Abstract
Chalk is a weak rock that occurs extensively in the UK. The mechanisms which control the mass compressibility of chalk are not fully understood and hence current foundation design practice favours a conservative, and therefore costly, approach. This research aimed to investigate the factors which control mass compressibility in order to improve our understanding and hence improve foundation design methods. This was achieved by the development of a hypothesis which described aspects of the deformation behaviour of both intact and discontinuous chalk. The hypothesis was then validated by a laboratory investigation of an intact high porosity chalk, using one-dimensional compression and triaxial tests, followed by a study of discontinuous chalk, which represented an idealised rock mass. The hypothesis, which was based on friction theory, indicated that the yield stress (strength) of the chalk was a material constant and was independent of the size of the chalk body. It also suggested that the primary control on mass compressibility was yield and failure of the asperities at the discontinuity boundary, with an associated increase in contact area and progressive discontinuity closure. Recent work has found that it is possible to use a unifying framework, based on Critical State Soil Mechanics, to describe the deformation behaviour of intact, bonded materials, including chalk. This approach was adopted to describe the behaviour of the intact chalk. It was found that the intact chalk behaved, at low stresses, like an intact, elastic rock which was able to exist in stress states that were impossible for reconstituted chalk, due to the contribution of structure. The chalk was described by a double yield model, where first yield (linear limit) was associated with the onset of bond degradation and second yield marked pore collapse and the start of grain crushing. At large strains, beyond second yield, the chalk behaved as a particulate, granular soil. The introduction of a smooth, planar discontinuity did not significantly alter the linear limit or yield surface from that found for intact chalk. It did cause a slight reduction in stiffness and an increase in accumulated strain. A series of uniaxial compressive strength tests on discontinuous specimens of varying contact area ratios confirmed the hypothesis that the yield stress of the chalk was a constant. Inspection of the specimens and analysis of the results confirmed that asperity crushing, with the consequent change in contact area and discontinuity closure, had been key factors in the deformation behaviour. Specimens with a contact area in excess of 51% demonstrated similar, limited axial strains, while those with a contact area of less than 44% demonstrated much larger axial strains. An empirical relationship was observed between both initial contact area and initial asperity height (aperture) and the amount of axial strain developed. Triaxial tests on 15% contact area specimens identified a linear limit and yield surface similar to that of intact chalk. Much lower stiffnesses and larger accumulated strains were recorded. The differences observed in the behaviour were thought to be a function of the reconstituted chalk at the discontinuity boundary, which was created during discontinuity closure. The main factors which governed the mass compressibility of chalk were found to be material yield stress, true contact area ratio and discontinuity aperture.