Since the commencement of mining at the strip coal mine at Goonyella, Queensland, Australia, serious stability problems have been experienced in both the spoil piles and the highwalls. A collaborative investigation was commenced in 1974 and the significant factors influencing stability have now been defined. This paper outlines the results of field and laboratory studies associated with the investigation and describes the various numerical methods used to analyse stability. The main conclusions, which are similar for both the spoil piles and the highwalls, are that the failures generally occur along two or three planar surfaces; that the shear strength in the basal areas is controlled by the sensitivity of the materials to moisture and shear strain prior to failure; and that tensile cracking due to stress relief and blasting is the dominant factor in the upper or escarpment areas. Possible control and remedial measures are discussed briefly.

SUMMARYOne major problem encountered when using the finite element method to simulate the load-deformation behaviour of an elasto-plastic soil mass is that the theoretical collapse load is generally exceeded, and in some cases the numerical solution fails to exhibit a collapse load. A mixed variational principle is used as the basis for developing the governing equations of deformation and the results from this are compared with those obtained from the use of the virtual power equation. The former is found to give improved results and is generalized to include description of finite deformation. An Eulerian frame of reference is used. The method of approximating configuration changes in the numerical solution procedure is found to determine the shape of the load-deformation curve. Finally, a simple method of accounting for anisotropy of yield is presented.

No abstract

Analyses of load tests on 100 instrumented bored piles in different weathering grades of different tropical geological formations of peninsular Malaysia enabled correlations of ultimate shaft and base resistance with standard penetration test (SPT) results and unconfined compression strengths. The data also enabled development of shaft resistance (t–z) and base resistance (q–w) models that are related to SPT, unconfined compressive strength and rock types. The t–z models can be used for strain softening and strain hardening while the q–w models are for strain hardening and stiffening behaviour. The models thus developed were applied for analysis of 35 piles in the database that were loaded until the load–settlement curves were significantly non-linear. Most of the analyses resulted in a reasonable match with the measured load–settlement and load-transfer curves up to 1.5 times the pile working load, regardless of whether the q–w function was strain hardening or stiffening. Accurate matching with measured load–settlement and load-transfer curves for 1.5–3 times working load was conditional on the correct choice of the q–w function. The models were further tested against 27 published load test results from across the world.