About

Deborah Levin is a Professor in the Department of Aerospace Engineering at the University of Illinois Urbana-Champaign. She holds a PhD in Chemistry from the California Institute of Technology and a BS in Chemistry from the State University of New York at Stony Brook. Her academic career includes positions as a Research Professor and lecturer at George Washington University, Associate Professor and Professor at The Pennsylvania State University, and since August 2014, she has been a faculty member at Illinois. Her research areas encompass combustion and propulsion, computational fluid mechanics, hypersonics, nanosatellites, and space systems. Levin's work involves the development and application of kinetic modeling, molecular dynamics, and computational techniques to study hypersonic flows, plasma interactions, and space vehicle design. She has contributed to chapters in books and authored numerous articles in journals, focusing on non-equilibrium flows, shock interactions, and plasma-surface interactions, among other topics.

Research topics

• Physics
• Materials science
• Nanotechnology
• Chemistry
• Chemical physics
• Atomic physics
• Meteorology
• Optics
• Computational chemistry
• Nuclear physics
• Molecular physics
• Medicine
• Optoelectronics
• Mechanics
• Computational physics
• Classical mechanics
• Organic chemistry

Selected publications

• Calibration of the Larsen-Borgnakke Model for NO-O Vibrational Relaxation

  Springer aerospace technology · 2026-01-01

  book-chapterOpen access

  Abstract Modeling of Nitric Oxide (NO) vibrational states in hypersonic reentry flows is necessary to interpret emission and absorption spectra. To that end, the vibrational relaxation number, $$Z_v^C$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msubsup> <mml:mi>Z</mml:mi> <mml:mi>v</mml:mi> <mml:mi>C</mml:mi> </mml:msubsup> </mml:math> , is an important parameter in the Larsen-Borgnakke (LB) model which is used to express vibrational relaxation rates for collisions. In this work, we use rates obtained from Quasi-Classical Trajectory (QCT) calculations of Andrienko et al. [2] for the NO-O vibrational relaxation process to arrive at a new fit for the $$Z_v^C$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msubsup> <mml:mi>Z</mml:mi> <mml:mi>v</mml:mi> <mml:mi>C</mml:mi> </mml:msubsup> </mml:math> parameter. This new fit for $$Z_v^C$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msubsup> <mml:mi>Z</mml:mi> <mml:mi>v</mml:mi> <mml:mi>C</mml:mi> </mml:msubsup> </mml:math> is implemented in Direct Simulation Monte Carlo (DSMC) calculations for the case of hypersonic flow over a cylinder where the effect of the new fit on NO macroparameters and vibrational state populations is studied.

  PublisherOA PDFDOI

• Ab Initio Calculations of Processes in Low Energy Electron-Molecule Scattering

  Caltech Library · 2026-04-28

  articleOpen access1st authorCorresponding

  Chapter I Calculations are reported for low energy e-N2 scattering cross sections in the static-exchange approximation. The approach used involves solving the Lippmann-Schwinger equation for the transition operator in a sub-space of Gaussian functions. New features of the method are analytic evaluation of matrix elements of the free particle Green's function and analytic transformation to obtain single-center expansion coefficients for the scattering amplitude. Results are presented for the total elastic and rotational excitation cross sections and the momentum transfer cross section for incident electron energies of 0.5 to 10 eV. Comparison with other theoretical and experimental data is included. The second paper of this chapter presents cross sections for e--CO scattering in the static exchange approximation. The method of calculation is the T-matrix discrete-basis-set method as updated by Fliflet, Levin, Ma and McKoy (previous paper) along with the variational correction approach of Fliflet and McKoy. We extract the 2π resonance parameters at equilibrium internuclear separation and compare with other theoretical and semi-empirical results. Momentum transfer cross sections are compared with the experimental data of Land and the theoretical calculation of Chandra. Chapter II Results are presented in this first paper for rotational and vibrational-rotational excitation of H2 by electron impact in the static-exchange approximation. Using the T-matrix discrete-basis-set approach as described in the first paper of Chapter I we solve the fixed-nuclei scattering problem at several internuclear separations. Comparisons of our results with the experimental data of Linder and Schmidt and with other calculated results are given. In the second paper of this chapter we obtain vibrational and vibrational-rotational excitation cross sections of N2 by electron impact via the 2Πg resonance in the static-exchange approximation. To obtain highly accurate phase shifts in the Πg channel we use the variational correction of Fliflet and McKoy applied to the discrete-basis- set method of Fliflet, Levin, Ma and McKoy. As in e-H2 vibrational excitation, the approach involves solving the fixed-nuclei scattering problem at several internuclear separations. From these calculations one extracts the parameters necessary to calculate resonant vibrational cross sections in a compound state model. Our results are compared with the experimental data of Wong et al., and other theoretical calculations. Finally, the last paper discusses a simple model to include polarization effects in shape resonances. The position and width of the 2eV shape resonance in e-N2 scattering are calculated by solving the T-matrix equations with the static-exchange field of the N2 case of the N2- compound state. Resonance parameters obtained at the equilibrium separation of the molecule agree well with semi-empirical results. Most importantly, the procedure can be readily applied at several internuclear separations. Chapter III As was mentioned in Chapter I, an important refinement of the original T-matrix method is the analytic evaluation of Gaussian matrix elements of the free particle Green's function. Previous calculations evaluated these matrix elements by a numerical quadrature which was in practice restricted to cases of axial symmetry. In this chapter the derivation of a method for generating higher order Gaussian matrix elements is presented. Although this procedure is applicable to polyatomic systems, we list here only the types of matrix elements necessary for Σ, Π and Δ symmetries of a linear molecule. Chapter IV In this chapter a method for obtaining scattering wave functions at arbitrary energies is presented. Minimization of the variance integral for a trial wave function expanded in discrete basis functions only provides a criterion for choosing the expansion coefficients of the wave function. By using a separable representation of the scattering potential only one new class of matrix elements appears in the evaluation of the variance integral which is not already required in the diagonalization of the Hamiltonian. The method is applied to some model potentials and to s-wave scattering for helium in the static-exchange approximation.

