Simulationsunterstützte funktionale Grenzlagenabsicherung

Description

In addition to a classical toleration of components and assemblies Computer-Aided Tolerancing (CAT) became an indispensable engineering discipline in recent years~-- especially in the context of a series manufacturing of a product. Taking 3D geometries, resulting lever ratios, assembly sequences and various reference systems as well as statistical approaches into account, it already enables complex tolerance calculations of products that are subject to variations arising from the manufacturing process. One of the main premises for these kinds of simulations refers to the presumed rigid behavior of all contributors.

As a direct consequence of that assumption this thesis describes a virtual approach, based on specified Key Product Characteristics (KPCs) that takes both elasticities and a possible non-linear system behavior in the simulation into consideration. This proceeding to include materialities and interactions between tolerance affected components increases the forecast quality of the tolerance simulation in terms of a more realistic prediction.

Based on the results of a 3D tolerance analysis with a subsequent Design of Experiments (DoE to increase the simulation effort) KPC-related predicted limiting positions will be made available to finite element simulations. By including materialities and contact conditions to the chosen system, which is subject to non-linear effects, corresponding elastic limiting positions occur regarding the specified KPCs. An appropriate strategy for coupling of the rigid limiting positions with their corresponding elastic counterparts will be derived from existing research approaches. This strategy will be applied in order to enable a better prognosis for overall limiting positions. Additionally, robustness evaluations that are based on the aforementioned finite element simulations will be presented regarding the probabilistic impact of the single contributors on the characteristics of the limiting positions. Finally, correlation studies are carried out to verify the plausibility of the results of that virtual approach. Therefore, with regard to the selected demonstrator, a comparison will be made with data from an experimental test rig.