Mixing processes in the transonic, accelerated wake of a central injector

Abstract

The compressible accelerated mixing layer of a central injector was thoroughly investigated experimentally to provide a data set that can be used for validating numerical simulations. A drop-shaped central injector was mounted upstream of a rectangular convergent-divergent nozzle, through which air was accelerated to a Mach number of 1.7. The free-stream Reynolds number at the point of injection was 6.245×10⁴. Four different measurement techniques – short-time illuminated schlieren imaging, laser schlieren, laser-induced thermal acoustics, and laser-induced fluorescence (LIF) – were applied to visualize the flow structures and to measure the predominant frequency of periodic flow features, the Mach number and temperature, and the injectant distribution. Instantaneous images show that the mixing layer was dominated by a series of alternating vortices. The mixing layer’s self-similarity could be proven by means of injectant mass fraction profiles, which were derived from LIF measurements. The growth rate of the mixing layer was shown to approximately follow the 1/2-power law. It was concluded from comparison to literature data that the growth rate is primarily determined by the free-stream Reynolds number, whereas the free-stream Mach number (compressibility effects) and the injectant amount play a minor role. These experimental data were used to validate three-dimensional (3D) unsteady Reynolds-averaged Navier-Stokes simulations using the shear-stress transport turbulence model. It was shown that the vortex shedding frequency and the mixing layer growth rate as well as the wake velocity deficit were underestimated by the simulations. This indicates that the flow physics of vortex formation were not entirely reproduced.