As I keep working on updating my pipeline for video of cyclist to 3D model to CFD aero drag test for my new beta service https://wind-tunnel.ai I keep running interesting tests and sharing them here and a couple other places.
This one is really just a quick peak because I only tested one speed and one yaw, lone rider, vs a rider drafting a rider, measuring the forces each one has to overcome from the wind specifically.
I’m already working to scale and try to answer this at many yaw angles, 3 speeds, perhaps different rider size configurations, maybe even more riders, etc.
So, how many watts does the rider pulling SAVE from having someone else in their Draft?
This is a quick test to ensure the setup was configured properly. More detailed comparisons and much more data are on the way!
Results at 0° Yaw, 22.4 mph (~36.0 km/h)
Rider Scenario
CdA (m²)
Power (W)
Approx. Reduction from Lone
Lone Rider
0.26
159
—
Front Rider
0.24
150
~6% less
Rear Rider
0.17
105
~34% less
Why does the front rider also gain?
The trailing rider “fills” the turbulent wake, reducing the low-pressure zone behind the leader—so everyone wins!
(Test subject is ~5’7” (~170 cm), ~145 lb (~66 kg). Keep an eye out for future posts more speeds and yaw angles/rider configurations!)
Hopefully you can do an IM legal draft pack of 4-6 riders at 12 m vs 20 m. Been meaning to do it myself in my spare time, but between work and tri (and some laziness), haven’t had a chance to upgrade my OF 6 to whatever is the new version and build the models. It would be great if IM could retain yor services do do this work, have you cold called them? It would go a long way that the increased draft zone in large packs doesn’t make a ton of difference when you have 4-6 guys ahead of you.mor maybe it does.
When I raced Barcelona, the draft effect was noticeable, even riding at legal distance- which most people didn’t. There’s a wall separating the road from railroad to one side, and 5 story buildings with no open space between them on the other. So the riders going through create air current within that tube.
hmm, I do have one video of my front end guy on my TT/Tri bike, I can start turning it into a CFDable model and work on setting up the case.
So, have one where each rider is spaced 12m apart vs 20m apart in a 6 rider peloton? Does that sound right?
I have no contacts in the IM world, I did one once a couple of years ago, but have since been focusing on MTB and Gravel, if I don’t run into any issues, I’ll whip it up (if you’re curious how I got into this, I was trying to make entire montain bike trails in 3D from a ride through with an insta 360, but couldn’t host the result, here’s a protype (for desktop, WASD and mouselook controls) Potree Viewer then I thought smaller, made custom point to watertight mesh software, and figured out how to run CFD tests on it with OpenFoam (I’m using blockmesh, snappyHexMesh, pimpleFoam solver, no slip on ground, slip on other walls, and rotate the inlet/outlet for yaw angles).
I can do -12.5 to 12.5 and 2.5 degree yaw increments, and graph them for each rider, create a image that showcases their drag in watts and CdA.
If there’s a windspeed/configuration/solver you think is best, let me know.
Might be worth DMing Mark Matthews or Kyle Glass from PTN on IG. They’re both good about responding to DMs and they are trying to get IM to make a move on this rule. I’m sure they’d be happy to help you get in touch with the right people. @talbotcox
I recall seeing a 3% benefit during a tunnel test; unfortunately can’t find the source. That being said the deltas are easily within the differences in protocols (rider size/position, drafting distance, wind speed, etc.).
Here’s a test done in the Spech Tunnel…though the protocol is a bit sketch.
Another idea based on all the drafting rule discussions: what advantage does a rider get drafting at various distances (e.g., 1m, 10m, 12m, versus 20m)?
There have been a lot of tests on this (example below). It’s fairly easy to do your own test: on a flat road and a calm day have a friend ride at a constant speed (ex. 35kph) while you follow X meters behind. You can then swap positions and see how much the power differs being in the lead. Second version is ride behind your friend and keep dropping back and see how the power required goes up.
Most of these tests probably overestimate the advantage, since they are typically done at zero yaw. Even a slight crosswind makes a rider following at a distance fall out of the draft (since the trailing rider has to remain in line and can’t echelon).
From J-PB’s graph I posted upthread (TT position @ 45kph):
5m = 20.3%
10m = 13.4%
15m = 10.6%
20m = 8.9%
Interpolate at your leisure.
And have a chew on the effect of motos ahead, behind, staggered ahead, staggered behind, alongside (2m, 1m). At 1m alongside 10% increase in the rider’s drag.
