Although the Lincoln town car is quickly becoming a thing of the past, this commercial aptly describes what consumers still look for in their cars today: power, style, and comfort. Car designers often provide driving comfort by adding more legroom, plush leather interiors and a quieter cabin. With the rise of the quiet electric car, attention has shifted towards flow induced noise. The protrusion of side mirrors from the body of the car are notorious for causing air flow separation that creates pressure fluctuations aka high-frequency aeroacoustic noise (over 1kHz). The noise is then transmitted via the window into the cabin where the amplitude can vary dramatically depending on vehicle speed.
In order to properly predict the underlying separation phenomena, Honda R&D worked with the Cascade team to conduct a cost/accuracy analysis between two different prediction methods. One of the methods was Wall-resolved Large Eddy Simulation (WRLES) which required a fine mesh of 5.5 billion cells to accurately calculate the separation. Wall-modeled Large Eddy Simulation (WMLES) was tested to see its ability to capture the relevant flow physics through a more targeted, but lower resolution Voronoi hexahedral grid. Combining the Voronoi mesh with the WMLES method, the solver was able to predict the separation with only 59.4 million cells.
“This investigation illustrated that the dynamic slip wall model and Voronoi meshing methodologies which were implemented in CharLES could predict boundary layer separation from a vehicle side mirror at a super-critical Reynolds number… the number of core-hours (this approach) required for such an LES is dramatically less,” and “LES becomes a viable option for industrial use.”
The complete report, “Comparison between Wall-modeled and Wall-resolved Large Eddy Simulations for the prediction of boundary-layer separation around the side mirror of a full-scale vehicle” was published and presented at the 2017 AIAA SciTech Conference in January. Special thanks to the Honda R&D team, Kei Ambo, Takashi Yoshino, Tesuhiro Kawamura, and Minoru Teramura. We appreciated the expertise you shared on your design challenges and enjoyed the opportunity for close collaboration during Kei’s visit to Cascade. Special acknowledgements for the use of the “K computer” provided by the RIKEN Advanced Institute for Computational Science through the HPCI System Research project (Project ID: hp140011).