If I understood you correctly, it seems that we should start adding dimples/ridges to undersides of our wetsuits…
Rocket Science sports actually has vortex generators on the back of their wetsuits and trisuits and dimples on the legs. They claim the vortex generators reduce swimming drag by up to 6%.
http://www.rocketsciencesports.com/Rocket_Race_20BPM_Page.php#
http://www.rocketsciencesports.com/Rocket_Race_Suit_Page.php
http://www.rocketsciencesports.com/Rocket_Wet_Suit_Page.php
No problem when you have a dimpled wallet!
To put it in perspective, the Re # over a bike tube (even a HT) is about 10,000 to 50,000 max at 13.4 m/s (30 mph). That’s x100 less than low speed flow over an airplane wing (assume 10x the size of a DT airfoil, 10x the speed = 30cm tall wing, chord of 1.5 m, going through 300 mph wind. Very different flow cases.
Interesting forum thread – this is fun stuff.
MC
Hi Mark,
Do you see any transitional areas (including rider) in your tests or simulations? Given that a Re based on tube member scale is of order 10^4, it seems that a Re based on a longer length scale (ie; rider or treating the whole bike as a slender body) would easily reach 10^5 to 10^6. I suppose this may explain the gains you saw with a dimpled suit on the rider. This is assuming, of course, that laminar separation has not already occurred - admittedly a poor assumption at higher yaw.
Chris
Oh yes, the rider’s body is entirely different. Airflow remains attached much longer over a rider’s back. Over a good back shape, you’ll see airflow attached all the way to the rider’s butt (usually at the large seam of the skinsuit). Dimple-ing/tripping of airflow is really only needed on the parts of the body that don’t run tangential to airflow (the upper arm, upper leg, calf). The back and shoulders are a good top edge of an airfoil and see Re up in a much higher regime. Thus airflow attaches longer on the back and doesn’t separate at the tallest point (like it would on a HT with a much lower Re).
Chris, I read your PM and as a Caltech aero guy, I’m sure you know much more of the theory here than I do. What I can say is that from my experience, the theory, CFD, wind tunnel flow visualization, and wind tunnel drag data all line up really well. There’s still a lot to learn and I don’t think the 3D aero shapes are perfect yet. I’d be interested in having a conversation about your research and would be glad to share mine as well. Our CFD models are pretty primitive and generally just show problem areas in the 3D rather than a home brew code that might yield an ideal airfoil shape. The CFD I do now is very different than the 2D I did in college.
Mark
But…when they are “prefect”…are we looking at a 3 hour IM bike split?
A fun sum I did a while back:
If a cyclist were to achieve the same aerodynamic shape (Cd) as a speed skier, they’d go from 25mph to 63mph.
Oh yes, the rider’s body is entirely different. Airflow remains attached much longer over a rider’s back. Over a good back shape, you’ll see airflow attached all the way to the rider’s butt (usually at the large seam of the skinsuit). Dimple-ing/tripping of airflow is really only needed on the parts of the body that don’t run tangential to airflow (the upper arm, upper leg, calf). The back and shoulders are a good top edge of an airfoil and see Re up in a much higher regime. Thus airflow attaches longer on the back and doesn’t separate at the tallest point (like it would on a HT with a much lower Re).
Chris, I read your PM and as a Caltech aero guy, I’m sure you know much more of the theory here than I do. What I can say is that from my experience, the theory, CFD, wind tunnel flow visualization, and wind tunnel drag data all line up really well. There’s still a lot to learn and I don’t think the 3D aero shapes are perfect yet. I’d be interested in having a conversation about your research and would be glad to share mine as well. Our CFD models are pretty primitive and generally just show problem areas in the 3D rather than a home brew code that might yield an ideal airfoil shape. The CFD I do now is very different than the 2D I did in college.
Mark
Mark Cote
MITAerobike (ST, Twitter)
Specialized Bicycle Components
Road Engineer/Aerodynamicist
Ok,
I see what you are saying now… My very basic issue is not really understanding the variables that go into the reynolds numbers (mark one for the engineer and zero for the molecular biologist). One thing from my previous thought that I still maintain is the part about boats (both above and below water)…do you know much about reynolds numbers of boat hulls. From what you are describing the reynolds numbers of a wing, then the reynolds number of a hull would be even higher…and they definately make use of a tripping mechanism to decrease drag (by a quite increadible amount). Most is available in peer review literature, and some can be inferred from unclassified military literature…you just need to know how and were to look. What are your thoughts?
Stephen J
Anyone else want to do naked stuff with Carry Byron?
yes
Isn’t that why everyone watches the show?
