Evaluation of Proposed Standards for Aircraft Flyover Noise Analysis Systems

Abstract

A unique test is discussed in which identical tape recordings of aircraft flyover noises were analyzed by 12 different organizations in the United States and Europe to determine the degree of uniformity in data analysis that could be achieved. Although all organizations did not use the same analysis system, each system conformed to standard specifications proposed by a Society of Automotive Engineers (SAE) Instrumentation and Analysis Subcommittee. The purpose of the test was to evaluate the proposed standard. The statistical analyses of the test results are presented. Also described are the proposed standard, the elements of the evaluation test, and the analysis systems used. INIMIZING the differences among organizations in flyover noise levels measured for the same type of aircraft has been a continuing objective. These measurement differences result from both data acquisition and data processing variations. Improved uniformity of techniques and procedures will increase the usefulness of data from different organizations. This uniformity is not easy to achieve because of 1) the complexity of flyover noise, which contains non- stationary random data and fluctuating periodic signals, and 2) the variety of noise measurement and data processing systems employed and the effects of their various operating modes upon the results. As a collective effort to examine this problem, SAE Committee A-21, Aircraft Exterior Noise Measurements, initiated a program to determine the magnitude of measurement differences, identify potential causes, and establish recommended practices to minimize the differences. A major effort early in this program was a flyover noise test to compare test techniques and results. The tests were con- ducted in 1968 at Brown Field, Calif, using Boeing 707 and 727 aircraft. Eight different organizations in the aerospace industry recorded and processed data using their own stan- dard methods. Results then were compared to determine the magnitude of the differences and identify potential causes. These results indicated reasonable agreement in some areas, but significant differences in others. It was concluded that the overall level of disagreement was undesirable and that the differences were largely because of data processing technique differences.!