Automotive engineers know the challenges of aerodynamic shape optimization. Some parts of the car body design are notoriously difficult to aerodynamically optimize, like a side mirror, because they have complex shapes that need to enclose mechanical components of a fixed size.
To keep a competitive edge, and to meet market demands, engineers need tools that can automate and simplify the shape optimization of car body designs.
How to Automate Shape Optimization of Car Body Designs
Engineers can use the ANSYS Fluent adjoint solver to automate shape optimizations — especially those hard to optimize components.
In this case, engineers need to set up a computational fluid dynamics (CFD) simulation and state that the optimization goal is to reduce the drag of the car body design. The adjoint solver then automatically morphs the geometry and mesh of the design, based on previous iterations, to improve the aerodynamic performance.
To demonstrate the capabilities of the adjoint solver, engineers used it to optimize the shape of one of the most complex, and challenging, parts of the car to aerodynamically design — the side mirror. These tests were performed on the DrivAer geometry from the Technical University of Munich.
Small features, mechanisms and mirrors need to fit into the side mirror enclosure without negatively affecting the rest of the car’s aerodynamics.
The position of side mirrors is rather set. They tend to fit on the car’s A-pillar to improve aerodynamics and a driver’s ability to adjust the view. To improve the aerodynamics, engineers need to conduct a shape optimization of the side mirror’s enclosure.
After only two iterations, the adjoint solver automatically morphed the geometry of the side mirror’s enclosure to improve the drag coefficient from 0.299 to 0.286. That represents an improvement of 13 automotive drag counts (or 130 drag counts in the aerospace industry).
Adjoint Solver Optimizes the Shape of Car Body Designs Where it’s Least Expected
There can be areas of a car body design that an engineer may not think to examine to improve the aerodynamics.
For instance, a car’s hood is a large, smooth surface that doesn’t typically add much drag. To save time, many shape optimization specialists focus on more problematic areas.
However, with the adjoint solver, another test showed that changing the geometry of the hood could reduce the car’s drag coefficient, by another 4%, by changing from 0.299 to 0.287, or 12 automotive drag counts (120 drag counts in the aerospace industry).
For more information about the ANSYS product range, please click here >>
This blog was originally posted on ANSYS’ website here.
