An overview of phonon-based heat conduction models and their solution

Abstract

Purpose– The purpose of this paper is to provide an overview and some recent advances in the models, analysis and simulation of thermal transport of phonons as related to the field of microscale/macroscale heat conduction in solids. The efforts focus upon a fairly comprehensive overview of the subject matter from a unified standpoint highlighting the various approximations inherent in the thermal models. Subsequently, the numerical formulations and illustrations using the current state-of-the-art are provided.Design/methodology/approach– This paper is dedicated to the approximate solution to the relaxation time phonon Boltzmann equation (BE). While original contributions are pointed out and addressed appropriately, the efforts and contributions will be focussed on a relatively complete overview highlighting the field from one unified standpoint and clearly stating all assumptions that go into the approximations inherent to existing models. The contents will be divided as follows: In the first section the authors will give an overview of semi-classical phonon transport physics. Then the authors will discuss the equation of phonon radiative transport (EPRT) and its approximations—the ballistic-diffusive approximation (BDA) and the new heat equation (NHE). Next the authors derive and discuss the C-F model. A numerical discretization method valid for all models is then presented followed by results to numerical simulations and discussion.Findings– From a unified treatment based on the introduction of an energy distribution function, the authors have derived the EPRT and its two well-known approximations: BDA and NHE. For completeness and to provide a vehicle for a general numerical discretization approach, the authors have also included analysis of the C-F model and the parabolic and hyperbolic descriptions of heat transfer along with it. The approximation of angular dependence of phonons in radiation-like descriptions of transport has been given special attention. The assumption of isotropy was found to be of paramount importance in the formulation of position space models for phononic thermal transport. For the thin film problem considered here, the NHE along with the proper boundary condition appears to be the best choice to approximate the phonon BE. Not only does it provide predictions that are in excellent agreement with EPRT, it does not require the discretization of phase space making it far more computationally efficient.Originality/value– The authors hope this work will help dispel the idea that since Fourier’s law describes diffusion (under limiting assumptions) and it has shown to be ineffective in describing heat transfer for very thin films, that diffusion cannot describe heat transfer in thin films and one should look to a radiative description instead. If one considers diffusion in the sense of random motion, as invisaged by the original builders of the subject (Smoluchowski, Einstein, Ornsteinet al.), instead of a temperature gradient, the idea that diffusion can govern thermal transport at this scale is not surprising. Indeed, the NHE is essentially a diffusion equation that describes the motion of particles up to the point of true randomness (isotropy) as well as thereafter.