No abstract

No abstract

Abstract. This paper summarizes a method to characterize conduction properties in soils at the particle-scale. The method set the bases for an alternative way to estimate conduction parameters such as thermal conductivity and hydraulic conductivity, with the potential application to hard-to-obtain samples, where traditional experimental testing on large enough specimens becomes much more expensive. The technique is exemplified using 3D synthetic grain packings generated with discrete element methods, from which 3D granular images are constructed. Images are then imported into the finite element analyses to solve the corresponding governing partial differential equations of hydraulic and thermal conduction. High performance computing is implemented to meet the demanding 3D numerical calculations of the complex geometrical domains. The effects of void ratio and inter-particle contacts in hydraulic and thermal conduction are explored. Laboratory measurements support the numerically obtained results and validate the viability of the new methods used herein. The integration of imaging with rigorous numerical simulations at the pore-scale also enables fundamental observation of particle-scale mechanisms of macro-scale manifestation.

No abstract