Wind tunnel Questions

Triathlon Forum
1

Hey Guys,

Going to the wind tunnel in Seattle on May 10th, looking for some collective wisdom from ST on suggestions for testing.

I have had a few conversations with Len Brownlee and Carl Hoefler and I believe I am in very good hands.

Just looking at some input as to how I should go about this. I read James Haycraft post about pre-testing and his blog about the results…great info.

To start, I own a P5 stinger disk and H3 front. As per advice (zero yaw tunnel) I don’t intend on changing wheels. I will cut shifter cables to make the transitions from run to run easier. Also the base bar is slammed, I like it that way and with hydraulics etc it would be hard to manipulate. Also I have a selector, but I may or may not have the ability to try other helmets. Only issue is that I changed the tire on my H3 and it is 23 mm (only one I had) it will be the same for all runs though

My thoughts on testing:

Run 1) Base line position (high V) with standard tri suit.

Run 2) Base line as above but with castelli body paint full sleeve (I believe old version 2.0)

Run 3) Base line as run 2 but with q-factor or pedal stance at 12 mm wider.
*I want to test this as I have not seen much info on q-factor and how it affects CDA, I have 2 sets of pedals with 49mm and 55mm axels. I am also not sure if wider is better or worse with say a P5 or other “super bike”

Run 4) Base line with elbows tight in about 3cm, will stay with the pedals that offer better CDA

Now……I have 6 runs left, Maybe 4-5.

Obviously I will go lower, I am debating to “waste” a run on road helmet vs selector, if there are no helmets. IE base line crap vs what I think is a fast helmet. Also debating extensions I have some straight and s-bend, it would be easy to swap out with out shifters.

Also debating an out of the saddle run, I think this would be interesting data but likely of little value in terms of making me faster. Also after thinking about it I believe that the side to side motion of real world climbing would be different than being locked in to a stand in the tunnel.

So….I was thinking,

Run 5) low position with elbows tight
Run 6) low with elbows back to normal width.

As some have said in other posts, leave room for trends. I have set up runs 1-6 in order to have the least amount of down time. I am thinking of runs 7-10 would free form or to follow trends or unexpected ideas/positions etc.

Appreciate any feed back.

Maurice

2

I’ve never been to a wind tunnel, but if I did, I would probably just pick 2 questions to answer with the given number of runs you have, to be sure I get them correct.

Like for Q factor I would want to do a run narrow, do a run wide, then do it all again to be sure I’m not fooling myself.

3

Q-factor decision should be based on what feels better when you are pedaling- not wind tunnel.
It is a fit parameter.
Try to work this out before hand.
I am not saying that q-factor could not affect aerodynamics. I am saying that if you ride narrow when your body wants to be wide you
could be inviting injury.

4

Goin fast ain’t easy.

Q-factor decision should be based on what feels better when you are pedaling- not wind tunnel.
It is a fit parameter.
Try to work this out before hand.
I am not saying that q-factor could not affect aerodynamics. I am saying that if you ride narrow when your body wants to be wide you
could be inviting injury.

5

my comment would be that you are not making small changes, you are making large changes and as such, might completely miss the optimal aero for width of drop or might completely mask any trend. 3cm width…I guess you cnat do much with that as it is the incremental change in the aduro pads(assumption of use) unless you simply move arms via measuring. hi to lo…4.5 cm?? quite a change. if not positive it might well be you completely blew past the best point.

helmets…what is the point in testing a road helmet other than to see how much drag you have in training? if not going to race it, there really is no point.

pedals, someone else covered that

out of saddle…hopefully you see the folly in this

you have limited time that you are paying for and you want to throw some of that time into the wind(tunnel) why not better spend the time verifying results, or doing more incremental tests?

keep in mind, some things you find at zero may not pan out in the field when not at zero

6

I think the Q-factor data would be interesting from an aero standpoint, but I agree that you should probably make that decision on what fits better, unless there is a huge difference. Then it might be worth making some sacrifices. I went to the wind tunnel recently as well and went a little narrower and less comfortable with my aerobars, but I have been able to sustain the position. There is definitely a balance there that you have to consider. Personally, I wouldn’t bother with things that would be considered ‘wasted’ runs. The time goes by pretty quickly and it is possible to make some big gains by focusing on key areas. I wouldn’t bother with a standing position. You know it is going to be far less aero. It has its place in racing, but really you should be maximizing your time on the aerobars. Same thing with a road helmet. Unless you have something like a Giro Air Attack or LG Course to compare against your Selector, you’re not likely to ever race in a standard road helmet. Focus on things that you will use and will give you useful data.

