An experimental investigation of interaction of crossflow instability with forward facing steps

Abstract

Experimental measurements are performed on a 45 degree swept flat plate model at the low speed laboratory (LSL) at the Delft University of Technology, in a low turbulence environment to stimulate the development of stationary crossflow. The swept flat plate model is equipped with two linear manual stages to create forward and backward facing steps. Preliminary measurements characterize the pressure gradient over the swept flat plate model under study. In the preliminary study , hot-wire anemometry (HWA) measurements characterize the flow over the swept plate without steps over a large chordwise domain with and without forcing by discrete roughness elements (DREs). The DREs are spaced at a spanwise wavelength corresponding to the overall maximum N factors from LST. The mean velocity contours and N factor trends presented in these measurements reinforced the need for DREs to control flow. Spectral content is monitored and the frequency bands associated with probe vibration and travelling crossflow interaction were delineated. Infrared thermography was employed to observe the movement of transition front with varying step heights and initial crossflow amplitudes. When the DRE height increases, the transition front moves upstream consistently for all step heights. Furthermore, when the DRE height is kept a constant , but the array is moved upstream and downstream of the neutral point, the transition front moves upstream for all step heights. In order to observe the flow in the vicinity of the step, HWA was once again used to quantify the interaction of crossflow with FFS. The clean, short FFS and supercritical step height configurations identified from the IR study, are studied for two initial amplitudes. For the supercritical step configuration, bandpass filtered fluctuations are found to align with a high wall normal and spanwise shear region which has been identified in previous work. It is postulated to be associated with a vortex shedding mechanism, for which frequency bands are delineated. Estimates of the range of recirculation bubble length were made and a flapping frequency range was also demarcated. In this study, a vortex shedding scenario is used to explain the presence of these near wall fluctuations. To conclude the report, recommendations are made for extending the present study for future work.