Laboratory tests on dry outcrop and artificial sandstone indicate that non-elastic deformation appears quickly upon loading from the forming state or a turning point on the stress path. Upon further loading the non-elastic part of the deformation may increase or decrease depending on the stress history of the rock. Upon unloading from the forming state or from a turning point of the stress path the non-elastic part of the deformation appears to grow linearly with stress alteration, starting from zero. The results may be utilized to obtain values for purely elastic stiffness on the basis of static measurements, at points where the stress path turns from loading to unloading. Combined with dynamic (velocity) measurements, this may allow for estimation of dispersion, or estimation of Thomsen’s d-parameter, without the use of complicated experimental equipment.
1. INTRODUCTION
Rock volumes situated in the vicinity of rock altering events are usually subject to stress changes, ranging from next to nothing and up to levels that may even induce failure. For instance, the magnitude of stress changes induced by the drilling of a well depends - according to linear elasticity - on the inverse square of the distance from the well, and hence vanishes asymptotically. Also for the rock around a reservoir, depletion induced stress changes decrease with distance in a similar way. For activities such as modeling of rock deformation and interpretation of time lapse seismics, it is of interest to know where and when the rock behaves like a linearly elastic material, and how the transition from such behavior to non-elastic behavior occurs.
We here present results from laboratory tests on relatively weak outcrop sandstone, where both static and dynamic (elastic) moduli are measured simultaneously. Particular attention is given to the behavior when the stress path makes a turn, which for such materials is the closest we get to a field situation where the stress state is shifted from its in situ equilibrium. Previous results [1, 2] show that the difference between static - generally non-elastic - and elastic behavior is small when the stress state is close to a turning point, while the difference grows as the stress state deviates more and more from the turning point.
We also present results from tests where artificial sandstone is cemented under stress, and subsequently subjected to stress alterations. In terms of stress path, this situation is even closer to the field situations we are considering.
2. LABORATORY TESTS
Some of the laboratory tests were performed on outcrop sandstone [2, 3], while some of the tests were performed on synthetic sandstone formed under stress [4]. The tests were performed on core plugs of size 3” (length) by 1½” (diameter). The test setup is shown schematically in Fig. 1. P-and S-wave velocities were measured in the axial direction by transducers mounted in the pistons, while transducers clamped to the sleeve were used to measure the P-wave velocity in the lateral (radial) direction. Axial deformation was measured by LVDTs while lateral deformation was measured by a chain.
