Statistical analysis is used in the development of solid rocket random vibration environments. This type of analysis relies heavily on the standard deviation in predicting statistical extrema. Often, however, there is a lack of data from which to accurately determine the standard deviation, resulting in environments developed using assumptions about the standard deviation. Historically, a 3 dB standard deviation assumption has been used for launch vehicles. This assumption is based on a single broadband value drawn over normalized standard deviation curves that are frequency dependent. This work reassesses these assumptions for solid rocket motors using five programs of varied size, performance, and structural characteristics. Based on a Monte Carlo and curve fitting extrapolation technique, the limit of the mean programmatic standard deviation suggests the historical 3 dB assumption could be reduced nearly 50% to 1.56 dB.
展开▼