Never miss a story from Computational Fluid Dynamics. Subscribe for in-depth analysis and articles.
Get ready to hold on to the edge of your seats, fellow F1 fans! The upcoming Formula One season promises to be an exhilarating and heart-stopping spectacle. The Fédération Internationale de l'Automobile (FIA) committee has proposed an array of new design regulations that will push the limits of driver and vehicle safety. Teams will be given the opportunity to get creative within the restrictions, with upgrades to floor flexing, side-view mirrors, fuel cooling, roll hoop design, and more. With new testing restrictions in place to ensure compliance with FIA regulations, the competition is bound to be fierce. The Miami Grand Prix witnessed Max Verstappen take the first position, and now, McLaren is gearing up for the Emilia Romagna Grand Prix at the Autodromo Enzo e Dino Ferrari. While the race dates have been postponed due to heavy floods, we can still look forward to an unforgettable season of high-speed thrills and spills!
Until 2008, there were no restrictions on aerodynamic testing hours in F1. Some top teams, such as BMW Sauber, Honda, Williams, and Toyota, were investing millions of dollars into operating wind tunnels. These teams also used computational fluid dynamics (CFD) tools in their design validation cycle. At that time, BMW Sauber used the Albert 3 supercomputers with 4000 Intel cores. Racing authorities soon realized that deploying testing restrictions was necessary to avoid the unfair advantage for a few teams that could invest their finances into 24/7 wind tunnel testing and multi-core CFD simulations. In 2009, as a step towards curbing the testing hours, the Formula One Teams Association (FOTA) signed a resource restriction agreement, and today, the teams at Formula One have to follow these testing restrictions. Failure to comply with these regulations will invite a penalty.
The 3D representation (physical or digital) of the vehicle design is referred to as Restricted Aerodynamic Test Geometry (RATG), and teams are allotted 6 Aerodynamic Testing Periods (ATP) per year. The first testing period is at the end of week 9. The 2nd, 3rd, and 5th testing period lasts eight weeks each. Ten weeks are allotted for the 4th testing period, including the 14-day summer factory shutdown. The last or 6th testing period ends by the 31st of December. Teams must follow the wind tunnel and CFD testing restrictions to validate their vehicle’s aerodynamic performance.
The wind tunnel to be used for aerodynamic testing has to be nominated by the team to the FIA, and only one wind tunnel can be used in a 12-month period. According to FIA, in restricted wind tunnel testing (RWTT), a single run commences when airspeed rises above 5m/s and ends when it falls below 5m/s. Once air speed is above 5m/s the RATG must remain fixed until below 1m/s. Changes to RTAG are permitted between runs of RWTT, and only two occupancy shifts are allowed per calendar day.
Other sets of RWTT restrictions include that the RATG must be less than 60 percent of full-size vehicles, the wind tunnel speed should not exceed 50m/s, only two models can be tested in 24 hours, and creating curved flow conditions for testing is restricted.
Restricted CFD simulations (RCFD) include all three steps in a CFD workflow, i.e., preprocessing, solving, and post-processing. CFD simulations conducted at scales other than 1:1 or using non-gaseous are also RCFD solutions that provide insight into gaseous flows on the full-size F1 vehicle. Machine learning, artificial intelligence, and deep learning-based simulation results are allowed only during the post-processing stage of RCFD simulation. Moreover, the team’s compute resources must contain a set of homogeneous processing units nominated to the FIA. The 2022 budget for each team was 6 million Allocation Unit hours (AUh). An AUh is equivalent to a core hour on a physical CPU core. Non-RCFD can be done for optimizing CFD methodology but must be on RATG that was simulated 30+ months ago.
Cadence recent investment in Fidelity Cascade Technologies brings high-fidelity CFD solvers to the table, including LES turbulence modeling with GPU acceleration for high throughput, and can do wonders in pushing the limits in racing and F1. The table below presents a variety of system-level high-fidelity multi-physics simulations that can be completed in just a few hours using the GPU-resident Fidelity CharLES solver. These simulations can take days and weeks on the CPU, making them impractical for design purposes.
GPU-acceleration and GPU-resident solvers allow the exploration of complex design spaces with unprecedented speed and accuracy. In fact, it is fundamentally changing the game. The graph below compares the simulation throughput of an ensemble of Fidelity multi-physics LES performed on CPU and GPU. For a given investment in compute, based on AWS pricing of CPU and GPU nodes, the NVIDIA GPUs deliver nearly nine times the throughput of the CPU.
Currently, F1 regulates CFD use because it significantly improves the performance of the cars. For example, while addressing the stability challenge of cars in F1, they used 7500 CFD simulations and 16.5 million core hours of compute to introduce new aerodynamic regulations, thereby improving the downforce from 55% to 86%. F1 does not currently allow GPU acceleration of the solvers due to compliance with the AUh definition. These strict limits confine the types of CFD simulation performed. While F1 is not benefiting from GPU, this transformation already impacts other areas of racing, commercial automotive, and many other industries.
Are you looking for a simulation suite that can help you reduce racing car design time? Cadence Fidelity CFD solutions might be just what you need, with or without GPU acceleration. If you want to explore and get hands-on with the new Cadence Fidelity CFD tools, you can request a demo today!