Computer simulation studies of dense suspension rheology : computational studies of model sheared fluids : elucidation, interpretation and description of the observed rheological behaviour of simple colloidal suspensions in the granulo-viscous domain by non-equilibrium particulate dynamics
Rheological properties of idealised models which exhibit all the non-Newtonian flow phenomenology commonly seen in dense suspensions are investigated by particulate-dynamics computer-simulations. The objectives of these investigations are: (i) to establish the origins of various aspects of dense suspension rheology such as shear-thinning, shear thickening and dilatancy; (ii) to elucidate the different regions of a typical dense suspension rheogram by examining underlying structures and shear induced anisotropies in kinetic energy, diffusivity and pressure; (iii) to investigate the scaling of the simplest idealised model suspension; i.e. the hard-sphere model in Newtonian media and its relationship to the isokinetic flow curves obtained through non-equilibrium molecular dynamics (NEMD) simulations; (iv) to preliminarily determine the effect of perturbations present in all real colloidal suspensions, namely particle size polydispersity and a slight 'softness' of the interparticle potential. Non-equilibrium isokinetic simulations have been performed upon ;systems of particles interacting through the classical hard-sphere potential and a perturbation thereof, in which the hard-core is surrounded by a 'slightly soft' repulsive skin. The decision to base the present work upon isokinetic studies was made in order to obtain a better under- standing of suspension rheology by making a direct connection with previous NEMD studies of thermal systemst(93). These studies have shown that the non-linear behaviour exhibited by these systems under shear is atttributable to a shear-induced perturbation of the equilibrium phase behaviour. The present study shows this behaviour to correspond to the high shear region of the generalised suspension flow curve.