Effective Density and Viscosity of A Suspension

This paper presents results of a series of experiments on the settling velocity of spheres in two‐component solid–liquid suspensions. Both sedimenting and fluidized suspensions are considered. Particular emphasis has been given to the effective values of the density and viscosity of the mixture which allow us to describe slow settling of a test sphere using the Francis [Physics 4, 403–406 (1933)] modification of the Stokes formula for the settling of a single sphere in a Newtonian fluid with wall effects. Our experimental results show that values of the effective density and viscosity are close to the average density of the mixture and to the viscosity of the mixture predicted by a correlation suggested by Thomas [J. Colloid. Sci. 20, 267–277 (1965)] when the test particles are of the same size as the suspended particles, or larger, but not too much smaller. The signature of the failure of the concept of effective fluid properties is traced to an increase of the variance of the experimental results and is not associated merely with a different than average value of the density; in particular, none of our results are well described by the use of the fluid density in the expression for the buoyant force. We also studied the settling of spheres in fluidized suspensions at high Reynolds numbers. The Barnea and Mizrahi [Chem. Eng. Sci. 5, 171–189 (1973)] model for the expansion of uniformly fluidized beds was modified in two different ways to describe the settling of particles in the fluidized suspensions. It is not possible to describe the modified drag laws with an effective Newtonian theory if for no other reason than the effective densities which enter into buoyancy are not in principle those which control inertia. The modified drag law is in agreement with experiments in sedimentation columns. Coarse agreement between the predicted and measured drag is found in the fluidized suspension even when the test and the fluidized particles are the same size.