Date of Award

May 2026

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Mechanical Engineering

First Advisor

Chonglin D. Zhang

Abstract

This work presents a numerical investigation of plume–surface interactionsusing the Direct Simulation Monte Carlo (DSMC) method, coupled with the devel opment of an automated adaptive mesh refinement (AMR) framework for rarefied flow simulations. A two-dimensional axisymmetric configuration is used to simulate the impingement of a rarefied exhaust plume on a flat surface under vacuum condi tions representative of experimental test facilities. Surface pressure, shear stress, and heat flux induced by plume impingement are computed and compared with published experimental and numerical results to validate the DSMC approach. To address the strong mesh sensitivity and computational inefficiency associ ated with uniform discretization in DSMC simulations, an automated AMR frame work is developed. The refinement strategy enforces a local cell-size criterion based on the mean free path, ensuring DSMC validity while concentrating resolution in the plume core, near-surface interaction region, and other high-gradient zones. The AMR workflow is fully automated using Gmsh and bash-based scripting, enabling mesh generation, refinement, partitioning, and solver execution from a single user-defined input file. The framework is designed to work on both CPU and GPU platforms, supporting parallel execution across multiple GPUs for large-scale simulations. Results demonstrate that the automated AMR framework significantly im proves solution accuracy while reducing computational cost compared to uniformly refined meshes. The approach provides a robust and scalable tool for high-fidelity plume–surface interaction simulations and establishes a foundation for future exten sions to plume-driven lunar dust entrainment and transport.

Download

Share

COinS