Automated analysis of the form error caused by springback in metal sheet forming

Biermann, D.^{1, a}; Surmann, T.^{1, b}; Sacharow, A.^{1, c}; Skutella, M.^{2, d}; Theile, M.^{2, e}

1)

Institut für Spanende Fertigung, Technische Universität Dortmund, Baroper Str. 301, 44227 Dortmund

2)

Combinatorial Optimization & Graph Algorithms Group, TU Berlin, Strasse des 17. Juni 136, 10623, Berlin, Germany

a) biermann@isf.de; b) surmann@isf.de; c) sacharow@isf.de; d) skutella@math.tu-berlin.de; e) theile@math.tu-berlin.de

Kurzfassung

In mechanical engineering the problem of form errors does often occur. Especially in metal sheet forming, the deformations caused by springback are very strong. The common way to handle this problem is by trial-and-error or by numeric simulation. However, the simulation of springback is only successful for specific cases, but in general case it is inexact. Thus, in order to keep the manufactur-ing process as efficient as possible, it is very important to detect and analyze the form errors of manu-factured workpieces as early as possible in the production chain. The aim of the analysis is the identi-fication of form errors like material allowance, bending, or torsion and to create descriptions that can help to generate appropriate rules for error compensation by tool modification or changing relevant parameters according to the error descriptions. In case of strong deformations, it is difficult to locate the characteristic parts of the CAD model in the digitized data. Thus, the calculation of an appropriate registration (bestfit) is difficult. This article pre-sents a new approach by combining traditional methods known in production (such as the ICP algor-ithm) with methods from discrete mathematics in order to compare the CAD model with the digitized data of the deformed workpiece. Among those tasks is the reduction of the input data without losing necessary information, an improved assignment of the data from both datasets as well as methods of global registration (bestfit) and a new approach of local registration. In this process, methods from combinatorial optimization like a clustering approach and a matching algorithm will be used.

Schlüsselwörter

form error, springback, bestfit

Veröffentlichung

In: Proceedings of the 3rd International Conference on Manufacturing Engineering, 1.10.-3.10. 2008, Kallithea of Chalkidiki, Greece, S. 737-746