The development and numerical modelling of a Representative Elemental Volume for packed sand

Abstract

The motivation of this thesis is the development of simple microscopic-scale model (representative elemental volume; REV) that can be used to conduct flow and heat transfer simulations from which closure coefficients can be established for the volume-averaged transport equations for porous media (packed bed). The thesis provides a brief introduction to the computational technique adopted for the geometric generation of the REV (YADE), followed by a parametric study undertaken to reveal the minimum number of particles inside the REV that are required to mimic the appropriate physics. Additional analysis was conducted with the goal of determining the influence of deviation in particle diameter. A wide range of particle temperatures was considered for analyzing the convective heat transfer and hydrodynamic behavior such the influence of property variation on temperature could also be studied. The key research output is a detailed comparison of the results related to the hydrodynamics and heat transfer closures produced by an idealized, YADE-generated model against a three-dimensional, digitized model of packed sand. It is shown that the YADE model gives results that are in very good agreement with the digitized packed-sand model in the range of Reynolds numbers considered. The YADE model results were then compared with experimental findings and it was found that while the trends in convective heat transfer were well-predicted, there was a difference in the magnitude of the convective coefficients predicted and measured in previous experiments.

Bayesian calibration of GPU–based DEM meso-mechanics Part I: Parallelization of RVEs

“…However, the packing fraction at the wall is generally lower than the rest of the sample, and the stress response at the wall is different from that of the center [8]. Periodic boundary conditions (PBC) are used in RVEs to reduce the effects of rigid wall boundaries [9].…”

Study of the evolution of transport properties induced by additive processing sand mold using X-ray computed tomography

“…(Soulaine, 2015) used FVM to perform a direct numerical simulation of fluid flow in a fully saturated porous media for the prediction of permeability from µ-CT image. (Thabet and Straatman, 2018) idealized a geometrical introduced to characterize small volume of packed sand in pore-level computations using YADE (Yet Another Development Engine) solver. In previous research by (Boek and Venturoli, 2010), it was proved that using LBM, to obtain permeability from a digital image is a reliable alternative to destructive traditional measurements.…”