Theory of Uncertainty Analysis With Application to Naval Hydrodynamics

Abstract

The modern methodology for quantifying the quality of experimental data is uncertainty analysis. Current methods are reviewed with some examples primarily from naval hydrodynamics. The methods described apply to fluids engineering. The history of uncertainty analysis, U.S. and international standards on uncertainty analysis, verification and validation standards for computational fluid dynamics, and instrument calibration are discussed. One important result is that random loading in force calibration can produce a lower uncertainty estimate than sequential loading. Statistically, the calibration results for the slope and intercept are the same for the two methods in the example thrust calibration, but the uncertainty in random loading is a factor of three smaller than sequential loading.