Data-Based Prediction Model for an Efficient Matching Process in the Body Shop

Achieving the optimal dimensional quality for automotive body parts today is a time and cost-intensive process still often based on trial-and-error approaches. There are two ways to improve the accuracy in the production process: Early modification of the tools in the press shop is one way to significantly manipulate the dimensional quality of parts, although resulting in high costs. The other—much more time and cost-effective—way is trying to change the geometry in the body shop, although providing a lesser adjustment range. Definition of a reasonable parameter adjustment in a single joining stage needs expert knowledge because the adjustment of a single fixture component can have a complex impact on the final assembly. In this publication, a new approach based on finite element simulation and statistical methods is presented being able to characterize the interactions between clamp settings and assembly geometry and to identify the main impact factors on the dimensional accuracy of assembled body parts. The surrogate model is based on smart data, gathered from FEM simulations.