In finalizing the shell for the Raisbeck Aviation High School Green Energy Team’s Solar Car, the objective was to find the best-designed aerodynamic model. As the structural and mechanical design lead, I worked to complete and test our models in both Autodesk Flow and Siemens STAR-CCM+, computational fluid dynamics (CFD) software. CFD software essentially creates a computer-simulated wind tunnel where objects can be tested for aerodynamic properties and real-world performance. We are looking for trouble areas which create excessive drag and slow our car down. Working with the team mentor Dr. Semet and team member Max Welliver, I ran several simulations on our four models and identified the best one, SymLow, with minimal drag forces. Going forward, we need to iterate on the best model and make it easier for us to build.
To make our car more efficient and to place the solar array for optimal energy generation, there must be an aerodynamic shell encasing the car frame. This shell will protect the driver and vital components such as the battery, electronics, and steering system. The top of the shell will hold solar panels, which supply energy to the battery (Fig. 1). We had four different models through previous deliberations and designs (AsymHigh, AsymLow, SymHigh, SymLow) for our car’s aerodynamic shell. I had already run preliminary tests on these models in Autodesk Flow Design to get the relative drag coefficient numbers. Since we were lucky to get access to Siemens STAR-CCM+ from our Siemens contact Mr. Chris Penny, I worked with Dr. Semet and teammates and ran simulations on the four models to calculate the relative drag forces and find potential problem spots that create extra drag. Common sense would have it that a body with the most gradual angles will have minimal drag forces and create the least turbulence.
In physics, a drag force is “the resistance force caused by the motion of a body through a fluid, such as water or air.” In fluid dynamics, drag force is essentially the air’s resistance as the car moves forward through space. The car has to push air out of the way as it goes forward, creating drag. A more streamlined car helps break up the air more efficiently, thus lowering the drag (source: https://www.thermal-engineering.org/).
To collect drag and turbulence data, I modified a preset simulation environment provided by Mr. Penny and ran multiple simulations on our four models. All simulations were conducted at 40 miles per hour. We chose this speed because we drove around 35 mph in the last competition and were hoping to drive a little faster this year and set our performance goals higher.
Our plan has four parts: first, we test all four models in the simulation. Second, we determine which model has the lowest drag force. Third, we will improve the best model using data from the simulations. Finally, we will design and test the fifth model. Test, refine, iterate.
Setting up several simulations was quite the task. Negotiating with a newer version of STAR-CCM+ than the tutorial video, importing files, and missing random components of the car in the simulation led to lots of hours troubleshooting. I understood the software to a sufficient degree through hours of work, allowing me to run the simulations consistently and create a tutorial video for future team members.
All of the data gathered from CFD simulations is relevant to our efforts in improving and finalizing the model. Our challenge has mostly come from the interpretation of the collected data. While 2D graphs, like Fig. 6, are easy to comprehend and draw a conclusion from, 3D data visualizations, like the blue isosurfaces in Fig. 5, can represent various data types and have very different implications on our design. In a recent technical assistance session with Mr. Penny, we learned how to understand some of the simulation data and modify the graphs and plots to provide more relevant and accurate information.
By testing different car models with the CFD software, especially Siemens STAR-CCM+, we determined that the SymLow model has the lowest drag force and therefore is the most promising. The SymLow model has gentler curves and overall convergence to a central plane at the rear wing.
Going forward, it would be best to further refine our model for ease of construction, create a more realistic model in Autodesk Inventor, and run more CFD simulations. A model more accurate to the car in real life will give us a more precise estimate of the drag forces we will experience on the race track in Texas. This all means we have more fun design and simulation work to do and a clear pathway forward. We appreciate Mr. Penny’s assistance and Siemens’ support to make this testing possible.
To learn more about:
Autodesk Flow: https://www.autodesk.com/solutions/simulation/cfd-fluid-flow
Autodesk Design Academy: https://www.slideshare.net/AutodeskDesignAcademy/module-1-drag-force