Multi-scale particle transport modeling: Kinetic high-fidelity DSMC approaches and system-scale Eulerian-Lagrangian simulations for high-speed flows

Marayikkottu Vijayan, Akhil

Permalink

https://hdl.handle.net/2142/121974 Copy

Description

Title

Multi-scale particle transport modeling: Kinetic high-fidelity DSMC approaches and system-scale Eulerian-Lagrangian simulations for high-speed flows

Author(s)

Marayikkottu Vijayan, Akhil

Issue Date

2023-11-22

Director of Research (if dissertation) or Advisor (if thesis)

Levin, Deborah A

Doctoral Committee Chair(s)

Levin, Deborah A

Committee Member(s)

• Villafane Roca, Laura
• Curreli, Davide
• Dreizin, Edward L

Department of Study

Aerospace Engineering

Discipline

Aerospace Engineering

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Multiphase flows
• Direct Simulation Monte Carlo
• Eulerian-Lagrangian

Abstract

The study of gas-particulate multiphase flows is indeed of significant importance in various aspects of life, including natural phenomena, industrial applications, and inter-planetary missions. These systems can be highly complex and occur in scenarios such as volcanic eruptions, coal mine explosions, fluidized reactors, combustors, and interactions between lander plumes and planetary dust in inter-planetary missions. One of the challenges in studying these multiphase systems is that research advancements have been more focused on low-speed, continuum regimes, while high-speed regimes have received less attention. In situations where the particle dimensions are comparable to the gas mean free path, the system becomes non-continuum, adding another layer of complexity. To approach the study of such systems, researchers classify them based on the characteristic length scale of interest. For larger length scales, where shock-dominated, high-speed multiphase systems are involved, an Eulerian-Lagrangian numerical approach is utilized. This approach allows researchers to study the temporal evolution of both gas and particulate phases in these systems. However, when dealing with much smaller length scales on the order of micrometers, where the gas mean free path becomes a significant factor, the traditional continuum-based methods may not be suitable. In such cases, the Direct Simulation Monte Carlo (DSMC) method by G.A. Bird is employed. The DSMC method is a kinetic approach that models the interactions of individual gas molecules and particles, providing a more accurate representation of high-speed, non-continuum regimes. By adopting a combination of numerical approaches, researchers can gain deeper insights into the behavior of gas-particulate multiphase systems across different length scales and flow regimes. This knowledge is crucial for advancing our understanding of natural phenomena, optimizing industrial processes, and ensuring the success and safety of inter-planetary missions. The development of the Eulerian-Lagrangian multiphase code is based on the framework of the FLASH research code, which was created at the Argonne National Laboratory. The Lagrangian particle modules are integrated into the FLASH code as modular subroutines, aligning with the parallelization strategy of the original framework. This allows for efficient handling of large-scale simulations involving gas-particulate multiphase flows. The code is versatile and capable of handling a wide range of particle concentrations, from dilute to dense. It incorporates both one-way and two-way coupling approaches for interactions between the gas and particulate phases. In cases of moderate particulate concentrations, where the dusty gas assumption is applicable, inter-particulate interactions are modeled as binary inelastic DSMC collisions. The DSMC method is a kinetic approach that accurately captures the behavior of gas molecules and particles at low densities. For dense regimes with high particulate concentrations, the code employs the multiphase particle in cell (MP-PIC) approach to model inter-particle collisions or interactions. The MP-PIC method allows for a detailed representation of particle-particle interactions and their influence on the overall flow behavior. The code's validity is established through validation against various analytical and experimental gas-particulate flow cases. This ensures that the numerical results are reliable and representative of real-world phenomena. Additionally, an in-house GPU-CPU hybrid code named CHAOS is used to model gas flow through micron-scale particles with complex shapes and configurations. This code leverages first-principle based modeling of gas-surface interactions, enabling accurate predictions of momentum transfer between gas and particulate phases. As a result, the forces and moments acting on the multiphase systems can be quantified and understood. CHAOS is also employed to study mobility parameters for particles with intricate shapes and the aerodynamic interactions between closely packed monodispersive particulate systems subjected to compressible, rarefied gas flows. These investigations contribute to a better understanding of gas-particulate multiphase systems. The first part of the thesis focuses on large-scale simulations using the Eulerian-Lagrangian (EL) FLASH code. The code was initially used to study the lifting behavior of