Relating Atterberg limits to rheology

Abstract

The combination of rheology and soil mechanics is a relatively rare occurrence. Both fields are generally applied in quite separate circumstances, not usually at the same time. However, the transport and storage of thickened tailings slurries has created a situation in which both of these fields do come together. Whilst a thickened tailings slurry is flowing, it can be considered as a fluid. Rheometric equipment can be used to measure its flow properties, and rheological models and theories can be successfully applied to describe its behaviour. Once the tailings slurry is discharged into a storage facility, it typically flows across a 'beach' of previously deposited tailings, and eventually comes to rest. From this point onwards, the tailings particles are often considered as a soil, and the models and theories of soil mechanics then apply in describing the behaviour of the material. Tailings engineers often measure both the rheology and plasticity (by the use of the Atterberg limits test) of tailings materials in order to design for the transport and containment of the tailings, but sometimes samples are not available for testing, and documented lab test data for only one of these two aspects may exist. This paper presents new empirical relationships for the estimation of one from the other, based on a data set featuring 26 different tailings samples. Much of the observed variability in rheology can be accounted for by the plasticity of the material. 1 Introduction Rheology is the study of flowing materials. Whilst rheology may be theoretically contemplated at an atomic scale, it is more often considered in an empirical context, in which the resistance to flow of a fluid at various temperatures, rates of shear and shear histories is used to define the flow behaviour of the fluid, without the need to consider any of the physical aspects of what is actually happening in the fluid at a microscopic scale. Rheology has primarily been exploited in the laboratory characterisation of the flow behaviour of various fluids, through the use of rheometers (also called viscometers). The data gathered by these methods can be of practical value in predicting pumping requirements for viscous fluids such as slurries. It can also be useful in predicting slumping, sheet flow, channel flow and extrusion behaviours of various fluids, particularly those that exhibit non-Newtonian flow characteristics (varying viscosities at different rates of shear and shear histories). The question of relating slurry concentration and yield stress has presented itself as a challenge to rheologists in the past. It has long been recognised that factors related to particle size and mineralogy have significant influences on the rheology of soil particle slurries, e.g. Sofrá and Boger (2002). Many rheologists have regarded these variations as material specific and only quantifiable by actual rheology testing. Johnson et al. (2000) and Zhou et al. (2001) experimentally investigated the effects of van der Waal's forces and other forces acting on slurry particles at a microscopic scale, but did not go so far as to provide any methods of predicting slurry rheology or yield stresses. Stickel et al. (2006) presented a complex model for the prediction of yield stress using complex finite element computational analysis of the Brownian forces and charged particle interactive forces acting on individual particles in suspension, but the limits of computational power currently prevent their model from being applied to full scale practical situations, due to the large numbers of particles that must be simulated.