Industrial Applications of Uncertainty Analysis

摘要

Uncertainty or tolerance analysis is a powerful technique for addressing many problems facing the practicing engineer. This paper reviews several of the common uses of this tool and introduces some novel applications. The paper begins with derivation of "Law of Error Propagation" and follows the developments presented earlier. In this development, explicit forms for the outcome of a process and its statistics are developed. These results are then applied to a number of common industrial applications. The applications include, error budgeting and requirements derivation. Included in the discussion of the derivations of requirements is a discussion of the need to include margin and contingency explicitly into the analysis. Next, an expression for the repeatability of a process is derived. This analysis is then used to identify the general conditions that lead to a repeatable process. Combining cost data or modeling, error budgets are extended into tools that can be used in Cost as an Independent Variable (CAIV) modeling and analysis. The combined model can then be used to determine cost as function of tolerance and the marginal cost of repeatability.