Recent developments of large digital computers and numerial methods such as F.E.M. have made possible, in principle the solution of any boundary value problem in rock mechanics and engineering. However, the realistic strain-softening behavior of soft rock has been neither clarified experimentally nor incorporated in the analysis yet. In order to breakthrough these barriers, the basic mechanical properties of soft rocks must thoroughly be investigated through experiments.
The stress-strain behavior of soft sedimentary rock, which usually shows a decrease in strength after reaching a peak value, has been recognized for quite some time. The loss of strength contributes significantly to the manner in which the stresses and strains are distributed and leads to a progressive type failure. A microscopic view of failure mechanism in a rock sample has been presented by many researchers, supported with the techniques of acoustic emission (AE)1) and fractography. The behavior of soft rock must be reviewed in this standpoint.
A new triaxial testing system which is servo-controlled was designed and constructed. The control signals and test data are processed in a microcomputer. In the compression and extension tests, the lateral deformation of the specimen (5 cm in diameter and 10 cm in length) is measured by both cantilever type displacement meters (at three points) and strain gauges. This paper describes the outline of experimental procedure and interprets the strain-softening behavior of soft saturated tuff by a energy theory, accounting for the growth of cracks in a test sample. A part of the work done to the specimen would be stored as strain energy in the specimen, and the other would be dissipated with in the specimen as a dissipated energy. The fracture in the specimen may be directly related to the nonrecoverable energy. The failure process of the rock sample will be interpreted as a energy transfer in this study.
A sedimentary soft rock called Funyu Tuff (Gs=2.65, n=29%, γ d=1.87 g/cm 3 uniaxial compressive strenth, 120 kgf/cm 2)is used for a conslidated undraind test. The sample is fully saturated by a vacuum. The change of pore water pressure in the sample during compression process' was measured by a small transducer. The lateral displacement of the cylindrical sample in compression was observed by three cantilevers as shown in Fig. 1. The consolidation (initial confining) pressure was 10 and 70 kgf/cm 2and the strain rate in the test was 0.12%/min. Data aquisition, prosessing and plotting are done by a microcomputer system.
Several different tests were performed. In order to varify the nonuniform distribution of strain and water content in the test sample, four strain gauges were attached on its surface as shown in Fig. 2. Water contents were measured at various point after completing the test. For the investigation of the energy dissipation during fracture process, three loading types as shown in Fig. 3 were adopted. A usual triaxial compression test is conducted by the constantly increased (CI) loading (Fig. 3(a)) And the energy dissipation at an arbitrary displacement may be calculated from the test result of gradually increased repeated (GIR) loading (Fig. 3(b)).