Topology optimization of crash-stressed vehicle structures

The aim of the BMBF-funded research project is the methodological and software implementation of an industrial operational procedure for topology optimization of the cross sections of extruded sections for crash load cases.

It is a joint project in which the following partners are involved:

  • Automotive Simulation Center Stuttgart e. V
  • DYNAmore GmbH
  • University of applied sciences Hamburg
  • SFE GmbH
  • Prof. F. Duddeck (TU München) as cooperation partner of SFE GmbH

In addition, the associate partners are:

  • Adam Opel GmbH
  • Daimler AG
  • Dr. Ing. h.c. F. Porsche AG
  • Goethe‐Universität Frankfurt Goethe-Center for Scientific Computing, Simulation and Modeling (G-CSC)

They act as consultants and input-provider and play an important part in the implementation of the project results as the companies which use the simulation process. The duration of the project is 24 months.

Motivation and Objective:

New and enhanced virtual design methods are necessary to maintain the competitiveness in the automotive industry and for the efficient derivation of new structural concepts for lightweight, specifically for hybrid and electric vehicles,  due to the change in new requirements. Special numerical optimization methods are required and are already partially integrated in the vehicle development process (FEP).  However, currently there are no industrially applicable topology optimization methods for nonlinear problems, especially for the crash. Under topology comes the geometric description of the location and arrangement of structural elements, thus the understanding of  the conceptual layout. It is therefore not yet possible to generate new structural topologies in an efficient and automated way. That is why a multi-stage optimization method is to be developed and validated by the example of extruded sections. A pre-optimization of the topology based on two recently published approaches in the research literature is to be evaluated and implemented to meet the needs of the automotive industry. A graph based methodology is attached to it which is already used successfully in the aeronautics and astronautics industry. This allows for highly efficient evaluation of the variety of topology variants. This is then coupled to a final shape optimization, which is necessary to perform the fine tuning of the geometry to the stresses. The whole concept is modular and can be used in its entirety or only in partial combinations.

In order to meet the increasing complexity of today's vehicle development, both manufacturing constraints, joining techniques and material characteristics as well as the number of load cases due to the different functional demands on the body structure are taken into account. This is ensured in the project through industrial validation. The proposed new approach opens a wide range of future developments and applications not only because of its modularity and adaptability to different problems about the expert catalog. Due to the proposed developments a significant progress can be achieved. Since the structural design is dominated by crash requirements, the new proposed methodology for the automated determination of structure geometries (topologies) will present a significant improvement regarding efficiency and lightweight.