  PublisherDOI

• DSMC Study on Unsteadiness of Supersonic Flow Over a Cylinder

  2026-01-08

  articleSenior author

  Unsteady flow over a 30 μm diameter cylinder at Mach 4 is investigated using Direct Simulation Monte Carlo across Reynolds numbers from 270 to 6750, spanning slip-flow to near-continuum regimes. Power Spectral Density analysis at discrete probe locations and Dynamic Mode Decomposition of flowfield snapshots independently characterize instabilities in both the stagnation and wake regions. In the stagnation region, a bow shock pulsation mode at 10-15~MHz frequency is identified across all cases despite a 24-fold variation in number density. In the wake, two distinct instabilities are observed. The canonical vortex shedding mode at Strouhal number St ≈ 0.2 dominates at lower Reynolds numbers. A second instability at St ≈ 0.5, consistent with previously reported slipline-driven oscillations in supersonic wakes, emerges at higher Reynolds numbers and progressively strengthens, becoming the dominant mode. In the studied Reynolds range, the wake transitions from no clear unsteadiness, to canonical vortex shedding at St ≈ 0.20, and reaching a regime where the St ≈ 0.5 mode becomes dominant, while also having a distinct independent bow shock pulsation possibly governed by its standoff distance.

  PublisherDOI

• Hybrid DSMC and PIC-MCC Method with Spatially Varying Pressure for Electric Propulsion Ground Testing

  2026-01-08

  articleSenior author

  Ground testing of electric propulsion (EP) thrusters is influenced by elevated chamber pressures that increase charge–exchange (CEX) and momentum–exchange (MEX) collisions, altering plume divergence and neutralization behavior. Fully kinetic Particle-in-Cell with Direct Simulation Monte Carlo (PIC–DSMC) simulations capture these effects with high fidelity but require substantial computational resources. More efficient Particle-in-Cell with Monte Carlo Collisions (PIC–MCC) methods require a prescribed neutral background, but assuming a spatially uniform value based on facility pressure gauges is problematic because such measurements do not resolve the localized pressure gradients near the thruster exit, limiting the accuracy of the model. This work evaluates a hybrid DSMC–PIC–MCC approach in which spatially varying neutral density and temperature fields from a neutral-only DSMC simulation are incorporated into a PIC–MCC solver. Three models are compared: (A) fully kinetic PIC–DSMC, (B) PIC–MCC with a uniform neutral background, and (C) PIC–MCC with DSMC-informed spatially varying neutrals. Under low-pressure conditions, the neutral-only DSMC field does not match the PIC–DSMC neutral distribution because it lacks wall-accommodated neutralized ions, but all three methods generate similar ion-beam properties due to minimal CEX activity. Under high-pressure conditions, DSMC accurately captures the spatial variation of neutral number density, as neutrals dominate over ions, but yields a nearly uniform temperature, unlike the elevated temperatures produced in PIC–DSMC by fast CEX neutrals. Despite this temperature discrepancy, Method~C reflects the major plume features governed by neutral-density gradients and provides better agreement with PIC–DSMC than the uniform-background model, especially in predicting CEX production and beam-wing broadening. Moreover, the hybrid method reduces computation time by roughly 90\% relative to PIC–DSMC and requires far less memory. These results indicate that DSMC-informed PIC–MCC offers a practical and computationally efficient alternative for modeling EP plumes in high-pressure ground-test environments where neutral-density variations strongly influence plume behavior.

