Timetrial Helmet Aero Testing Round 2
Two consecutive days in the Centaur wind tunnel. I'm feeling it. This old guy is used to getting up early, but 4:45am is a bit earlier than I prefer.
Ah, but it's a small price to pay for the knowledge gained. Plus I get to see the sunrise; and when I get home I get to have another breakfast.
Sunday's session was a continuation of the previous day's look at Timetrial helmets. Saturday I looked at my Giro Advantage II and a borrowed LG Rocket. Today my plan was to look at Andy's LG Rocket and a Spiuk Kronos he'd borrowed.
Why, you ask, did I want to test another LG Rocket? My reason had to do with the helmets' setup. Andy's LG has the vents taped over and the tail left open. The LG I borrowed has open vents, but the tail is taped over, enclosing it on the underside. I wanted to see whether enclosing the tail had any effect on CdA.
My Giro Advantage was included in the test to serve as a control, a commonality between tests to allow comparisons across tests to be made.
The test procedure was the same as the previous day's: Four runs in each direction through the wind tunnel with each timetrial helmet. Two of the runs at low speed, and two of them at race speed. With 3 helmets to test, that meant a total of 12 full laps.
The chart tells the story. (And here's a bigger version) Basically, I didn't find very much difference between these three helmets. Of the three, the Spiuk Kronos appears to be the fastest helmet. The problem is, it doesn't win convincingly. The margin for error in the tests is probably larger than the margin of victory.
Still, there was some small margin; and believing you have an advantage might give you the motivation to dig deeper and go faster. How much of an advantage? How much more aerodynamically slippery did the Spiuk Kronos test to be? According to my math, at 45 kph, you'd save about 5 Watts with the Spiuk. That equates to about &frac2;second per kilometer, or roughly 20 seconds in a 40 km timetrial.
So is the Spiuk Kronos the one to buy? Hard to say. For one thing, I'd like to test the LG with the tail enclosed and the vents taped over to see what it can do "full aero". For another, I've been offered a chance to test a Rudy Project Syton Supercomp.
Then there's this rumor floating around about a whole rash of new aero helmets to be introduced later this year...
Search
 |
Aero testing
Mark, this series of posts is very interesting, thanks and please keep them coming. As a result I have identified a local "wind tunnel" and intend to do some tests of my own. Just waiting for some windless conditions!
Just one question. Why do you do the runs in both directions? I can't find a completely flat stretch of road, but the one I have in mind is quite sheltered by trees but rises by about 4 feet over a kilometer. I thought that if I kept to the same protocol and always went uphill at least I would have consistent data.
Aero Testing
Colin - Thanks. I'm delighted that you enjoy them. I'll keep them going as long as it takes me to optimize my results, which is likely to be a very long time indeed.
Our "wind tunnel" is very heavily shielded by foliage. The shielding virtually assures any wind will be aligned with the tunnel. And there is wind to be sure. Even on the most quiet and calm mornings there is some air movement, however slight.
We do runs in both directions at any given speed to give the regression method an opportunity to cancel out the effect of the wind. The two data points produced afford a good visual check of the data. When charted, we like to see the pairs bracket the regression curve; and that's easier to see if we know which data points belong together.
And even if we were assured there were no wind at all, having 2 data points would still be good practice. Together they serve as a sort of control, each for the other.
You will need to account for the elevation change in your CdA tests. Even the 0.12% grade I calculate for your wind tunnel represents an appreciable amount of extra work put into and taken out of the system as you traverse through it.
I used some of my own numbers to mock up a couple of virtual runs through your tunnel. One run I set the elevation change to zero - a perfectly flat wind tunnel. The other I tilted the tunnel by raising it up 4' on one end. Both "tests" assumed no wind. That small change caused a significant 0.015 change in CdA.
thanks...
for the larger pic :D
looks good. your partner in crime is stirring up a hornets nest with his results over on ST...
all good stuff - i have three more TTs then back to testing again - i have some funky cranks etc i need to tease out the differences of...
g
Post new comment