Hand position was fairly significant for me. Since you have multiple extensions available, I think that would be good to try. Without changing any parameters and holding your head in a ‘turtle’ position could also be valuable information. I have been doing that as much as possible when racing since visiting the tunnel.

I did a write up of my experience as well if you’re interested in reading it. Since the A2 wind tunnel is local for me, my write up was geared more towards people who I know locally and who might not be as familiar with aero testing (aside from including CDA data). I haven’t posted it to Slowtwitch before so I doubt many people here have seen it. Here is the link…

http://coolbreezetrekbikes.com/derek-kidwell/visiting-a2-wind-tunnel

7

I want too try out the Floyd Landis position. My extensions don’t tilt up separately but wanted to see if a little tilt up would hurt or help.

8

Personally, I wouldn’t bother with things that would be considered ‘wasted’ runs. The time goes by pretty quickly and it is possible to make some big gains by focusing on key areas. I wouldn’t bother with a standing position. You know it is going to be far less aero. It has its place in racing, but really you should be maximizing your time on the aerobars. Same thing with a road helmet. Unless you have something like a Giro Air Attack or LG Course to compare against your Selector, you’re not likely to ever race in a standard road helmet. Focus on things that you will use and will give you useful data.

http://coolbreezetrekbikes.com/...iting-a2-wind-tunnel

What Derek and Jeff said holds very true. Test what is actually useful to you. That other stuff isn’t useful.

Armpad width
Hand position
Head position
Helmets
Apparel
Hydration/storage

You say you’re “happy” with your armpad stack, which I find interesting. You have no idea whether it’s “fast” or not? What if raising your armpads 1cm loses 2 watts but allows your hip angle to be more open by several degrees…?

9

Perhaps I should re-state, I am happy with the base bar where it is in terms of comfort, cornering, handling, safety etc. I cut it last year when initially dialling in fit.

I was caught in the trap of big changes are better than small changes in stack. I see Jeff’s points about 1cm changes, arm pad stack is one of the main things I want to validate or change. So it looks like I can cut a spacer and have enough room to raise 1cm I also have 2.5cm pad spacers which look like they would work.

So I could test from high-v down to low in 1 cm in about 4 runs.

Maurice

10

Thanks for the input everyone,

A few take aways,

Yes, out of the saddle is a wasted run.

Looks like there will be other helmets available, I have access to Selector (my own), an old bell vortex from the 90’s (they say was a very fast helmet) and maybe a few others. So there will be helmet options, but I won’t waste 10 runs testing helmets, will look for guidance on this one.

I agree road helmet is a wasted run.

The Q-factor question interests me, if anyone has any data around this I would be interested. I have 2 pairs of Key-wins, one normal and one at Neg 6mm. In terms of fit, injury etc no issue. I used the normal ones for about 4 years and switched to narrow about 7 years ago. I think the 1.2cm is worth exploring as it could be a fair amount of frontal area. My Q or stance is as narrow as possible now.

Narrow elbows, I am at the most narrow position on the Arduro but am fabing a piece of flat bar to get tighter,

Like I say multiple extensions, I will look for advice from Len and others as to the direction I should head in this regard. IE when a trend is established, what is the quickest way (least runs) to validate or explore

For those of you who have tested, It looks like most people test clothing last.

Is this correct? Validate changes in position first, then maybe head, hands and helmets (or in tandem)….at the end clothing? As I said experts will be there so I will look to them for guidance.

Thanks,

Maurice

11

having plotted out a few different testing sessions a couple of times I think you’re going about it backward.

I’ll echo a few things. Figure out the q angle before you go, it’s a wasted run, the only exception is if you can’t decide and both are perfectly acceptable to you. Then maybe figure it out in the tunnel. it’s a position change, see below.

You need to test your position changes first. Position changes impact how clothing fits. If you test clothing first you need to go back and retest.

Second: test helmets

Then if time permits clothing/storage etc.

