapes and different methods of restraint including passive steel frames and active loading. Consideration was also given to the way in which heat was applied and to the effects of friction between the sample and external components of the test set up. Generally, it was found that the application of restraint can significantly affect the development of the stress states inside concrete samples exposed to high temperatures and that the precise effect is dependent on the specific configuration of the test with some producing more severe conditions than others. The development of compressive stresses near the hot face of a sample seems to correlate well with experimental observations of spalling and test configurations that promote this seem to spall earlier. In summary, it is clear that both the shape of the sample and the nature of the applied restraint influence the stress states in the sample and as a consequence, may then affect the spalling behaviour. However, the conditions necessary to initiate spalling seem to be able to develop with or without restraint.