Deformation and failure of rock

Sammanfattning: In this doctoral thesis work the relationship between the deformation and the initiation of failure around an underground excavation was studied. In this work data from observed failure and measured deformations was used. This work is based on information obtained from laboratory test, underground field cases and numerical analysis. Field data from the underground cases was used as input data in the numerical analysis. The underground field cases provided good information for analysis of failure and deformation. These cases provided valuable information due to the following characteristics: (i) spalling and deformation can be studied from the Garpenberg raise and Zinkgruvan exploration drift cases using numerical analysis, (ii) bending and shear can be studied from the Kristineberg mine since deformation measurements were conducted in situ, and (iii) wedge failure can be studied from the Kiirunavaara drift case since it was observed and deformation was measured in situ. The failure-deformation process was analysed for fifteen tested rock types. Rock characteristics such as grain size and mineral compositions for these rock types were compared with a number of strain quantities. The comparison revealed that the crack damage lateral strain was strongly influenced by the grain size. Furthermore, the mineral composition slightly influenced the examined quantities. Both crack volumetric and volumetric strain quantities seems to be sensitive to grain size. The comparison of the strain quantities, representing different deformation stages obtained from laboratory tests using information from limestone and quartzite, and calculated for the Garpenberg raise and the Zinkgruvan exploration drift cases using numerical analysis, helped to interpret the failure process around the opening in these two cases. The distances for strain concentration bands from the linear-elastic brittle and linear-elastic perfectly plastic analyses that were closest to the boundary were similar to the depth of the observed fallout for both cases. The deformation pattern of a monitored stope in Kristineberg mine helped to determine that bending and shear failure took place in the HW and FW side of the stope. The shear failure was first observed by borehole camera surveys inside the walls, and then later daylighted on the surface of the stope. The onset of the shear failure was represented for a convergence of 56 mm. In the Kiirunavaara drift case, the results from two local models (2D and 3D) showed qualitatively good agreement with the field observations. The models are able to simulate the wedge and determine the stability of the wedge. It can be inferred that the wedge did not fallout and remained stable. In fact the rock reinforcement installed in the field was not necessary to keep the wedge in place. This work conclude that the failure process of each failure mechanism develop differently. The failure process due to spalling can occur very fast, while for bending and shear failure it is slow. The behaviour of wedges can be monitored as long as they are stable during the excavation process. Monitoring data from extensometers, total stations, convergence measurement, borehole camera surveys, combined with appropriate numerical analysis can be used for estimating the failure and deformation of spalling, bending, shear and wedge failure. Deformation values were estimated for all failure mechanisms. Strain based failure criteria have to be developed and other input parameters have to be collected. Keywords: failure, deformation, spalling, bending, shear, wedge, laboratory test, underground cases, numerical analysis, monitoring, onset of failure, crack initiation, crack damage, fallout, stability, failure criterion.