Author(s): Christoph Rettinger; Sebastian Eibl; Ulrich Rude; Bernhard Vowinckel

Linked Author(s): Christoph Rettinger

Keywords: Sediment transport; Rheology; Dense suspension; Direct numerical simulation; Lattice-Boltzmann

Abstract: We investigate the rheological behavior of sheared sediment beds composed of polydisperse spherical particles in a laminar Couette-type shear flow. Classical scaling relationships for rheological quantities such as the macroscopic friction factor, the maximum volume fraction and the effective viscosities have become increasingly popular for closures of two-phase flow modeling, but rheological models have, thus far, been derived for monodisperse particles only. Our aim is to extend these considerations to a more realistic polydisperse particle distribution that covers the variance of fine to coarse sand grains, which corresponds to a non-uniformity factor of 10 separating the smallest from the largest grains. To this end, we present the time-averaged depth-resolved profiles of the relevant physical quantities that determine the rheology models, i.e. local shear rate of the fluid, particle volume fraction, total shear and granular pressure. The data was generated by means of fully coupled, grain resolved Direct Numerical Simulations using a combined Lattice-Boltzmann discrete-element method [1-3]. A total of 4 massively parallelized simulation runs of different uniformity with up to 26,000 particles were carried out on a supercomputer with 7680 processors and a runtime of 48 hours each. We compare our results against seminal rheology modeling frameworks of effective viscosities and the macroscopic friction factor. Hereby, we find good agreement for the monodisperse benchmark case, which demonstrates the applicability of our simulation approach to systematically investigate the effect of polydispersity on the rheological model. We utilize our results to improve upon classical rheological models by enhancing the parameterization of the empirically derived coefficients. Our data suggests that simple scaling laws can be used to predict these coefficients by explicitly taking the effect of polydispersity into account. [1] Bauer, M., Eibl, S., Godenschwager, C., Kohl, N., Kuron, M., Rettinger, C., Schornbaum, F., Schwarzmeier, C., Thönnes, D., Köstler, H. and Rüde, U., 2020. waLBerla: A block-structured high-performance framework for multiphysics simulations. Computers & Mathematics with Applications. [2] Eibl, S. and Rüde, U., 2018. A local parallel communication algorithm for polydisperse rigid body dynamics. Parallel Computing, 80, pp.36-48. [3] Rettinger, C. and Rüde, U., 2020. An efficient four-way coupled lattice Boltzmann-discrete element method for fully resolved simulations of particle-laden flows. arXiv preprint arXiv:2003.01490.

DOI: https://doi.org/10.3850/IAHR-39WC25217119202253

Year: 2022