  PublisherDOI

• Potential Fluctuations of Emissive Sheaths in Collisional Boundary Layer Plasmas Using a Kinetic Approach

  Springer aerospace technology · 2026-01-01

  book-chapterOpen accessSenior author

  Abstract Emissive plasma sheaths are subject to various instabilities, including two-stream (TS) and self-spike (SS) oscillations, both of which play critical roles in the sheath dynamics. This study extends our previous work by examining the effects of these instabilities on potential field fluctuations and the electron velocity distribution functions (EVDFs) in boundary layer plasmas. Using a fully kinetic PIC-DSMC approach, we demonstrate that the TS instability induces strong oscillations in the sheath potential, exceeding the sheath potential by up to tenfold in large domains. In contrast, the SS instability produces sawtooth oscillations with instantaneous, non-amplifying disturbances. Additionally, the TS instability significantly alters the EVDFs, creating highly non-Maxwellian distributions, while the SS instability causes instantaneous shifts in the EVDF peaks during SS ejections.

  PublisherOA PDFDOI

• Correction: Effect of a Backward-Facing Step on Drag of a VLEO Satellite

  2026-01-12

  article
  PublisherDOI

• Understanding in-chamber plasma behavior using a dimensionally scaled gridded ion thruster in three-dimensional kinetic particle-in-cell simulations

  Journal of Applied Physics · 2026-02-25

  articleOpen accessSenior author

  We investigate facility effects on a reduced-scale gridded-ion-thruster plume using a fully kinetic, three-dimensional particle-in-cell/Monte Carlo collision solver coupled with a direct simulation Monte Carlo neutral background. This approach enables detailed examination of key plasma processes governing beam neutralization and wall interactions under ground-test conditions. We find that inelastic electron cooling is essential for achieving a physically consistent, neutralized beam. Increasing the background pressure enhances ion–neutral collisions, leading to more charge- and momentum-exchange events that reduce ion mean energies, broaden the beam, and increase sidewall losses. Inelastic processes flattens the potential, sustains quasi-neutrality, and preserves beam collimation farther downstream. Single-particle trajectory analyses show that primary electrons undergo mixed escape and temporary trapping, while low-energy post-inelastic electrons remain confined, sustaining the neutralization cloud. Sheath diagnostics reveal that at the beam dump, classical Child–Langmuir and Hutchinson models underpredict the sheath length due to residual electrons, while near the sidewall, the sheath is truncated by beam-sheath interference within the compact domain. Current-flow analysis indicates that higher background pressure conditions yield lower beam energies and increased sidewall currents.

  PublisherDOI

• Modeling Laser Absorption Measurements in a Shock Tube using the Photon Monte Carlo Method

  2026-01-08

  articleSenior author

  The photon monte carlo (PMC) method has been used to model laser absorption measurements in a shock tube by solving the radiative transport equation for a one-dimensional computational domain. A new wavelength selection procedure has been implemented to select wavelengths for photon bundles at initialization which has been shown to be significantly faster than other various root-finding algorithms. The PMC method was applied to simulate tunable diode laser absorption spectroscopy (TDLAS) measurements targeting nitric oxide (NO) ro-vibrational transitions in the infrared (IR) region behind a normal air shock for different laser beam powers.

  PublisherDOI

• Effect of a Backward-Facing Step on Drag of a VLEO Satellite

  2026-01-08

  article

  Spacecraft in very low Earth orbit (VLEO) at altitudes of 150 – 300 km can be prevented from rapid reentry and maintained in orbit with an air-breathing propulsion system that overcomes aerodynamic drag without the need for on-board propellant storage. A conceptual generic satellite shape of an airbreathing VLEO satellite with an electric propulsion thruster and backward-facing step in free molecular flow is presented with a focus on (i) a fundamental examination of drag on a relatively simple vehicle shape, and (ii) fundamental computations for drag and drag reduction on a vehicle shape having a backward-facing step. The magnetoplasmadynamic electric propulsion option is discussed.

  PublisherDOI

• Effect of Scaled Angle on Unsteadiness Characteristics of Large Separation Bubbles in High Speed Flows

  Springer aerospace technology · 2026-01-01

  book-chapterOpen access

  Abstract In this work, high-speed flows at moderate Reynolds numbers over compression ramps with large scaled angles were computed using the Direct Simulation Monte Carlo (DSMC) method. The free-stream conditions and ramp angles were selected to ensure that the resulting scaled angles were large, leading to correspondingly large separation bubbles. The study revealed that once the separation bubble reached a certain size, further decreasing the wall temperature and increasing the free-stream speed and physical ramp angle destabilized the flow, and resulted in the formation of secondary recirculation regions. Moreover, configurations that had secondary recirculation regions were found to exhibit unsteady behavior. This unsteadiness in the near-reattachment region was particularly pronounced across all cases.

  PublisherOA PDFDOI

Frequent coauthors

• Sergey Gimelshein

  University of California, San Diego

  81 shared

• Vassilios Theofilis

  University of Liverpool

  47 shared

• Robert J. Collins

  Cardiff and Vale University Health Board

  46 shared

• Jiaqiang Zhong

  43 shared

• Е. В. Титов

  40 shared

• Graham V. Candler

  University of Minnesota

  39 shared

• Zheng Li

  39 shared

• Rakesh Kumar

  37 shared

Awards & honors

• Alumni Awards and Endowments
• Alumni Loyalty Award
• Outstanding Recent Alumni Award