Multidisciplinary Optimization in Bogie Development using HyperStudy and Middleware

Motivation and Objectives
Many optimization methods are well established and implemented in commercial tools. These tools are powerful and have many features. However, the implementation within the product development process is not always satisfying and therefore the potential of these tools is not fully used. This is especially true, if more than one discipline is taken into account (“multidisciplinary optimization”). Here, beside the knowledge of optimization methodology, the user also has to deal with complex data handling caused by different input decks, solver executables, result files, variable linking etc.  These difficulties could prevent “normal” users from using multidisciplinary optimization within the product development process.

Approach and Method
To overcome these problems it is sometimes helpful to introduce a middleware between the optimization tool and the solvers, which organizes the data handling and the communication between all parts. This middleware helps to make the setup process much easier by lowering the effort to configure various interfaces.

Contents of Presentation
In this presentation it will be shown how the process of multidisciplinary optimization can be simplified by the use of such a middleware tool. The architecture of this optimization system will be demonstrated and some results of the optimization process will be presented. 

 Values and Benefit
Complex data handling has always been a barrier for using multidisciplinary optimization. A new approach uses a middleware solution to simplify setup process and data flow in the interaction of optimization tools and solvers. This new method has now been successfully tested in the field of railway engineering and will be implemented in the product design process in the future at Siemens AG Austria.

The Author

Michael Alb (Virtual Vehicle)
Heinz-Peter Kotz (Siemens AG Österreich)