Industry Summary Articles

Mubea is one of the world leading manufacturers of lightweight components for the automotive industry. Since 2001 Mubea delivers Tailor Rolled Blanks (TRB®) to nearly all automotive OEMs. The technology of flexible rolling enables the production of sheet material with varying sheet thickness.

Automotive body parts are designed with respect to several disciplines such as function, cost, design or manufacturing. Often the resulting structures are a compromise of these disciplines. When manufactured from constant sheet material these parts are inhomogeneously loaded. With the application of TRB, it is possible to take advantage of this uneven material usage and reduce sheet thickness in areas with low loading.

In order to leverage the full potential of this technology, optimization strategies become the key enabler for efficient design. Optimization strategies are well known and widely applied in the automotive industry. There is a wide range of software solutions that offer process integration and design optimization (PIDO) capabilities. The increasing complexity in optimization tasks made it necessary for Mubea to review its PIDO strategy and explore alternatives in order to find the best solution for future optimization tasks. After evaluating 5 different PIDO tools, pSeven became the solution of choice for TRB optimization task in the upcoming years.

Tailor Rolled Blanks are manufactured through a special technological process chain and hence require a special CAE simulation chain. Mubea has developed a dedicated in-house software solution to handle the evaluation of a design point in order to enable automatic design exploration. This solution is based on the Python code.

“The integration of the in-house Python code into a pSeven workflow is quite easy and enables the re-use of existing logic. Another benefit of pSeven is the availability of the algorithms as a pure Python library pSeven Core. With this library, it is possible for us to make use of the provided algorithms from Python directly. This enables the easy development of our own optimization strategies.“

TRB parts are mostly designed with respect to highly non-linear and numerically noisy load cases. The use of fast and robust predictive models is the key to optimize parts with short turnaround time.

“In our benchmark, pSeven showed the best model quality when compared to the other software. With SmartSelection models can be trained with good prediction capabilities automatically. Another interesting feature in pSeven is the Dimension Reduction capability. Based on a Design of Experiments (DoE) of the complete problem dimensionality, the number of dimensions can be reduced. This helps us to overcome the curse of dimensionality and train predictive models for a high number of dimensions.”

To view the original press release, please click here.

Search for Mubea on CIMdata.com