Eco-Runner Team Delft Uses Rescale’s Big Compute Platform to Design a Fuel-Efficient Vehicle
Guest post by Paul Hulsman, Team Manager, Eco-Runner Team Delft at TU Delft
Eco-Runner Team Delft is a student engineering team at the TU Delft. Since 2005 our mission has been to design and build the most efficient hydrogen-powered vehicle possible. Every year, we compete in the Shell Eco-marathon competition, which pits student-designed hydrogen-powered vehicles against each other to complete a 16-kilometer course using the least amount of fuel. There are two vehicle assemblies that are critical to winning: a fuel-efficient propulsion system and a body with minimal vehicle resistance or weight. The weight of the vehicle is crucial to fuel efficiency because the 2017 Eco-marathon course featured a hill. Thus, our design efforts this year focused on making every component of the car as light as possible to reduce the energy requirement to climb the hill. We used Rescale’s ScaleX platform to explore and simulate many weight-saving design options while maintaining structural integrity. Rescale’s cloud-enabled simulation allowed us to quickly design a lightweight, but strong vehicle on a constrained project schedule and a student budget.
For the past decade, we have used ANSYS Mechanical for our FEA to reduce the weight of the car’s carbon fiber body. Using previous models of the Ecorunner as reference points, we ran FEA on various parts of the car to determine the lay-up plan of the carbon fiber and to determine whether each component would withstand the loads applied.
FEA simulation is compute-intensive requiring multiple cores to run a fast simulation, and our team needed a final design within 3 months in order to compete on event day. Rescale’s cloud big compute platform provided turnkey access to those compute resources. We were able to get simulation results almost immediately after signing up for Rescale’s platform. We ran the majority of our simulations on 192 cores of the Gold hardware configuration (15 GB memory, 32 GB storage, 10 GB/s interconnect), which was the most cost-effective configuration for our simulations. The ability to scale out on unlimited cores on demand via the Rescale platform reduced our typical simulation turnaround time from a couple of days to a matter of hours.
Critically, this fast simulation turnaround kept us on schedule to compete on event day and enabled the team to quickly explore a variety of new concepts and design features before iterating to a final design. Rescale enabled a comprehensive design exploration phase, in which we discovered and implemented several design features that reduced the total weight of the carbon-fiber body by 5%, thereby increasing the energy-efficiency of our hydrogen-powered car.