Purdue University

Purdue's Formula SAE student racing team recently achieved the best result in their team’s 30-year history, thanks in large part to the use of the Rosen Center for Advanced Computing (RCAC)’s Scholar cluster.

Every year, the Formula SAE team designs and builds a single-seater race car from scratch, competing against teams from other schools around the country and the world. In 2022, the team took advantage of Scholar, a cluster RCAC makes available for educational use, to perform large fluid dynamics simulations and finite element analysis, which helped them achieve a 16% performance gain from the previous season.

Scholar’s powerful computational capabilities were critical in the design of the aerodynamics package, which is the primary design tool for the car. The team performed over 250 simulations in the span of five months, with around 7 to 8 hours spent on each simulation. This helped the team optimize the downforce and drag of the car, resulting in improved performance on the race track.

“We could not have achieved this result without access to Scholar,” says team member Chenkai Zhang, a junior in aeronautical and astronautical engineering. Zhang collaborated with a senior from the same major, Arpit Agarwal.

Lev Gorenstein, senior computational scientist for RCAC, worked with the team to identify needs and help them find solutions.

“Lev has been a huge help in the project. He really cares about what we do, and he shows it. He is always quick to respond to any questions we have, and always provides a ton of details,” says Zhang.

In addition to helping the team gain performance on the track, Scholar proved to be an important educational tool. Many team members gained valuable experience using high-performance computing and doing computational fluid dynamics, finite element analysis, aerodynamics design and programming.

With Scholar at their disposal, the team is looking forward to continuing to push the boundaries of open-wheel race car design and construction in the years to come.

Says Zhang, “In the near future, we will also be using Scholar to perform sensitivity studies on the aerodynamics package, and the data we gathered will help us in understanding the car better and extracting the performance of the track.”

Scholar is Purdue’s computer cluster for teaching HPC, offering seven login servers and 28 batch worker nodes. It provides a desktop-like environment and Linux remote desktop, Jupyter notebook server, or R Studio server for interactive use with HPC software and tool chains.

Access to Scholar is free for Purdue students and can be scheduled for classes through the Class Account Request page. All students registered for the class will automatically receive login privileges to the cluster.