particulates when exposed to a moving planar shock wave in a one-way coupled approach. The simulation results were validated against experimental data by Merzkirch et al. and numerical studies by Gosteev et al., showing good agreement. Building on this confidence in the code's capabilities, the thesis proceeded to investigate the lifting of a monolayer of spherical particulates when subjected to an expanding blast wave generated by an electrostatic discharge. The study identified the gas velocity shear induced Saffman lift force as the primary lifting mechanism for this system. The numerical results were compared with a physical experiment conducted with research collaborators at the New Jersey Institute of Technology, and a good agreement was observed. The research also explored the impact of varying the distance between the wall and the electrode on the particle lifting height and velocity. It was observed that increasing the wall-electrode distance led to a decrease in both the lifting height and velocity of the particles. To further enhance the capabilities of the EL code, inelastic DSMC collision modules were incorporated into the solver. This allowed the study of the evolution of a particle-laden blast wave (PLBW) system, where the effects of inter-particulate collisions and aerodynamic interactions on the particle front were analyzed. It was found that in the early stages of PLBW evolution, aerodynamic interactions tended to accelerate the particle front, while inter-particulate collisions had a decelerating effect. However, in the later stages of the system's evolution, a significant presence of inter-particulate collisions resulted in higher magnitudes of kinetic energy for the particulate phase. Finally, the EL approach was extended to study particle lifting from dense beds of particulates exposed to blast waves generated by electrostatic discharges. In this case, a two-way coupling approach was adopted due to the increased particulate concentration. Unlike the monolayer study, the effects of drag, pressure-gradient, and added-mass were found to dominate the behavior of the dense particulate system. Overall, the research presented in the thesis demonstrates the capability and versatility of the Eulerian-Lagrangian FLASH code in investigating various gas-particulate multiphase flow phenomena, ranging from lifting mechanisms to complex interactions in dense particle systems exposed to blast waves. The numerical results obtained provide valuable insights into these complex phenomena and show good agreement with experimental data, validating the code's reliability. The second part of the thesis focuses on particle-scale simulations aimed at understanding the mobility parameters of particulates through momentum transfer between phases. The Direct Simulation Monte Carlo (DSMC) code CHAOS was extended with subsonic boundaries to handle the subsonic parametric spaces considered in these studies. To validate the approach across different Mach numbers and Knudsen numbers, an extensive study of drag developed on spherical particulates was conducted. The results showed that the approach accurately captures the effects of compressibility and rarefaction on particulate mobility, and the drag parameters agreed well with the formulation of Loth, which served as the base model for further studies. The thesis then investigated drag, lift, and pitching moments developed on irregular fractal-like aggregates. It was found that the drag on these particulates consistently exceeded that of a single sphere or monomer under the same flow conditions. The lift force generated on these particulates due to their irregular shape decreased with increasing Mach number, as the gas tended to perceive the particulates as approximately spherical entities at these conditions. Additionally, it was observed that the shape effect on particulate mobility decreased with increasing gas rarefaction or Knudsen number. Furthermore, the study explored the mobility of particulate distributions. The results revealed that distributions of particulates experienced different fluid-induced drag and lift compared to an isolated particulate under the same conditions. The average drag on particulate distributions showed a non-linear trend with volume fraction. For a volume fraction of 1 %, the average drag was higher than that of a corresponding single sphere, but as the volume fraction increased further, the average drag decreased. The thesis also examined the criteria for the formation of separated versus collective shocks in aerodynamic systems exposed to supersonic gas flows. This analysis shed light on the shock dynamics in such systems and provided insights into the behavior of particulate distributions in high-speed flows. Overall, the particle-scale simulations presented in the second part of the thesis contribute to a better understanding of mobility parameters for particulates, drag and lift forces on irregular particulate shapes, and the behavior of particulate distributions in various flow conditions. These findings have implications in fields such as aerodynamics, particle transport, and environmental sciences.

Graduation Semester

2023-12

Type of Resource

Thesis

Copyright and License Information

Copyright 2023 Akhil Marayikkottu Vijayan

Owning Collections