As a part of the VP Research & Innovation portfolio, UBC Advanced Research Computing (ARC) supports the computing and data management needs of researchers across campus. At the core of their offerings is UBC ARC “Sockeye”, a powerful, cutting-edge computing platform with nearly 16,000 CPU cores, 200 GPUs, and 3 petabytes of storage capacity. Recently, UBC ARC provided free access to Sockeye’s extensive computational power to support an innovative, high-load project in the development of UBC Formula Racing’s race car. This project, led by a UBC student engineering team, was entered into the annual Formula Society of Engineers (FSAE) Michigan competition this past May, an international competition where they placed in the top 10 for design with the help of the Sockeye platform.
Fast computing for fast cars
Led by UBC student Tom Rotbart, the Formula UBC Racing’s aerodynamics team opted to leverage ARC’s resources to assist in running simulations, a key step in the process of formula racing design. To effectively convert the air flow around the vehicle into downforce and cooling, advanced computations are used to support informed engineering decisions prior to manufacturing. In an effort to understand the aerodynamic performance of the race car design, the aerodynamics team used ARC to extract the coefficient of lift, drag and aero balance percentages; look through aerodynamic geometry pressure distributions; analyze line integral convolution plots to understand surface flow behaviours; and observe pressure and velocity fields. When asked about his experience with ARC, Tom recounted:
“Having access and support from ARC allowed us to save thousands of dollars and hundreds of hours from trying to acquire the computational resources required to run our simulations elsewhere or running these simulations on our local underpowered devices.”
Tom and his team are now using Sockeye to break new ground in understanding dynamic performance in pitch, roll, and yaw to further refine their designs. The success of Formula UBC’s project car serves as an example of the wide-ranging applications of ARC and how it can be leveraged for use-cases in engineering and manufacturing research.
About UBC Advanced Research Computing Sockeye
The ARC Sockeye platform is offered free-of-charge to the UBC community. Designed to meet the needs of UBC researchers across all disciplines, Sockeye is suited for projects with needs that cannot be handled by standard computing infrastructure alone. Projects that require research involving big data, large computational loads, or advanced modelling or visualization are perfect candidates for utilizing Sockeye.
To learn more and to apply for their services, visit the ARC website.
Article written by Aaren Lin