Research
Our research uses large-scale stochastic molecular simulation as a bridge between computational/experimental chemistry and continuum gas dynamics. We have developed a range of modeling strategies for nonequilibrium gas flows and gas-surface interactions where the model-regimes overlap. This allows for rigorous scaling-up of computational chemistry (the study of small groups of atoms), through large-scale stochastic molecular simulation, to continuum gas dynamics and thermodynamics. Such a strategy is enabled by the inherent statistical nature of dilute gases combined with modern computer power. We collaborate with experimental researchers studying physical processes at both molecular and macroscopic scales. We use our multiscale simulation framework to primarily investigate hypersonic shock-layer physics and high temperature gas-surface chemistry for Air Force, Navy, and NASA applications. However, research also extends into areas such as aerosol science and microflows, surface catalysis, gas flow through porous media, as well as satellite drag and rocket plumes.
Molecular Gas Dynamics and Direct Simulation Monte Carlo (DSMC)
The direct simulation Monte Carlo (DSMC) particle method simulates a gas flow from the molecular perspective. Millions of simulated molecules move through the domain colliding with each other and with solid surfaces using molecular-level collision models. Macroscopic variables such as temperature, velocity, and pressure, result from averaging over the particles residing in each computational cell. Performing accurate and efficient simulations on large computer clusters (billions of particles) presents many exciting challenges.
Professor Schwartzentruber recently co-authored a textbook on the subject. The first half of the textbook focuses on nonequilibrium and molecular gas dynamics, while the second half of the textbook focuses on the direct simulation Monte Carlo (DSMC) particle method.
Iain D. Boyd and Thomas E. Schwartzentruber, Nonequilibrium Gas Dynamics and Molecular Simulation, Cambridge University Press, 2017.
DSMC simulation of a satellite re-entering the atmosphere under high-altitude, rarefied, conditions. We use the Molecular Gas Dynamic Simulator (MGDS) code, developed in our research group.
Gas-Surface Interactions and Chemistry
Many heat shield materials are carbon-fiber based with a complex microstructure. It is now possible to image actual heat shield microstructure using tomography and triangulate a surface grid for use in a DSMC simulation. Relative to the fiber (or pore) size, the reactive gas flow is in the transition regime between free-molecular and continuum. Of interest is the transport of reactive species (i.e. atomic oxygen) to the fiber surfaces and the oxidation reactions that result in loss of the heat shield material. Our group collaborates with experimentalists involved in molecular beam scattering, materials science and high temperature ceramics, and high-enthalpy plasma wind tunnel facilities.
Our research group is part of a new five-year NASA project that combines computational and experimental researchers across five universities (the Advanced Computational Center for Entry System Simulation - ACCESS). Our research uses molecular-level simulation and understanding to create advanced models for large-scale computational fluid dynamics (CFD) simulations.
Current research grants:
NASA "Advanced Computational Center for Entry System Simulation" -- collaborative project with University of Colorado (lead), University of Kentucky, University of Illinois, University of New Mexico State, and University of Minnesota (Candler group)
NASA "Predictive Simulations of Chemical and Structural Failure of Porous Ablative Materials" -- collaborative project with University of Kentucky
NASA "Air-Carbon Boundary Layer Chemistry for Hypersonic Ablation" -- collaborative project with University of Colorado, and Northwestern University
DSMC simulation of gas flow through a porous, carbon-fiber material, similar to that used in NASA heat shields. Fiber diameters are 10 microns and each pixel in the flow flow field is a single DSMC cell containing approximately 50 simulated molecules.
DSMC simulation of a high speed, high temperature, boundary layer flow over a porous, carbon-fiber material. Heat flux, shear, and atomic oxygen flux to the fiber surfaces is computed. Conditions match the NASA Stardust re-entry.
Nonequilibrium and Reacting Flows (Hypersonic Flows)
Each DSMC particle has a molecular velocity, a rotational and vibrational energy (classical or quantized), as well as a species type (N2, O2, N, O, NO, etc.). Instead of using stochastic collision algorithms, it is possible to perform trajectory calculations for all collisions (i.e. integrate each collision on a potential energy surface (PES) via molecular dynamics). In this manner, a macroscopic flow solution can be obtained where the sole model input is a PES that governs the interactions between atoms (provided by chemists/physicists). Such an approach directly connects the fields of computational chemistry and continuum aerothermodynamics. We refer to such calculations as Direct Molecular Simulation (DMS).
