PF22 Cooling System Design and Manufacturing
I wrote a very simple calculator to calculate how much airflow we would need to cool our powertrain appropriately. If you read through the code comments, it is pretty crude, but it gave ballpark estimates that we started to compare to real life testing data. In this case a simple method was chosen over a transient model because it is so much easier to validate. Heat exchangers are difficult to get right, as the heat transfer coefficient is hard to get without bench testing radiators. Testing throughout the season helped us to collect enough data to estimate this value for the radiators we used.​​​​
​
Testing data is KING! Correlations are so satisfying.
As chief engineer I never should have let the left ducts be built onto the car.
They do not pass the "does it look remotely correct" test...
​
My hunch was correct, the initial air ducts we sized were much too small, due to a massive mass flow prediction error (400%) from the aerodynamics team. I have been bitten before by bad CFD, and transient cooling models that dont work. It was at this point I wrote the MATLAB script shown above. My predictions that we were not getting enough air were correct. Luckily, the radiators were large enough, and we could increase air mass flow with bigger upstream ducting. We iterated on a testing day with cardboard proof of concepts, and finally new carbon fiber ducts for the final version.
​
We ran the system with a 30psi cap. The rules limit us to run distilled water, so max pressure = max temp margin.
We originally designed the system to be fanless, however we saw coolant temps rise significantly in lower wheel speed sections of the racetrack. We added some cooling fans that would only turn on at certain coolant temps and wheel speeds. We did this to save on electrical power. Bypass flaps were added to the post-radiator ducting so ensure ram air would not be choked by the fan.
​​​Coolant temp compensations for ignition timing and redline were also incorporated by the Powertrain Lead.
More aluminum welding! Those inlet tubes are 0.035in wall. The radiator tank caps are about 0.125in wall. Took a few tries to get them to hold pressure.
