ORCID Identifier(s)

ORDID 0009-0000-1801-6244

Graduation Semester and Year

Summer 2024

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Mechanical Engineering

Department

Mechanical and Aerospace Engineering

First Advisor

Albert Y. Tong

Second Advisor

Chaoqun Liu

Third Advisor

Miguel Amaya

Fourth Advisor

Dora E. Musielak

Fifth Advisor

Dezhi Dai

Abstract

Two-phase flows are critical in a variety of engineering applications that span multiple length scales. These applications encompass macroscopic phenomena, such as dam breakage and wave-structure interactions, as well as microscopic events, including droplet impacts and boiling, and mixed-scale issues like spray dynamics. Accurate representation of evolving interfaces is essential for numerical simulations of these problems, and the Volume of Fluid (VOF) method combined with the Piecewise Linear Interface Calculation (PLIC) approach is employed to fulfill this requirement. To tackle the computational challenges associated with high-resolution meshes in deforming two-phase flows, Adaptive Mesh Refinement (AMR) is utilized to enhance mesh resolution near interfaces. The accuracy and efficiency of VOF simulations with dynamic mesh refinement are influenced by various factors, including AMR load balancing, VOF field initialization, background mesh sizes, AMR tolerances, AMR levels, buffer layers, and VOF field mapping schemes.

A shell script-based load balancing method is proposed for parallel VOF simulations utilizing dynamic mesh refinement in OpenFOAM. The methodology was evaluated through a typical dam-breaking problem to analyze CPU time consumption during two-phase flow simulations, both with and without load balancing. Various simulations were conducted with different load balancing time steps. The results indicate that the proposed load balancing method significantly enhances computational efficiency. Although there is increased overhead associated with decomposition and reconstruction, the overall computational time was markedly lower in the load-balanced scenarios, underscoring the effectiveness of load balancing in optimizing resource utilization.

This study investigates the effects of VOF field initialization, AMR tolerances, AMR levels, and VOF field mapping schemes by solving a spherical interface advection problem within a non-uniform three-dimensional flow field. The VOF field initialization using AMR is compared to the traditional approach that does not utilize AMR. The results indicate that initializing the VOF field with AMR yields significantly smoother interfaces, thereby highlighting the advantages of this method in accurately capturing interface dynamics. Further analysis focuses on the accuracy and efficiency of various AMR tolerances. By adjusting these tolerances, the study assesses how refinement criteria impact both the precision of the solution and computational performance. Additionally, the investigation explores the number of AMR levels, providing insights into the trade-offs between computational cost and the detail captured in the interface. Finally, the effectiveness of two VOF field mapping schemes is evaluated regarding their ability to preserve interface continuity and accuracy throughout the advection process.

Furthermore, this study systematically evaluates the impact of various AMR control parameters, including background mesh resolution, AMR tolerances, AMR levels, and buffer layers, across a range of engineering applications such as dam breakage, wave-structure interaction, droplet impact on a liquid pool, and spray in cross-flow scenarios. The research proposes optimal AMR parameters tailored for different two-phase flow scales. This comprehensive analysis provides valuable insights into optimizing the performance of the PLIC-VOF method and enhances the understanding of its interaction with AMR parameters in diverse flow regimes. Ultimately, this research advances the accuracy of two-phase flow simulations and offers critical guidance for numerical simulations of two-phase flows with dynamic mesh refinement across a wide spectrum of engineering applications.

Keywords

Multiphase flows, Volume of fluid (vof), Adaptive mesh refinement (amr), Load balancing, Droplet, Wave, Spray, Meshing, Plic

Disciplines

Aerodynamics and Fluid Mechanics | Computational Engineering | Ocean Engineering | Other Mechanical Engineering

Download

Available for download on Thursday, August 13, 2026

Share

COinS