Current research grants:
NASA "Recombination Modeling in High Enthalpy Air Flow"
NASA "Air-Carbon Boundary Layer Chemistry for Hypersonic Ablation" -- collaborative project with University of Colorado, and Northwestern University
AFOSR "Spectroscopic Measurements and Nonequilibrium Modeling for High-Enthalpy Air" -- collaborative project with Caltech, Stanford, Truhlar and Candler groups (Minnesota)
Potential energy surface (PES) dictating the forces between N atoms (left). Individual collision simulation between two nitrogen molecules, using the PES (right).
Shock wave calculation in oxygen using Direct Molecular Simulation (DMS).
Particulate Interactions with Hypersonic Flow and Material Impact Damage
Professor Schwartzentruber is leading a Multidisciplinary University Research Initiative (MURI) project, funded by the Office of Naval Research (ONR), called "Particulate and Precipitation Effects on High-speed Flight Vehicles". This project includes computational and experimental researchers from five universities. The goal is to understand how small aerosol particles and large precipitation (cloud and rain droplets) transit through the shock-layer surrounding a hypersonic vehicle and to predict their properties upon impact. This research involves close collaboration with the Candler and Hogan groups at the University of Minnesota. Our research also aims to predict material damage caused by high speed particle impact through close collaboration with the University of Hawaii and by using their PISALE (Pacific Island Structured-amr with ALE) code. Finally, our group conducts similar research for NASA to understand heat shield erosion due to hypersonic entry into the atmosphere of Mars during a dust storm.
Current research grants:
ONR FY2020 MURI "Particulate and Precipitation Effects on High-speed Flight Vehicles" -- collaborative project with the Hogan and Candler groups (Minnesota), University of Hawaii, University of Maryland, University of Illinois, and Steven's Institute of Technology
NASA "Investigation of Particle Effects on a Hypersonic Mars Entry" -- NSTRF Fellowship for Michael Kroells
Time accurate DSMC simulation of a particle passing through a normal shock wave. Disturbances caused by the particles and deceleration prior to surface impact are studied. DSMC is appropriate since the particle size is comparable to the mean-free-path.
Table-Top Shock Tunnel (TTST) Development
A novel experimental facility for measuring shock layer physics and gas-surface chemistry, called the Table-Top Shock Tunnel is now operational in the Minton Lab at the University of Colorado, Boulder. The basic idea is that a pulsed molecular beam, when targeted at a small blunt object, can generate a hypersonic shock layer. Unlike large shock-tunnel facilities, the beam, and therefore the shock layer, can be generated (pulsed) 2 times per second for hours with repeatable conditions. Such test frequency and repeatability may enable existing optical diagnostic techniques to measure thermochemical quantities with unprecedented accuracy at a fraction of the cost compared to existing shock tunnel facilities. We perform DSMC simulations to help optimize the TTST facility, design new experiments, and validate our models against the unique experimental measurements.
Current research grants:
AFRL "Computational Modeling of a Table-Top Shock Tunnel Concept" -- collaborative project with the University of Colorado
[TOP] Image of the TTST in operation; emission from trace-species is visible to the eye. [BOTTOM] DSMC simulation of the shock layer surrounding a test article.
Hybrid Particle-Continuum Methods
High-altitude, hypersonic flows exhibit large variations in the local mean-free-path (the average distance travelled between molecular collisions). In regions where the mean-free-path becomes large (rarefied wakes) or becomes to comparable to small length scales of interest (sharp leading edges, shock waves, boundary layers, etc.), continuum assumptions inherent in CFD methods may become inaccurate. Research focuses on coupling DSMC and CFD methods within a single simulation. CFD algorithms are used to accurately and efficiently simulate continuum regions, and the DSMC method is restricted to non-equilibrium regions where increased physical refinement is required.
Current research grants:
ONR "Hybrid DSMC/CFD Method Development for High Altitude Hypersonic Flows" -- collaborative project with Candler group (Minnesota)
AFOSR "Validation of Hypersonic Flow Simulations via Molecular-Scale Physics" -- collaborative project with Candler group (Minnesota)
Hybrid DSMC-CFD simulation of hypersonic flow over a hollow-cylinder-flare geometry. Top shows a full DSMC simulation. Bottom shows a hybrid simulation where only DSMC regions are shown.
Microscale Rarefied Flows
In order to characterize nanoparticle pollutants, particles can be given a known charge and accelerated via a known electric field. The measured time of flight is then directly related to the drag force on the particles caused by the surrounding air. If this drag force can be related to nanoparticle cluster geometry then nanoparticle pollutants in a given sample could be characterized. Such flows lie in the regime between continuum and free-molecular (image right).
Other projects involve polymer transport and deposition, in the transition regime, for surface coatings and additive manufacturing,
