Application of statistical parameters in defining inlet airflow dynamics.

Abstract

During wind-tunnel and flight testing of several inlet duct designs, duct-pressure time histories were recorded on magnetic tape. Preliminary examination of the data indicated the pressure variations to be random processes having Gaussian distribution patterns. Reduction of these data by electronic analog instruments was made to obtain specific statistical parameters; namely, variance, time auto correlations and cross correlations, and power spectral densities. This paper discusses the utilization of these statistical parameters in helping to define the dynamic properties of the inlet duct and the state of the airflow turbulence. Examples are given illustrating the methods and techniques involved. HE inclusion of random phenomena as important governing factors in the design of modern aircraft systems introduces an element of uncertainty into the predictions of system responses. Consider the problem of flow instability in inlet ducts and its relation to engine stalls occurring in modern high-speed aircraft. As the inlet system operates more supercritically, boundary-layer flow separation increases, resulting in increased turbulence and spatial distortion at the engine compressor face. The onset of such conditions brings about stall in an apparently random manner; i.e., stall sometimes occurs immediately, sometimes several minutes of operation are possible without stall. If the occurrence of engine stall was a true random phenomenon, then it could not be controlled and, consequently, no further studies, statistical or otherwise, would prove useful. The engineering assumption, of course, is that stall is not a random phenomenon, but a function of nonrandom and controllable effects that can be determined from the flow properties of the inlet-engine system. The basic usefulness of statistical analysis for this study, therefore, lies precisely in its ability to separate the nonrandom effects from what seems to be the more or less random and unpredictable data. From pressure data taken during both wind-tunnel and flight test of several different inlets, statistical parameters were determined through data-reduction techniques offered by the electronic analog data-reduction instruments. The data were recorded during all stages of inlet operation, inAVERAGE PRESSURE AREAS (A+C+E-B-D) tf-tj