An experimental study on viscous resuspension

Daniel Caichac, Aldo Tamburrino, Christian Ihle

Tuesday 30 june 2015

12:42 - 12:45h at Mississippi (level 1)

Themes: (T) Sediment management and morphodynamics, Poster pitches, (ST) Sediment transport mechanisms and modelling

Parallel session: Poster pitches: 5A. Sediment - Erosion

Resuspension of an initially settled layer of particles has been observed not only under the action of shear by a turbulent flow, but also by the action of a laminar shear flow. The paper reports the results and analysis of a set of experiments performed in a fully developed laminar Hagen-Poiseuille channel flow of clear fluid over an initially settled layer of solid particles, both monodisperse and slightly bidisperse (dmax/dmin=1.4) mixtures. A total of 30 experiments were carried out, for particle volume fraction __s between 0.05 and 0.3 and flows with a mean velocity between 1.2 and 0.4 cm/s. Resuspension heights of the particles forming the bed were recorded and compared with the ones obtained by previously developed theoretical models by Leighton & Acrivos (1986) and Shauly et al. (2000) related to hydrodynamic segregation. It was found that the model proposed by Leighton & Acrivos had an acceptable agreement with the experimental results, even though it doesn’t have any adjustable parameter. On the other hand the model proposed by Shauly et al., despite didn’t show an initial good fit with the data. Shauly et al.’s model contains an adjustable parameter K, which is a dimensionless empirical coefficient. By means of dimensional analysis, a relationship was found for K, as a function of the solids volume fraction __s and a modified Shields number _, which is a measure of the relative importance of viscous forces to those of gravity. Considering a modified parameter K, now depending on the flow and volume concentration, the theoretical model by Shauly et al. showed a good agreement with the experimental data, better than that of Leighton and Acrivos (1986). The results of the experiments with bidisperse suspensions were also compared with those theoretical models considering a monodisperse layer but adjusting the maximum particle volume concentration __0 with the models proposed by Shauly et al. (1998) and Gondret & Petit (1997), obtaining a good agreement between measured and computed data. The theoretical velocity profiles that were obtained with simplified numerical solutions of the previous models had a good agreement with the ones measured by Schaflinger et al. (1995) for a similar geometry.