Wouldn’t make the most sense to have dimpled tri suits and helmets before dimpling the frames…I mean we are the largest force in play with drag
Been done: Troxel helmet and Nike Spinsuit. Neither one showed any measurable improvement, and in some cases, tested worse. The Spinsuit, especially, tested quite poorly when Jens tested one, and IIRC, when they actually did get good numbers with it on Lance, it was held in place using all kinds of impractical methods according the folks at LSWT.
I’ve tested the Nike Swift Skinsuit and had excellent results. This was many years ago, but the suit was roughly 100 gF savings over a stock, normal lycra skinsuit. The key was that it needed to fit the athlete. A smaller (stretched out) skinsuit actually tested worse than one that was properly sized – pulled open the fabric too much was our hypothesis.
Sure there have been other great suits designed in the past decade, but that Nike suit really was a step forward for cycling aerodynamics.
MC
I would not say that it tested poorly. It’s just that a Pearl Izumi suit I bought from the late Steve Larsen was a bit faster.
-jens
http://en.wikipedia.org/wiki/Reynolds_number
Re= (densityVelocityLength)/(dynamic viscosity) is the equation I use most.
Wouldn’t make the most sense to have dimpled tri suits and helmets before dimpling the frames…I mean we are the largest force in play with drag
Been done: Troxel helmet and Nike Spinsuit. Neither one showed any measurable improvement, and in some cases, tested worse. The Spinsuit, especially, tested quite poorly when Jens tested one, and IIRC, when they actually did get good numbers with it on Lance, it was held in place using all kinds of impractical methods according the folks at LSWT.
I’ve tested the Nike Swift Skinsuit and had excellent results. This was many years ago, but the suit was roughly 100 gF savings over a stock, normal lycra skinsuit. The key was that it needed to fit the athlete. A smaller (stretched out) skinsuit actually tested worse than one that was properly sized – pulled open the fabric too much was our hypothesis.
Sure there have been other great suits designed in the past decade, but that Nike suit really was a step forward for cycling aerodynamics.
MC
I would not say that it tested poorly. It’s just that a Pearl Izumi suit I bought from the late Steve Larsen was a bit faster.
-jens
Sorry, that’s my bad Jens. I think the new custom PI suits for the Garmin Team are wicked. Apparently they have non-stretch panels that pull the shoulders into a hunch. That’s cool.
…as a Caltech aero guy, I’m sure you know much more of the theory here than I do…
Is this a subtle dig of some sort? =P.
Once again, thanks for your willingness to share info! I’m curious if you guys have investigated whether transition to turbulence is happening over a rider’s back (maybe with a white skin suit and oil film). I ask because given that the flow stays attached so far downstream and Re at mid-back is of order ~10^5, it seems like a pretty robust BL would be needed to navigate the back curvature (perhaps I am over estimating the actual amount of curvature on a “good” back shape). I’m interested because if this is indeed happening, then maybe it’s possible to realize similar benefits by forcing earlier transition on people with less “optimal” back shapes. Anyways, my ramblings are becoming off topic in this thread…I’ll PM you with my contact info so we can discuss further. I’m definitely interested in your research experience (I have some more CFD questions as well).
Chris
i want to set my suit up with velcro to hold the tail of the aero helmet to my back
.
i want to set my suit up with velcro to hold the tail of the aero helmet to my back
.
i want to set my suit up with velcro to hold the tail of the aero helmet to my back
Honestly, I’d use rare earth magnets. Sew one into your suit and then glue one into the tail of your helmet. Use ones about the size of a quarter and they’d hold really well. Plus they will find each other much easier than velcro will. Just put them in proximity and they’ll snap into place. Voila.
good idea, easier to pop them loose when you need to turn yer head
.
How about putting a magnet in the jersey and a ferrous emergency multi-tool in the helmet?
Oh and Chris and MIT - don’t discuss in a PM; it was just getting interesting!
ive said this before but, a multitool just behind the base of my head…no!
=)
How about putting a magnet in the jersey and a ferrous emergency multi-tool in the helmet?
Oh and Chris and MIT - don’t discuss in a PM; it was just getting interesting!
The magnet idea is brillian, Jordan. I’m gonna give that a go with a TT2.
And no Chris, no Caltech dig – I meant what I said. I did very little theory in college, mostly just experimental stuff in the tunnel. You’ll have to talk slowly when we sit down and discuss aero.
BTW – I watched this episode last night because of this thread. Really sparked some interest but it could confuse a lot of people because of the blanket statments being placed on the concept of dimple-ing. And I don’t know what others think, but I couldn’t see any appreciable difference between the cars in the water tunnel (btw running at 1 m/s).
Mark