Development of a Multi-Camera PIV Imaging System for Studies of Shock/Boundary Layer Interactions

This paper describes the development of a multilaser, multi-camera PIV system for the investigation of shock wave/boundary layer interactions. The PIV system uses up to four 1k×1k pixel CCD cameras and two dualcavity Nd:YAG lasers to capture large field-of-view images of the entire interaction, or four-image time sequences. The large field-of-view images – encompassing the upstream boundary layer, separation shock, separated flow and the nominal reattachment region – are ideal for investigating the global structure of the interaction. For the time sequences, the time between PIV frames can range from 30 to 150 μs. In this paper, two-image time-sequences are demonstrated, where the delay time was 40 μs. To validate the PIV technique in this flow, a series of experiments was conducted to determine the response time of the seed particles through an oblique shock. Titanium dioxide particles were shown to have a particle response time constant of about 2.6 μs, which is sufficiently small that most of the flow features in the interaction will be faithfully reproduced. Our preliminary results show that this system can produce high quality PIV data that can reveal new and interesting details of the flow structure. In the future, this system will provide a powerful tool for investigating the physics of shock-induced turbulent separation. 1.0 INTRODUCTION Shock-induced turbulent boundary layer separation is a critical issue in the design of nearly all supersonic/hypersonic aircraft, missiles and projectiles. This phenomenon is problematic because it is associated with high fluctuating pressure and heat flux loads, which can lead to structural fatigue and thermal problems. Of particular concern is the unsteadiness of shock wave turbulent boundary layer interactions (SWTBLI), which can be characterized as exhibiting a broad range of frequencies and scales of motion. Of particular interest is the low frequency, large-scale motions of the shock foot, since it is this type of unsteadiness that has proven to be the most difficult to describe with computational and theoretical models. Erengil and Dolling studied the correlation between the wall pressure fluctuations beneath the incoming boundary layer and the shock foot velocity. They concluded that the small-scale motion of the shock is caused by its response to the convection of turbulent fluctuations through the interaction. The large-scale motion is a result of the shock's displacement due to the expansion and contraction of the separation bubble; however, no mechanism was identified for the cause of the separation bubble's low frequency, large-scale pulsating motion. Beresh et al. used planar flow visualization and fast response pressure measurements to monitor the shock foot location in a compression ramp interaction. They visualized the flow by using planar laser scattering (PLS) from a seeded alcohol fog. In that study the seeding density was sufficiently high that the shock foot could be seen in most of the images. Double-pulsed image pairs, separated in time by 15 to 30 μs, showed that large-scale structures in the upstream boundary layer would greatly distort the outer region of the separation shock, but the shock foot did not move appreciably in this time scale. This result was consistent with previous studies using wall pressure measurements, which reported shock frequencies that did not exceed 10 kHz. They concluded from this that largescale structures in the outer part of the boundary layer were not primarily responsible for the motion of the separation shock. This observation differs from the observations of Wu * Graduate Student † Associate Professor, Senior Member AIAA ‡ Professor, Fellow AIAA Copyright © 2002 by Y.X. Hou, N.T. Clemens and D.S. Dolling. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission.