Supercomputers Simulating As Close As Possible To Reality

Trucks drive thousands of kilometers through Europe every month, taking oranges from Greece to Scandinavia, delivering Spanish vegetables to German wholesalers, and collecting milk from farms in the region to take it to central dairies. To make sure the tires, wheel rims and other parts will survive the many kilometers without breaking down, the manufacturers test prototypes in test rigs to discover their service life.

Such a test often lasts several weeks, yet it can be rendered useless by malfunctions, such as when bearings or sensors wear out. In the Computer Aided Robust Design (CAROD) project, research scientists from seven Fraunhofer Institutes are devising methods with which malfunctions of this nature can be simulated ahead of time. The researchers are using the results to develop sturdy test rigs for life-cycle tests.

“Today the development and testing of prototypes – be they entire cars or individual components – takes place mainly in the computer,” says Andreas Burblies, spokesman for the Fraunhofer Numerical Simulation Alliance. But this simulation only reflects reality to a limited extent. “As a rule, there are no parts or manufacturing processes in which all product or process properties are identical.

But the developers always get the same simulation results if they enter the same pa-rameters.” This is where the researchers come in with their Computer Aided Robust Design. The goal is to develop new methods and software that makes it possible to factor the real deviations into the simulation calculations. In this way mechatronic systems, crash tests or laser processing methods can be made even less vulnerable to errors and variations.

One of the pillars of the new technology is the Taguchi method. The Japanese scientist Genichi Taguchi developed a method of making products, processes and systems resistant to interference. It is already applied in quality management, enabling the industry to achieve the optimum product quality. The task of CAROD is to improve quality by taking faults, variations and breakdowns into account during the virtual design phase.

“We are aiming to get as close to the natural manufacturing conditions as possible with our simulations,” says Dr. Tanja Clees, project manager at the Fraunhofer Institute for Algorithms and Scientific Computing SCAI in Sankt Augustin. Right now it is still early days for the new simulation software, but the experts are confident of achieving good results very soon. CAROD can be seen at the Hannover Messe in Germany from April 21 through 25.