Files in this item



application/pdfPAN-DISSERTATION-2020.pdf (10MB)Restricted Access
(no description provided)PDF


Title:Development of a high-fidelity thermochemical model for the O2+O system using direct simulation Monte Carlo
Author(s):Pan, Tzu-Jung
Director of Research:Stephani, Kelly A
Doctoral Committee Chair(s):Stephani, Kelly A
Doctoral Committee Member(s):Glumac, Nick; Lee, Tonghun; Panesi, Marco; Panerai, Francesco
Department / Program:Mechanical Sci & Engineering
Discipline:Mechanical Engineering
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):direct simulation Monte Carlo
non-equilibrium flow
Abstract:The presented work is about the investigation of reactions in a thermo-chemical non-equilibrium condition, which is especially with the application with the flows encountered during atmosphere entry/re-entry missions. The particle-based computational method, direct simulation Monte Carlo(DSMC), has been widely used for the hypersonic study. However, due to the lack of detailed information about particles’ collision, the developed DSMC chemical models in the past are usually with the assumption that the rotational and vibrational energy modes of gas molecules are in equilibrium with its translational mode, i.e. the rotational and vibrational energy distribution are assumed as Boltzmann distribution. Therefore, the objective of this work is to develop a ``state-to-state’’ DSMC model which is applicable for the strong thermo-chemic non-equilibrium flows and is used to investigate the O2+O system. The work is done with three aspects for DSMC: (i) develop a rovibrationally resolved state-to-state model for energy transfer, dissociation and recombination reactions; (ii) investigate the coarse-grain models to improve the expensive computational cost of the state-to-state model; (iii) examine the state-specific elastic collision models. The development of the state-to-state model for DSMC includes three aspects: (i) investigate the state-specific cross-section databases for O$_2$ + O system; (ii) develop a compatible gas-phase recombination model with a state-specific framework;(iii)construct a general DSMC procedure with the state-specific framework. Two cross-section databases by Esposito/Capitelli and Andrienko/Boyd are employed to examine the effect of vibrational and rovibrational nonequilibrium, respectively. It is discovered that the widely used phenomenological models in the past cannot capture the mono-quantum jump preference of rovibrational transition cross-sections, and the dissociation/recombination cross-sections have a strong dependency to the rovibrational energy state the molecule dissociated from/recombined to. It also is discovered that the equilibrium dissociation/recombination rate coefficients of the vibrationally resolved database are smaller than rovibrationally the resolved database in DSMC, which is mainly due to the nature of the coarse-grain process in DSMC. Two types of coarse-grain models are examined in this work. The vibration-based method merges all rovibrational states which share the same vibrational quantum number into the same bin; while the energy-based methods combine the states based on its rovibrational energy. For energy-based methods, the type of uniform rovibrational collisional coarse-grain (URVC) models are considered in this work. The fixed-size, variable-size, and adaptive-size binning strategies for URVC models are introduced. In the spatially homogeneous, rovibrational heating/cooling test cases, it is discovered that the vibration-based method behaved worse than the energy-based methods with a similar number of bins in both cases. For energy-base bins, it is found that using 20 bins is sufficient to reproduce the results of using a fully resolved database. With the same number of bins, the fixed-size model predicts the smallest error in the relaxation rate. The variable- and adaptive-size bins have a similar relaxation rate in the heating case, however, it is found that the adaptive-size bin performs better in the cooling case. Finally, the state-specific total collision cross-sections are investigated to construct a comprehensive and consistent treatment for state-to-state DSMC. The methodology of defining a cut-off scattering angle in quasi-classical trajectory calculation to evaluate the state-specific total collision cross-sections and scattering patterns are proposed, denoted as ``true'' data. The state-specific VSS parameters computed by collision-integral fit and cross-section fit are also proposed. It is found that the ``true'' total cross-section is an order of magnitude higher than VSS-type models. The distribution of scattering angles of ``true'' data is most within a range of small angles, which is much different from the VSS-type models. The simulation procedure using the ``true'' data is developed and implemented in the mature DSMC code, SPARTA. It is found that the result of VSS-type models are influenced by the rovibrational temperature, while the ``true'' data shows a subtle influence by rovibrational temperature.
Issue Date:2020-09-29
Rights Information:Copyright 2020 Tzu-Jung Pan
Date Available in IDEALS:2021-03-05
Date Deposited:2020-12

This item appears in the following Collection(s)

Item Statistics