Capstone · Test & measurement
Aerodynamic Load Measurement
A two-person capstone developing measured aerodynamic loads and a line of action for comparison with CFD.
A joint capstone by Carlos Medina Gutierrez and David Stout.
Overview
This joint capstone with Carlos Medina Gutierrez is developing an on-track method to estimate downforce, drag, pitching moment, and the aerodynamic load’s line of action for comparison with CFD. I completed the September theoretical dynamics model and synthetic sensitivity study, which showed how low downforce can make the inferred ground intersection unreliable. The study supports method selection rather than measured vehicle accuracy. A direct calibration-matrix approach remains under investigation, and physical sensor calibration, independent validation, and comparison of two wing settings are still pending.
My Contribution
I completed the theoretical dynamics-model and synthetic sensitivity work documented in the September update, including report preparation. My proposal roles are aerodynamic load analysis, experimental test design, sensor calibration, and software development. Carlos leads aerodynamic modeling/simulations, instrumentation design, vehicle testing/driving, and manufacturing; our final system and validation are shared capstone work.
Engineering Approach
The first model combines tire normal forces, vehicle acceleration and pitch dynamics, rotating wheel/engine effects, and a calibrated drag relation to estimate downforce, drag, and pitching moment. A ground-plane intersection represents the line of action when sufficiently defined. I ran 1,500 synthetic trials per noise case, along with intentional assumption violations and boundary tests. Illustrative noise and reporting thresholds are not measured sensor specifications.
Results & Current Status
At 180 N reference downforce, the synthetic study’s 95th-percentile raw ground-position error was 0.06 m. At 4 N it grew to 33.28 m, and only 0.1% of estimates passed the illustrative reporting thresholds. This exposes the low-downforce sensitivity of dividing by a small force. A direct 3×3 calibration-matrix approach remains an alternative under investigation. Physical sensor calibration, independent validation, and wing-setting discrimination have not yet been completed.
What the Sensitivity Study Revealed
| Case | Reference downforce | Raw ground-position error | Reportable estimates |
|---|---|---|---|
| Steady baseline | 180 N | 0.06 m | 100% |
| Low downforce | 4 N | 33.28 m | 0.1% |
Force and moment can remain meaningful when a ground intersection is unreliable. The model therefore withholds the intersection at low downforce rather than presenting a precise-looking point. The illustrative threshold still needs to be replaced by a criterion based on calibrated sensors and measurement uncertainty.
The same study also found that a missing drag contribution can leave the ground intersection nearly unchanged while producing the wrong force direction. A plausible position alone is insufficient evidence that the recovered aerodynamic load is correct.
Team Credits
Joint capstone: David Stout and Carlos Medina Gutierrez. Faculty advisors: Dr. Shreyas Narsipur and Dr. Vivek Khare. September model and study work on this page: David Stout.
Next Steps
Supporting Files
- Project proposal (signature page omitted) (download PDF)
- September progress report (download PDF)
- Theoretical model & MATLAB sensitivity study (download ZIP)
Downloads are public copies with identifying file metadata removed. Project status and team credits above describe their scope; source packages include notes on their use.