Once you nail down position changes the rest of the testing changes are quick. You can change a helmet faster than the operator can save the data and set up for the next run.

12

I use zipp wings to get narower
.

13

having plotted out a few different testing sessions a couple of times I think you’re going about it backward.

I’ll echo a few things. Figure out the q angle before you go, it’s a wasted run, the only exception is if you can’t decide and both are perfectly acceptable to you. Then maybe figure it out in the tunnel. it’s a position change, see below.

You need to test your position changes first. Position changes impact how clothing fits. If you test clothing first you need to go back and retest.

Second: test helmets

Then if time permits clothing/storage etc.

Once you nail down position changes the rest of the testing changes are quick. You can change a helmet faster than the operator can save the data and set up for the next run.

Yes, see above I think I am slowly getting that.

Q is not a fit parameter for me (well maybe…wide hasn’t been proven with power in several years) but I see your point, Q at the back of the session if I have time, it is for general curiosity more than anything, but as I am paying money I should be looking at specific curiosity and trends.

Fit first, clothing last, helmet, head, hands (I guess those are fit) in tandem or more likely the middle.

To Jeffp yes I looked at the Zip adapters, won’t be able to get them in time though. If it proves out I will order.

Thanks,

Maurice

14

a) Only test configurations that have a good chance of being optimal. Don’t use your limited time checking things that you are pretty sure will not be the best.

b) Retest the ones that seem the best a few times. That means get off the bike, reconfigure, retest, etc… not just taking a new data point during the run (which you should also do).

With 6 runs max I’d pick two configurations that you want to compare against each other and test each 3 times. Or maybe try 3 configurations and test each 2 times. But the fewer the repeats, the lower the confidence.

15

a) Only test configurations that have a good chance of being optimal. Don’t use your limited time checking things that you are pretty sure will not be the best.

b) Retest the ones that seem the best a few times. That means get off the bike, reconfigure, retest, etc… not just taking a new data point during the run (which you should also do).

With 6 runs max I’d pick two configurations that you want to compare against each other and test each 3 times. Or maybe try 3 configurations and test each 2 times. But the fewer the repeats, the lower the confidence.

This is a new experience for me so I am trying very hard to go in with out bias, as stated above some ideas just don’t need to be tested (out of saddle or road helmet) or should be tested at the end, assigned perhaps a lower priority at the end (Stance which I can not find info on but am curious about)

I think my initial theory on getting as many runs in as soon as possible was off. As DD stated, or at least what I think he is getting at spend more time initially adjusting for position, IE in the first hour it’s better to have 4 quality runs as opposed to 6 or 7 average runs where you are all over the map with helmet clothes and position.

After that runs should go quickly, maybe defining and adjusting for just one parameter (clothing, hydration etc) allowing for maybe 6-7 runs in the second hour or about 9-11 total.

Thanks,

Maurice

16

Why are you testing only at “0” yaw?

17

That is my understanding of the set up at U of W in Seattle.

Maurice

18

I have tested there and that tunnel is fully capable of full yaw runs. The tunnel is accurate so not using the capabilities will be short changing yourself. Testing at just “0” is ok for cars and airplanes but doesn’t lead you to very good answers for racing a bicycles outdoors. Tunnel time is expensive but leaving with a bunch of bad data and assumptions is just a bad plan. In a two hour test, including setup, you should be able to make 10 solid 0 - 20deg. runs with data every 5deg. this will lead you in the right direction for future development. Don’t waste time on Q factors, test aero helmets only, test a couple of skin suits, use an adjustable stem if possible and work on bar heights and make those changes 2cm at a time. Hand heigth to elbow height would be valuable but I wouldn’t spend the limited time worrying about 1 or 2 cm of elbow width. The time will go very quickly so start saving up for your next test session. Have some fun and take a lot of pictures, it will really help you later on.

19

Thanks, possible I may have misunderstood their run set-up. I’ll re-confirm. The 2 cm drop is a good call as I can make that work for tunnel purposes on the P5 with out touching the base bar.

Maurice

20

You should team up with some academic guy who knows something about Design of experiments (DoE and optimization), possibly found within the field of statistics and pharmaceutical drug research.

I haven’t got the time myself to set up an experiment for you, but I would think that would be the best way to go for you, when having limited runs.