Did you know the way you inflate your tires can ruin the results of a wind tunnel study all together? Data from a wind tunnel test, or any test for that matter, is only as good as the protocol used to collect it. Jon and I have always been very transparent with our testing protocol and we wanted to share some information with you today that might surprise you.
Anytime you test in a wind tunnel you want to minimize the number of variables you have. A variable is anything that is not consistent. Every time you introduce a variable into your test, you risk skewing the data. If the variables are large enough, the data collected in the test can become useless. Here’s a very easy to understand example of a variable, and how introducing it would render the data in a test useless.
Let’s assume that you are comparing two wheel brands in a wind tunnel. You are comparing a FLO 60 to a comparable 60mm deep wheel manufactured by “Wheel Brand X” to determine who has the fastest wheel. You mount a 23c Continental GP 4000s II tire on the FLO 60, and a Specialized Turbo Cotton 24 on the Wheel Brand X wheel, then you run the test.
Our Tire Study data shows that the difference in drag produced by those two tires alone can be up to 196.4 grams. That difference accounts for nearly 60 seconds in an Ironman race and makes it very clear why the “different tire variable” would render the data useless.
When the Variables aren’t Obvious
What’s surprising is how many hidden variables exist. These are variables that many companies and independent testers simply miss when testing. I’m going to illustrate my point with tire pressure.
Jon and I conducted a tire pressure study to illustrate how tire pressure affects aerodynamic drag. In that study, we found that altering tire pressure by only 5 psi can change the aerodynamic drag by up to 57.5 grams.
The solution to this problem seems obvious. Take your floor pump to the wind tunnel, and make sure that you inflate all of your tires to the same pressure. The problem that many testers miss however, is that the accuracy of a standard floor pump is extremely low. According to Silca, the industry standard for accuracy of a floor pump is +/- 5% of full scale. Assuming the scale of the pump is 0-160 psi, each reading is only accurate to +/- 8psi. This means two wheels pumped by the exact same pump can have a pressure differential of up to 16psi. Even when the person conducting the study has good intentions and is trying to keep the pressure the same, they are unknowingly introducing an error of up to 57.5 grams of drag and potentially more.
Always on a quest to find better and more accurate data, Jon and I used Silca’s “The Truth” pressure gauge the last time we were in the tunnel. We have since purchased our own version of the measuring device that we’ve aptly named “The Truth 2.0”. The Truth 2.0 is the most accurate digital pressure gauge made by Ashcroft. This $1200 pressure gauge is accurate to 0.05% of full scale and comes with it’s own calibration certificate. Our calibration certificate shows that the pressure readings taken with our gauge have no more than 0.02% error. With a 200psi full scale, this means the pressure measurements were accurate to within +/-0.04psi. Even considering the labeled worst case of 0.05% accuracy, our gauge is accurate to +/-0.10psi.
Does a small pressure variance exist when using our testing device? It does, but I think we can safely say that we have done all we can to eliminate any variability in our data created by pressure differential.
What Does this Mean?
The next time you are looking at the results from a wind tunnel study, ask yourself how the tire pressure was set during the test. If the study was designed to compare wheels and the tire pressure was set using a standard floor pump, there may be 57.5 grams of error in the measurements you are reading, and potentially more if a the specific tire used in the study displays a different drag variance than the Continental GP 4000s II we used in our study. If the study was designed to compare different bikes and the tires on the bikes were pumped up with a floor pump, you can then double the potential error. This means the results could be off by up to 115 grams of drag simply because the tires were inflated incorrectly.
Jon and I are continuing to improve our own testing protocol and will discuss all of our changes and some other commonly missed variables in this blog series.
I hope you have enjoyed this article. Please leave your questions and comments below.