To determine the Huggins coefficient in Eq. 14, a set of simulations were carried out using 1, 3, 5, 10, 17, and 25 monosize spheres. In this case the highest solid fraction was φ ≈ 0.2. Figure 3 shows the simulation data and, for comparison, experimental data based on sheared suspensions of silica particles [de Kruif et al. (1985)] is included. Clearly the agreement with experiment appears quite good in the regime shown here. The intrinsic viscosity, obtained from the intercept of the vertical axis is consistent with that obtained using a single sphere as described in the previous subsection. The Huggins coefficient, obtained from the slope, is in good agreement with predictions of KH ≈ 6. Since the Huggins coefficient results from an effective interaction between spheres, this is important confirmation that the code does account reasonably well for longer range hydrodynamic interactions in the fluid.
Figure 3. Determination of the intrinsic viscosity (y intercept and Huggins coefficient (slope) for a semidilute suspension. The solid circles represent simulation data and the +'s are derived from experiment. The lines correspond to a Huggins coefficient of 7 (solid) and 5 (dashed). Statistical uncertainties in the simulation data were approximately 10% or smaller.