Statistical Tolerance Analysis With Sensitivities Established From Tolerance-Maps and Deviation Spaces

Abstract

Math models aid designers in assessing relationships between tolerances that contribute to variations of a dependent dimension that must be controlled to achieve some design function at a target (functional) feature. The Tolerance-Maps© (T-Maps©) model for representing limits to allowable manufacturing variations is applied to identify the sensitivity of a dependent dimension to each contributing tolerance of the relationship. For each contributing feature and tolerances specified on it, the appropriate T-Map is chosen from a library of T-Maps, each represented in its own respective local reference frame. Each chosen T-Map is then transformed to the coordinate frame at the target feature, and the accumulation T-Map of these is formed with the Minkowski sum. The shape of a functional T-Map/deviation space is circumscribed (fitted) to this accumulation map. Since fitting is accomplished numerically by intersecting geometric shapes, T-Maps/deviation spaces are constructed with linear half-spaces. The sensitivity for each tolerance-and-feature combination is determined by perturbing the tolerance, refitting the functional shape to the modified accumulation map, and forming a ratio of the increment of functional tolerance to the perturbation. Taking tolerance-feature combinations one by one, sensitivities for an entire stack can be built. For certain loop equations, the same sensitivities result by fitting the functional shape to the T-Map/deviation space for each feature, without a Minkowski sum, and forming the overall result as a scalar sum. Sensitivities are used to optimize tolerance assignments by identifying the tolerances that most strongly influence the dependent dimension at the target feature. Form variations are not included in the analysis.