OMS - Optimierung mechanischer Strukturen

Development of an optimization scheme for shape and topology optimization of deep drawable sheet metal constructions

Sheet metal parts are commonly applied lightweight structures incurring low manufacturing costs in mass production. A good design of sheet metal structures covers the mechanical requirements at prescribed load cases and the requirements of the manufacturing process. Structural optimization is a tool for iterative improvement of components. Structural calculations are performed and the structure is adopted based on the component’s performance. This loop is processed until no more significant improvement is expected. In this thesis, tools for topology optimization are utilized. Topology optimization is a type of structural optimization allowing for holes to occur or vanish. Thereby the topology and the shape of components are optimized, so that any complex structure can emerge. To obtain sheet metal parts, new manufacturing constraints are implemented in the topology optimization. They allow for the optimization of sheet metal parts, which are manufactured by single-step deep drawing at room temperature. The density method is chosen as approach for the topology optimization. It improves components according to their sensitivities. First this method is extended, so that numerous structural responses (mass, stiffness, strength, eigenfrequency, buckling etc.) can be addressed during the optimization. Afterwards a manufacturing constraint for sheet metal structures without undercuts is introduced. Thereby the sensitivities of the objective function are manipulated for elements that are far away from the current mid surface. Furthermore, manufacturing constraints for minimal draw radii and for the prevention of tearing during the deep drawing are implemented. In addition to the structural calculation with the prescribed load cases, a deep drawing simulation is performed. Its results are used to smooth the mid surface in critical areas. These manufacturing constraints are realized as heuristics.

Shaker Verlag, ISBN: 978-3-8440-6196-3

Complete dissertation as PDF

Weitere Infos über #UniWuppertal: