Systems Design Using Solution-Compensation Spaces with Built-In Tolerance Applied to Powertrain Integration

Abstract

Complexity in systems design can be reduced by computing permissible ranges for some crucial design variables that need to be defined in an early design phase. These ranges are calculated such that there is sufficient tolerance for the remaining design variables in later design phases, while still achieving the overall system design goals. A new algorithm for this approach is presented and applied to the design of a vehicle powertrain mount system. The results show large permissible ranges for mount positions while maintaining sufficient tolerance for mount stiffnesses.

Optimizing Requirements for Maximum Design Freedom Considering Physical Feasibility

“…As for the extensions of the classical method, many different approaches have been developed with the goal of increasing the size of the resulting solution space. Solution-compensation spaces, for example, separate the problem into early-and late-decision variables in order to give more design-freedom to the early-decision variables, while still ensuring there's a degree of built-in tolerance for the late-decision variables (Vogt et al, 2018;Stumpf et al, 2022). A different approach involves introducing pair-wise coupling of design variables, either through enabling rotation of the box or by introducing 2d-polygons for paired variables, which decreases the restrictiveness of the problem and thus increases the overall size of the solution space (Erschen et al, 2018;Harbrecht et al, 2019Harbrecht et al, , 2021.…”

“…The solution spaces approach was first developed and used in the context of vehicle crash design (Lehar and Zimmermann, 2012;Zimmermann and von Hoessle, 2013;Fender, 2013;Zimmermann et al, 2017) in order to deal with the high degree of uncertainty present in the development process. Since then, it has also seen use in the development of product families for electric vehicles (Eichstetter et al, 2015;Rötzer et al, 2020Rötzer et al, , 2022 and on the integration design of a powertrain (Stumpf et al, 2022).…”