Aerodynamics are key attributes in new cars. Airflow over a vehicle is critical to gas mileage, can produce annoying wind “buffeting “and affects the vehicle’s quality and success in the market in many other ways. As with all other design features, auto companies want to use HPC simulations to predict aerodynamic performance. However, most automobile designs are not particularly aerodynamic, so airflow is very complicated and difficult to simulate accurately.
Cray and ANSYS recently published an application brief on an aeroacoustics (wind noise) simulation on an Alfa Romeo Giulietta automobile. It’s a very interesting example for many reasons, including:
- It demonstrated the technical partnership between Cray and ANSYS.
- It was a commercial production simulation, scaling to over 8,000 cores on the Cray® XC40™ supercomputer.
- It provided a good description of the source of vehicle noise and the challenge of reducing it.
In addition, the brief illustrates yet another example of the increasing range of simulation, and hence value, that HPC can bring to the automotive design process.
In previous posts, I have highlighted the impact (no pun intended) of crash/safety simulation and structural vibration (NVH) in the design process. The use of high-fidelity CFD simulation has also become a critical tool in the automotive industry. Vehicle noise level is a key buying criterion, and it even impacts passenger safety. High-end CFD simulation not only aids acoustic design, but also helps address vehicle drag (which affects fuel efficiency), interior ventilation (passenger comfort), engine cooling (reliability) and several other key qualities. Hence, the use of CFD, especially highly accurate and computationally demanding CFD, is emerging as a core technology in the automotive design process.
As the brief notes, in automotive aeroacoustics simulation it is especially important to accurately model the side-view mirrors and other elements of the vehicle in order to capture the “fine details of the flow, which can be achieved only through accurate prediction of complex flow features.” This simulation leveraged both features in the Fluent software to enable extreme fidelity and also the extreme scalability of the Cray XC40 system to deliver acceptable turnaround time. The result is a solution that can be applied to a wide range of automotive design problems.
HPC compute power in the automotive industry has kept growing during the past 30 years. Large auto companies already have petascale HPC environments (approximately 30,000 compute cores) to support this workload. This growth was driven primarily by crash and safety simulation. With the ability to perform high-fidelity aeroacoustic simulation, and the associated value from the simulation, we can expect extreme fidelity and extreme scaling CAE simulation to become even more critical to a competitive automotive design environment.
Read the application brief.