Predicting Hypersonic Laminar-Turbulent Transition with Direct Numerical Simulation

Poggie, Jonathan (Purdue University)

Andrews, Geoffrey


Accurate prediction of laminar-turbulent transition is essential to the design of hypersonic thermal protection systems. Progress is limited, however, because of acoustic noise in hypersonic ground test facilities. There are no currently operating quiet wind tunnel facilities above Mach 6. Using direct numerical simulation (DNS), we aim to predict acoustic noise and transition in existing, conventional (noisy), hypersonic wind tunnels to make these facilities more useful for vehicle design. With the present investment of computer time, under a DoD HPCMP Frontier project, towards understanding the effects of tunnel noise, we may be able to save the DoD time and expense and give the United States a competitive advantage in hypersonic technology.

There are two main sub-projects: the first to characterize the freestream acoustic spectrum in a wind tunnel environment (noise prediction); the second to examine the effects of freestream noise on boundary layer transition (receptivity). The aim of the noise prediction sub-project is to simulate true wind tunnel nozzle geometries, and to predict the spectrum of freestream acoustic fluctuations, including amplitude, phase, and direction of wave propagation. The receptivity sub-project focuses on the response of the boundary layer to freestream noise. A particular goal is to explain an unexpected difference in receptivity between a cone and a hollow cylinder found by Stetson et al. (AIAA Paper 1991-1639) at Mach 8 in AEDC VKF Tunnel B.

Preliminary calculations with the full-scale geometry of the Stetson et al. cylinder and cone were carried out with forcing representative of the measured tunnel noise spectrum. These calculations demonstrated the feasibility of the approach, but they were not sufficiently resolved to capture transition to turbulence. Calculations with a smaller circumferential extent, but better resolution, have now been carried out. These have displayed transition to turbulent flow at location matching experiment, and they have captured some of the anomalous low-frequency content observed experimentally.