Investigation of moment models for population balance equations and radiative transfer equations

More about the book

This thesis explores moment approximations for population balance and radiative transfer equations, focusing on laser-induced thermotherapy for liver tumors. Moments are created by averaging over a variable against basis functions to generate spectral approximations. In the context of the population balance equation, three closures are examined: minimum entropy, polynomial, and Kershaw closures. The polynomial closure is purely spectral, but it cannot guarantee the realizability of moments, meaning the underlying distribution function may not remain non-negative. The minimum entropy method addresses this issue using a non-negative ansatz, though it incurs higher computational costs due to the need for an optimization algorithm to reconstruct the distribution function. The Kershaw closure, which preserves realizability, employs a linear combination of Dirac-delta distributions informed by realizability boundaries. The study also considers full and partial linear moment models alongside the diffusion approximation. To enhance the accuracy of the steady-state radiative transfer equation, an iterative solution scheme is accelerated using a multigrid approach based on Lebedev quadrature rules. Additionally, a more relevant model is proposed using the incompressible Navier-Stokes equation to simulate blood and cooling fluid flow, addressing the heterogeneous nature of the heat and radiative transfer equations. The Runge-Kutta discon