No abstract

The stress–strain behaviour of sands is highly nonlinear, even at stresses well below the peak strength of the sand. The hyperbolic model is a reasonable conceptual model for representing the stress–strain behaviour of sand, but some empirical curve fitting is required to obtain a more realistic model for calculation purposes. This can readily be performed for reconstituted samples of sand using laboratory tests. Recent evidence shows that the stiffness of natural sands is often much greater than that of the same sand when reconstituted at the same density and stress state in the laboratory, and it is therefore necessary to use in situ testing methods to determine the stress–strain behaviour of such sands. In this paper, the finite element method is used to simulate pressuremeter tests in a soil modelled using a hyperbolic-type model. It concentrates on the behaviour in unload-reload loops, which are often included in pressuremeter tests to measure shear modulus. The effect on the whole unload-reload loop of varying some of the model parameters is examined. The results are compared with a high-quality pressuremeter test in sand. It is concluded that, though the results to date are encouraging, some further experimental work is required to verify some of the features of the model. Key words : pressuremeter test, hyperbolic model, nonlinear behaviour, initial shear modulus, sand behaviour.

No abstract

No abstract

During the placement of fine-grained cemented mine backfill, the high placement rates and low permeability often result in undrained self-weight loading conditions, when assessed in the conventional manner. However, hydration of the cement in the backfill results in a net volume reduction—the volume of the hydrated cement is less than the combined volume of the cement and water prior to hydration. Though the volume change is small, it occurs in conjunction with the increasing stiffness of the cementing soil matrix, and the result in certain circumstances can be a significant reduction in pore-water pressure as hydration proceeds. In this paper, the implications of this phenomenon in the area of cemented mine backfill are explored. An analytical model is developed to quantify this behaviour under undrained boundary conditions. This model illustrates that the pore-water pressure change is dependent on the amount of volume change associated with the cement hydration, the incremental stiffness change of the soil, and the porosity of the material. Experimental techniques for estimating key characteristics associated with this mechanism are presented. Testing undertaken on two different cement–minefill combinations indicated that the rate of hydration and volumes of water consumed during hydration were unique for each cement–tailings combination, regardless of mix proportions.

In current underground mining practice, it is common to use tailings, without added cement, to fill mined-out voids ͑"stopes"͒. If fine-grained tailings are used, the high placement rates and low permeability can often result in undrained loading conditions and, hence, lower effective stress, when assessed in the conventional manner. Where cement is added, the cement modifies the consolidation characteristics in a number of ways, including increasing the strength and stiffness, reducing the permeability, and inducing volumetric changes associated with the hydration reactions leading to "self-desiccation." As a result, conventional consolidation-analysis techniques are unsuitable for assessing the behavior. The one-dimensional mine-tailings-consolidation program ͑MinTaCo͒ has been modified, and renamed CeMinTaCo, to couple cement hydration with conventional consolidation analysis. The fundamental theory behind the modifications is presented. The model is used to undertake a sensitivity study, which highlights some of the important features of the behavior of cemented backfill, and shows how complex interactions between the various properties produce some outcomes that are counterintuitive.

In current underground mining using “open stoping” methods, it is common to backfill mined-out voids (“stopes”) using hydraulically placed backfill, which is commonly composed of tailings, to which cement is often added. Knowledge of the stress state within a backfilled stope is required for safe design of drawpoint barricades and for other operational reasons. This stress state depends, inter alia, on the degree of “arching” that occurs, resulting from the development of shear stress between the fill and the stope walls. This paper presents a numerical modelling study of aspects of the arching phenomenon, carried using the computer code Plaxis. The backfill is characterized using the Mohr–Coulomb soil model, and both dry backfill and saturated backfill are considered to completely cover the full spectrum of backfill types that are used in practice. It is shown that even with dry backfill, the behaviour is governed by a complex interaction between the soil parameters. The behaviour is more complex with saturated backfill, with a key parameter being the permeability of the backfill relative to the rate of filling — i.e., whether the backfilling operation can be considered to be “drained” or “undrained” or somewhere between these two extremes.

The behaviour of granular materials when subjected to rotation of their principal axes of stress or strain is not fully understood. In the late 1980s a new apparatus was developed to allow general plane strain conditions to be applied to an analogue soil (Schneebeli cylinders). This paper presents a short description of this apparatus and its capabilities. Experimental results from different types of cyclic tests involving the continuous rotation of principal axes are given, mainly in terms of volume change and the co-axiality of stress and strain increment tensors. The results show that cyclic rotation induces a general tendency to compaction, and that the principal axes of stress and strain increment are not coincident as it is generally assumed in isotropic elastoplastic modelling.