NASCAR racing teams are using a cutting-edge automated simulation workflow pioneered by D2H and Ansys to improve engineering design. The work reduces hands-on development time while integrating high-performance computing (HPC) to enhance and speed designs.
With just a week to prepare between races, NASCAR teams have traditionally spent hundreds of thousands of dollars on rigorous and time-consuming wind tunnel testing to advance race car aerodynamics. D2H and Ansys have created an automated Fluent workflow that eliminates much of the wind tunnel testing.
The simulation process accelerates the aerodynamic design and lets teams produce more designs without extra development time. It resolves issues in hours instead of days. “Our main development path is to use our aerodynamic experience and flow-field analysis of a current simulation to guide our next move,” Noah McKay, engineering director at D2H, told Design News. “We combine that with periodic, more complex simulations that can provide direct insights into what changes should be made and in what approximate magnitudes.”
Using Workflow and HPS to Accelerate Design
McKay noted that the workflow greatly reduces hands-on development time and resolves design issues quickly. The simulation accomplishes this without sacrificing while remaining highly accurate. “To generate accurate results requires very high-quality meshes. These are typically very hard to achieve on the extremely complex geometry of a full race car model,” said McKay. “Simplification of the geometry is also not an option to get accurate enough results to drive the development of a very refined car where each change can be less than 0.5%.
The design process requires an integrated thread of software tools that McKay describes as a simulation chain. “We have built an automated process that can check and give CAD corrections to get to clean, full-detail geometry and then mesh and submit to solve on a high-performance compute cluster,” said McKay. “The automated workflow combines all of the software tools in the simulation chain including Ansys Fluent and Ensight and gives nearly a single push button from start to end of a process including results and analysis.”
What makes the simulation possible is HPC. The process accelerates the speed of the simulation and guides design adjustments. “HPC has definitely helped. Large solutions run faster than ever, and increased accuracy has driven us to larger and larger models. We have also leveraged improvements in Ansys Fluent to drop solve times,” said McKay. “There are efficiency gains in the computations as well as more efficient processing of large datasets. Additionally, with the higher quality meshes, Ansys Fluent has made it possible to shorten simulation iterations with more aggressive solution controls to drive to convergence much faster than in the past.”
Simulation Can Match or Beat Real-World Testing
As engineers began to utilize simulation – and as the simulation process improved – it became clear that the results from simulation were less time-consuming than wind tunnels. “This is a complicated comparison because we are comparing very different tools. The wind tunnel can produce a result on a single item very quickly with good accuracy. With CFD simulation in Ansys Fluent, we can now achieve the same accuracy,” said McKay. “That is a recent development in CFD to achieve a resolution to understand the small magnitude changes. These large models take hours to solve from start to finish, so they are slower than a single wind tunnel run by orders of magnitude.”
The simulation lets the team increase the amount of testing overall. “The difference is a wind tunnel test can produce approximately 35 runs over a 10-hour test window and depending on your budget, you may do that as often as once per week,” said McKay. “There are now cost containment regulations that limit wind tunnel testing to approximately 100 hours per year, making the total data collected from a wind tunnel much less than can be done in CFD. With CFD, we can run every day and night producing around 5-10 runs per 24 hours or even significantly higher with added HPC and CAD designer budget.”
The Cost and Time Savings of Simulation
The HPC process also offers more information about each potential change in the design. “So over the course of the week, we can at least match, and with appropriate resources, triple or quadruple what can be done in the tunnel. On top of that, every CFD run gives many orders of magnitude more information about each change,” said McKay. “Not only do we know the overall forces and if there is a performance improvement, but we also get insights into the full flow field where we can pinpoint the effect. That allows us to develop much more effectively.”
Race cars are often altered between weekly races. While the cars are optimized at the beginning of the season, they’re often changed on the fly between weekly races. “Racing is a weekly-evolving sport like any competitive professional sport. Top-level teams build new cars almost weekly to incorporate all new developments derived from recent testing,” said McKay. “These developments can be all over and under the car to affect the flow field in advantageous ways.”
Rob Spiegel has covered manufacturing for 19 years, 17 of them for Design News. Other topics he has covered include automation, supply chain technology, alternative energy, and cybersecurity. For 10 years, he was the owner and publisher of the food magazine Chile Pepper.