Advanced Iterative Solvers and Optimal Design of Industrial Machine Components.pdfVIP

JOHANNES KEPLER UNIVERSITAT LINZ Advanced Iterative Solvers and Optimal Design of Industrial Machine Comp onents Gundolf Haase and Ewald Lindner Department of Analysis and Computational Mathematics Johannes Kepler University A Linz Austria Institutsb ericht Nr July INSTITUT FUR ANALYSIS UND NUMERIK A LINZ ALTENBERGERSTRASSE A USTRIA Advanced Iterative Solvers and Optimal De sign of Industrial Machine Comp onents Gundolf Haase and Ewald H Lindner Institut f ur Analysis und Numerik Abteilung f ur Numerische Mathematik und Optimierung Johannes Kepler Universitat Linz Linz Austria Abstract We consider the optimal design of a machine frame under several stress constraints The shap e optimization is based on a QuasiNewton Metho d We need to solve the plain stress state equations in a complex domain for each evaluation of the ob jective therein The complexity and robustness of the opti mization dep ends strongly on the solver for the p de Therefore solving the direct problem requires an iterative and adaptive multilevel solver Although we started with a p erfected typ e frame we achieved another reduction in mass Intro duction Some kilometers away from Linz the industrial company ENGEL Maschinenbau GesmbH in the eld of mechanical engineering is situated One of their elds in constructing fabricating and selling industrial machine comp onents is in injection moulding machines The classical construction is based on bars but ENGEL Maschinenbau hold