A multiphysics modeling approach for heat conduction in metal hydride powdersis presented, including particle shape distribution, size distribution,granular packing structure, and effective thermal conductivity. A statisticalgeometric model is presented that replicates features of particle size andshape distributions observed experimentally that result from cyclic hydridedecreptitation. The quasi-static dense packing of a sample set of theseparticles is simulated via energy-based structural optimization methods. Theseparticles jam (i.e., solidify) at a density (solid volume fraction) of0.665+/-0.015 - higher than prior experimental estimates. Effective thermalconductivity of the jammed system is simulated and found to follow the behaviorpredicted by granular effective medium theory. Finally, a theory is presentedthat links the properties of bi-porous cohesive powders to the present systemsbased on recent experimental observations of jammed packings of fine powder.This theory produces quantitative experimental agreement with metal hydridepowders of various compositions.
展开▼
