Jerome E. Manning
Cambridge Collaborative, Inc., 689 Concord Ave., Cambridge, MA 02138
Statistical energy analysis (SEA) is generally used to predict acoustic environments at high frequencies for random excitations. SEA can also be used to predict the environment at low frequencies for pure-tone excitations. The use of SEA for such problems requires consideration of both the mean and variance of the statistical prediction. Assuming a normal distribution of acoustic levels, 95% of measured data would be expected to fall within the limits set by the mean plus and minus two times the standard deviation. The use of SEA to predict both the mean and standard deviation of the cabin noise for a propeller-driven aircraft will be shown and comparisons with measured data will be presented. Methods to reduce the predicted variance will be discussed.