INTRODUCTION
Samples of outcrop chalk have been deformed to failure using an advanced triaxial testing facility. The apparatus allows precise control of axial, confining stress and pore fluid pressure, which enables a test sample to be deformed following a specified stress path. A series of identical chalk samples were loaded under conditions of fixed lateral to vertical stress ratio (K) where K was set to different values between 0 (laterally unconfined) and 1 (isotropic stress) followed by undrained triaxial loading if the sample did not reach failure along the K defined stress path. During each experiment the states of axial, radial and volumetric strain were recorded. These strain data show consistent trends that permit interpretation of the deformation mechanisms active during different stages of loading along the prescribed stress path. The data presented illustrate a consolidation surface contoured with respect to strain and dominant deformation mechanisms. The data provide the basis of a model for the deformation of bonded porous media and have important implications for assessment of strain states in compacting hydrocarbon reservoirs.
Previous experimental work has indicated that the behaviour of chalk although similar to a particulate soil in the sense that the critical state is attained at failure, is complicated by the proximity of a low stress elastic yield surface (Leddra & Jones 1990, Leddra 1990). Undrained triaxial tests on the chalk show that at stress states external to the elastic yield state, the material readily deforms and can, in certain instances, flow (figure 1). Other experimental studies (Addis 1987, Addis & Jones 1990, Leddra 1990, Leddra & Jones 1990), which have included uniaxial strain (ko) tests, have highlighted the transition from intact, competent, elastic rock to particulate sediment with increasing levels of confining stress. To further understand how stress conditions affect the structure and behaviour of chalk and to refine the rheological model, a series of drained triaxial tests have been completed on identical samples of outcrop chalk. These tests have differed from standard drained tests (Bishop & Henkel 1962) in that the samples were loaded under a fixed lateral to vertical stress ratio (K) where K was set to different values so that strain behaviour could be investigated within a spectrum of stress conditions. These experiments are similar to those reported by Teufel, Rhett & Farrell (1991) for North Sea chalks but have the advantage that all test samples were identical companions. The experiments also access a far wider range of K values than those of Teufel et al.
In previous studies the deformation behaviour of the chalk has been represented by various forms of stress-strain curves and by stress path diagrams. The stress invariants, q (deviatoric stress, o'1-63) and p' (mean effective stress, (o'l + 20'3)/3) are plotted against each other in stress path diagrams and define the state of stress surrounding the sample. The value of stress path diagrams is however limited, because they do not describe the strain behaviour of the sample, whilst stress-strain curves do not easily allow rapid and direct comparisons to be made between deformations on different stress paths. In this study strain data has been incorporated into stress path diagrams as contours of equal strain (isostrains). This method has been employed independently for the axial, radial and volumetric strains measured during the stress path tests. Previous experimental data (Leddra 1990, Leddra & Jones 1990) have been processed in a similar way, permitting comparisons with the new data reported here.
The chalk used in this study and in previous work is from a quarry