From pressure fluctuations to dynamic loads on axisymmetric step flows with minimal number of kulites

The present study focuses on two kinds of reconstruction methods designed to capture the dynamic loads acting on the reattaching part of an axisymmetric step flow. The pressure signals on discrete points of the geometry constitute the starting point of this study. Data are taken from a validated numerical simulation in order to maximize the number of sensors at the wall. First, an a priori methodology for the load reconstruction without any hypothesis on the flow behaviour is exposed. The influence of the number of sensors and their spatial distribution over the geometry is investigated. It is shown that azimuthal and longitudinal resolution have different effects on the reconstructed power spectral density of the load. Then, it is found that an experimental process with few unsteady sensors cannot provide an accurate prediction of the load. Then a second kind of reconstruction method based on the physical knowledge of the flow is presented. The original Nguyen approach initially developed for internal side loads is optimized with a spatial analysis and gives a satisfactory prediction of the load with only a minimal number of sensors.