This preview shows page 1. Sign up to view the full content.
Unformatted text preview: lerances are then identified with an STA sensitivity analysis. The design iterations are efficient, as they do not require repeated simulations or remeshing. 3. GEOMETRIC COVARIANCE As explained above, when part data is not available, i.e. the parts are not in production, the geometric covariance is modeled because STA does not provide any covariant information. [Bihlmaier, 1999] shows that to accurately model the geometric covariance, the surface variation of the mating parts must be accounted for. It is useful to divide the surface variation into three frequency domains; 1. Warping – Wavelengths longer than the part-length 2. Waviness – From one to five wavelengths over the part-length 3. Roughness – More than five wavelengths over the part-length Predicting Deformation of Compliant Assemblies 325 Figure 2: Examples of types of surface variation
where Figure 2 shows examples of each type of variation. The surface variation on any surface is composed of a combination of variation from the th...
View Full Document
- Spring '10
- The Land