DSMC Study of Hypersonic Boundary-Layer Unsteadiness Using Data-Driven Analysis

Abstract

The prediction of transition remains challenging because the receptivity amplitude is usually prescribed rather than measured. This work uses Direct Simulation Monte Carlo (DSMC) to study an unforced Mach 6 nitrogen boundary layer over a 120 mm flat plate, with no imposed inlet wave, wall signal, modal shape, or disturbance spectrum. The retained unsteady field is the intrinsic finite-particle stochastic fluctuation field carried by the DSMC representation. Changing the DSMC particle count provides an independent numerical control parameter that sets a predictable finite-particle noise floor while preserving the time-averaged boundary-layer flow. Dynamic Mode Decomposition (DMD) is then applied to unforced DSMC snapshots. The DMD eigenvalues and spatial modes estimate the quantities that appear in the LST normal-mode description, including frequency, temporal growth or decay, wavenumber, wavelength, phase speed, mode location, and amplification envelope. LST is used after extraction to test whether the data-derived structures occupy the expected second-mode band. The results show that DSMC data can connect a measured finite-particle stochastic disturbance level to second-mode instability dynamics without imposed disturbance spectra.

arXiv comment11 pages, 4 figures. Presented at the AIAA AVIATION 2026 Forum, 8-12 June 2026, San Diego, CA. Published by AIAA, DOI: 10.2514/6.2026-4737. Copyright 2026 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved. Author version; reference list typos corrected