OMS - Optimierung mechanischer Strukturen

Graph and Heuristic based Topology Optimization of crash structures under axial compression loads

Due to their superior energy absorption capabilities, extruded profile structures are broadly used as crash structures in the front and rear ends of vehicles. For their efficient crash design, topology optimization methods can play a key role. Due to the occurring nonlinearities in vehicle crash simulations, established topology optimization methods are usually inapplicable. Therefore, new methods such as the Graph and Heuristic based Topology Optimization (GHT) were developed. This method uses mathematical graphs for a simplified description of the profile cross-section. Furthermore, competing heuristics based on expert knowledge are used to modify the topology. These heuristics were developed for the optimization of profile structures exposed to lateral compression and/or bending loads (e.g., vehicle rocker under pole impact). However, for profile structures under axial compression loads (e.g., crash box in front crash), progressive buckling is usually the desired mode of deformation. Compared to laterally loaded profile structures, different considerations must be made to provide high specific energy absorption capabilities while initiating and maintaining a robust crash behavior. To address these challenges, a new set of heuristics is developed in this work. Since the structural responses of profile structures under axial loads can also be sensitive to small variations in the finite element model, the automated model creation process is enhanced to consider geometrical details. Furthermore, trigger mechanisms are established to control the structural behavior. Hence, the method allows for a parameter-based application of trigger mechanisms. The effectiveness of the topology optimization method is demonstrated in various application examples. The method proves to be suitable for an industrial development environment due to the implemented manufacturability checks, the consideration of geometrical details as well as its effectiveness regarding the required amount of function calls.

Complete dissertation as PDF

Weitere Infos über #UniWuppertal: