A new type of experimental investigations of the spalling stress thresholds has been designed. Experiments on large cylindrical rock cores containing a notch have been conducted. The far field strains and local strains at the notch were measured together with AE-measurements on the specimens during the loading. The formation of the spalling chips in the laboratory experiments looks the same as in field experiments, which indicates that the stress state and failure processes are similar in both cases. The spalling stress threshold was higher than expected in accordance with previous suggested laboratory experiments. The identification of the various crack stages provided some uncertainty and the number of tested specimens were few. The position of the localization of the spalling obtained at the experiments agrees well with calculations.
Several researchers found that the in-situ spalling strength of a rock mass is significantly lower than the uniaxial compressive strength (UCS), even when the rock mass is massive and homogeneous (Diederichs et al. 2004, Martin & Christiansson 2009). Martin and Christiansson (2009) suggested that the spalling strength is fundamentally an extension mechanism and that it corresponds to the stress level associated with crack initiation in uniaxial compression tests, which occurs at about 0.4 to 0.5 of UCS. A fundamental issue for the design and stability of underground openings is to be able to predict the stress level that will initiate spalling. Field observations and experiments have increased our understanding of the initiation and progress of spalling (Martin 1993, Andersson 2007). However, field experiments are expensive to conduct. Moreover, it is difficult to determine the insitu stress state with a high degree of accuracy at which spalling is initiated. Different types of laboratory experiments have therefore been proposed for investigating spalling.