Seeding Bias in Particle Image Velocimetry Applied to Dual-Mode Scramjet

Abstract

Uniform seeding is important for particle image velocimetry due to the fact that all velocity data are derived from the movement of particles as fluid tracers. However, in some wind tunnel applications, uniform seeding is not possible due to the severe fouling of windows. One such application is in the University of Virginia Supersonic Combustion Facility. Past particle image velocimetry measurements in this facility have potential velocity bias errors due to nonuniform seeding, since particles were only introduced in the fuel stream and not the freestream of a dual-mode scramjet combustor. The present study seeks to experimentally quantify the velocity bias errors associated with introducing seed particles into a single stream of a fuel and freestream mixing and combusting region of the scramjet flowpath. The velocity bias is quantified by measuring the velocity field at the combustor exit using stereoscopic particle image velocimetry under three seeding scenarios: 1) fuel stream only seeded, 2) fuel and freestream seeded, and 3) freestream only seeded. The case of seeding both the fuel and freestream was taken as the baseline, or most ideal solution. The results indicate that the effect of the seeding bias is relatively small since seeding the fuel only results in an average seeding bias error of 3.7% in mean velocity and 2.5% for the case of freestream only seeded. For the rms velocity, the average error induced by the seeding bias was 6.6 and 4.1%, respectively.