CAD on the Cloud

HPC-Competence Center

Arctur prides itself in not being just another HPC infrastructure provider, but rather one which always provides personal support even to users which are not so knowledgeable of HPC, so ensuring that every HPC project achieves or even surpasses user expectations. Their motto is “More than just HPC.” As an SME, they understand the issues that other SMEs face when dealing with HPC.

logo11.cdr

The Enterprise

Stellba is a Germany-based SME working on hydropower plant maintenance, repair and overhaul, engineering and manufacturing one-of-a-kind products for the green energy sector with the goal to optimize energy efficiency.

 

How HPC makes the difference

stellba_blade

Stellba’s process to design a new blade on a HPC system is in fact running roughly 25 times faster than before

This success story is an outcome of CloudFlow project(*).

In the production process for any company, saving time during the design phase and optimizing the final product design are big challenges. Existing general-purpose CAD systems offer a way to create good designs for a multitude of products, but knowing the particular type of product in advance enables the development of dedicated functionality, thus minimizing design time and avoiding repetitive tasks and errors.

Due to the complexity of the underlying free-form shape, the 3D design of a hydraulic turbine blade at the end user Stellba is a long process. In the traditional CAD design approach it is necessary to use a lot of basic operations when modeling such a blade. These operations are mostly repetitive and similar for each design.

Stellba’s process to design a new blade on a HPC system is in fact running roughly 25 times faster than before, reducing the design time from eight hours to less than 20 minutes. Consequently, Stellba is now able to create more designs in a given period and to try out more possibilities to improve the quality of their blades. In addition they can provide their design models to other applications via the Cloud-based PLM, e.g. for a successive simulation step.

 

(*)The CloudFlow project has received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement No 609029.