Backpressure unstart detection for a scramjet inlet based on information fusion

Abstract The process of a two-dimensional hypersonic inlet from start to unstart is experimentally demonstrated in this article by recording the corresponding time history of wall static pressures. The careful analysis of pressure signals indicates that T8, located the upstream last ramp compression shock, is the optimum transducer for unstart detection because it can denotes obviously the start mode and unstart mode. The derivative-based detection developed in the current paper is a new and easy-implementation technique. Comparing it with other reported techniques including the STD-based detection and the spectrum-power-based detection, the derivative-based detection has no data to be stored and no expensive calculation benefiting the detection in real-time, and it can be achieved by an analog circuit in engineering. A particular of interest focuses on the detection of the non-oscillatory violent pattern. The information of pressure magnitude of T8 is introduced to eliminate the false detection in the non-oscillatory violent pattern. Apply the given detection method to six runs under different freestream conditions, and the results indicate that it can detect unstart in time and shows a good performance.