New Mexico Supercomputing Challenge

Lowrider to Flowrider - Vehicle aerodynamics and aeromodding

Team: 53

School: Los Alamos High

Area of Science: Physics and Engineering


Interim: Computer modeling of 3D objects is not a new idea, and neither is the modeling of aerodynamics of a moving vehicle. Our goal is to build a program with the ability to easily import 3D models and calculate the air drag on the model. We will then make changes to the model, possibly based on suggestions from the calculations, which we will then mirror on a real-world car to verify our model.
According to the Department of Energy, wind resistance accounts for 8-10% of losses in a conventional vehicle. Most of the losses in a vehicle are due to inefficiencies in the internal combustion engine, which is why many manufacturers are turning to electric engines which usually only lose 5-10% of the energy put into them, rather than 60% lost by gas engines. As engines become more efficient, air resistance becomes a large percentage of energy lost, making it more valuable to target. Electric vehicles need to aim for at least 300 miles to meet the demands of an average American.
We plan to create the aerodynamics model in C with the ability to import vertices and faces from Google Sketchup. Our focus will be on resolving the geometry of our solid and moving our fairly simple CFD model to the GPGPU.
We have a simple 2D clipping algorithm working using a traditionally graphics-oriented algorithm to determine new vertices for clipped cells from normal rectangular mesh. We know we need to export values for new cell volume, area of cell interface, and normal of boundary in the cell as the relevant data in the cut-cell CFD. We have a starting point on a simple, regular grid CFD code. The Harlow text discusses the same kind of simple up-winding CFD model and putting in objects by blocking off entire cells. Once we integrate the CFD code and cell clipping based on modifications presented in the cut-cell simulation of atmospheric flow, we can use predictions of objects obstructing flow given in Harlow as a basic quick check. To import models from Sketch-up, we have written an interface with ruby to export geometry information.
The real-world tests have not proceeded as planned due to hardware difficulties. We have two electric vehicles to use to test, but neither has run for long enough to use. One vehicle has an unfortunate selection of different lithium cells which seem to have failing wiring, the other has just been finished using a controller which had to be imported from China. Unfortunately, it was ordered as a 120V controller, but we discovered after installing it that it is actually a 72V controller! The problem is there is nowhere to buy a high-powered controller for our type of motor. Therefore, the project has taken a slight diversion to build a new controller based on open plans. The controller will need slight modifications to work with our motor. We are confident we can build it, and the people who published the plans have made a suggestion for the modifications we need.
The control board for the new motor controller is already finished. Soon we should have the parts for the power board. Once we know the controller works as it should, we will make the modifications needed to make it work with our motor.
3D cell clipping is not a straightforward as 2D because it requires testing for whether a point is contained in a polygon.
After the completion of our model, we will publish it for use by anyone wanting to aeromod their car. Because the model will have been tested against the power usage of a real vehicle, it will be accurate enough for anyone to use to consider the impact of modifications.

1. Beauchamp, Warren. "The Recumbent Bicycle and Human Powered Vehicle Information Center." The Recumbent Bicycle and Human Powered Vehicle Information Center. Recumbents.com, May 2009. Web. 21 Oct. 2012. http://www.recumbents.com/car_aerodynamics/.
2. "Introduction to Finite-Difference Methods for Numerical Fluid Dynamics" Evan Scannapieco, Francis H. Harlow
LA-12984 (UC-700 (??)) Issued Sept 1995
3. "Open ReVolt." EcoModder. N.p., 13 Oct. 2012. Web. 11 Dec. 2012.
4. Ivan Sutherland, Gary W. Hodgman: Reentrant Polygon Clipping. Communications of the ACM, vol. 17, pp. 32-42, 1974
5. "Well-balanced compressible cut-cell simulation of atmospheric flow" R. Klein, K. R. Bates and N. Nikiforakis
Philosophical Transactions of the Royal Society A - Mathematical, Physical & Engineering Sciences (Phil. Trans. R. Soc. A 2009, 367, 4559-4575 dpi: 10.1098/rsta.2009.0174)


Team Members:

  Rachel Robey
  Simon Redman
  Gabriel Montoya

Sponsoring Teacher: Lee Goodwin

Mail the entire Team

For questions about the Supercomputing Challenge, a 501(c)3 organization, contact us at: consult @ challenge.nm.org

New Mexico Supercomputing Challenge, Inc.
Post Office Box 30102
Albuquerque, New Mexico 87190
(505) 667-2864

Supercomputing Challenge Board of Directors
Board page listing meetings and agendas
If you have volunteered for the Challenge, please fill out our In Kind form.