Velodrome drag testing

Triathlon Forum
1

Attempting to sort out which helmet and visor to use. I’ll also be trying to determine the impact of my near-foot tool kit, front hydration/nutrition system and tire selection.
Saturday’s tests will be 2 miles each at as near a constant average power as I can produce.

Proposed protocol:
1st run Giro Selector without visor
2nd run LG Rocket without visor
3rd run Giro Selector with visor
4th run LG Rocket with visor

Reverse order and repeat three more times for a total of 16 runs.

A similar pattern will be performed on Sunday and Tuesday to determine the impact of my near-foot tool kit, front hydration/nutrition system and tire selection.

Any advice to make the tests and data more telling is appreciated.

2

Indoor or outdoor track?

3

Outdoor.

4

More trial runs? If it’s anything like the track near me, the wind can be a factor.

5

Rather than doing the whole thing at constant power, you could do a range of six speeds for each set up and get CdA via regression. Here is a guide and here is a spreadsheet set up just for this purpose.
Please post your results.
Cheers,
Jim

6

I agree with Jim about not limiting yourself to just constant power. I would suggest the “Shen method” (http://nyvelocity.com/node/3350) and easily pry apart CdA and Crr. You could do this by just plugging the power file into Golden Cheetah’s “Aerolab”.

7

If you send us your power files, we can analyse them. The quality of the results will vary according to how much wind there was, and how closely (and smoothly) you followed the black line.

Do sets of 3 iso-power laps, increasing power from as low as you can go to as high as you can go.

Make sure your power meter is zeroed on each set, and that your tire rolllout, weight, and air density is accurately measured. One of these is highly recommended:
http://www.nkhome.com/kestrel/kestrel-4250/
It will also help to calibrate your power meter. Absolute CdAs are not essential to answering your questions, but it’s always nice to double-check that you are measuring things that are in the right range.

8

Thanks! I’ll do 6 x 2-mile constant speed runs per set up, across a range of slow to fast speeds. I’ll also add Tom’s suggested run, a steadily accelerating 2-mile ride. I’ll post results for feedback, then adjust my protocol, if needed.

9

Are you suggesting all my runs build, or that I add one run of the Shen Method to the six runs Jim recommends?

10

Does the Shen Method cover your suggestion? Is one or two constantly increasing power runs per set up enough?

11

Are you doing this on a moto?

12

Does the Shen Method cover your suggestion? Is one or two constantly increasing power runs per set up enough?

My suggestion is just a variant of the Shen Method. Yes, one or two increasing-power runs should do it.

13

Hello Andy and Tom:
How does Golden Cheetah and Shen method account for lean angle on a velodrome? In our spreadsheet we have a tab for road and a tab for velodrome. The velodrome tab accounts in a crude but effective way for the difference between the speed of the wheels and the speed of the center of mass. This may not matter too much in an A-B comparison but then again it might.
Also to Damon: I can understand the attraction of longer trials. However, the downside to longer trials is the total time taken can allow substantial changes in weather conditions. If it were me, I would recommend trials of perhaps 2 laps. The key is to get the speed at the start and end of the interval so that you can account for changes in kinetic energy.
Cheers,
Jim

14

Sounds like a ridiculous amount of trouble to go through…

I’m fairly confident any helmet would be more aero with the purpose-built visor attached, but if the visor isn’t comfortable then I suppose I could understand the value in trying to figure out whether one helmet without the visor is more aero than the other with the visor.

As for the tool kit, what is testing with/without going to do for you? I imagine that you’d use it regardless, so why bother testing it? Same for the front nutrition system.

I don’t think an outdoor velodrome is a controlled enough environment to produce meaningful aerodynamics tests. You’ve got uncertainties in winds, distance covered, lean angle, power data collection hardware, and air density, among other things. It seems the tools you’re using simply aren’t accurate enough to truly discern the minute difference that you will expect to find. Take the tires for instance…based on AFM data you’re looking at rolling resistance differences of somewhere on the order of 10-15w maximum, assuming you’re comparing tires that are fairly good to begin with. That’s well inside the range for what I’d consider “noise” on the average power meter (powertap…+/- 5% @ 200w= 10w). If you’re using a more expensive SRM system you have a much higher chance of getting repeatable data, but still, I think you’ll have trouble controlling/accounting for all the variables…

15

Hello Andy and Tom:
How does Golden Cheetah and Shen method account for lean angle on a velodrome? In our spreadsheet we have a tab for road and a tab for velodrome. The velodrome tab accounts in a crude but effective way for the difference between the speed of the wheels and the speed of the center of mass. This may not matter too much in an A-B comparison but then again it might.

I think this is a good feature to have. Wheel speed has more variation, allowing you to identify features a lot better. But the physics lends itself better to tracking the COG, since that’s what Newton’s Second Law is describing, I guess.

16

Sounds like a ridiculous amount of trouble to go through…

Or you could fly yourself, a coach, and all your gear thousands of kms to the nearest wind tunnel… :wink: (I’m not dissing wind tunnel testing, here, btw. It is the gold standard, and permits quick equipment changes. But it’s not what you’d call more convenient or accessible to the average Joe, by any stretch. )

(Edit: Added the word “more” in above paragraph.)

I’m fairly confident any helmet would be more aero with the purpose-built visor attached, but if the visor isn’t comfortable then I suppose I could understand the value in trying to figure out whether one helmet without the visor is more aero than the other with the visor.

That’s the mystery that is aerodynamics. Surprisingly, visors don’t always improve aero efficiency.

I don’t think an outdoor velodrome is a controlled enough environment to produce meaningful aerodynamics tests. You’ve got uncertainties in winds, distance covered, lean angle, power data collection hardware, and air density, among other things. It seems the tools you’re using simply aren’t accurate enough to truly discern the minute difference that you will expect to find.

You might be surprised at the accuracy of a carefully-controlled VE test on a track. Modern sensors can do wonders…

Take the tires for instance…based on AFM data you’re looking at rolling resistance differences of somewhere on the order of 10-15w maximum, assuming you’re comparing tires that are fairly good to begin with. That’s well inside the range for what I’d consider “noise” on the average power meter (powertap…+/- 5% @ 200w= 10w). If you’re using a more expensive SRM system you have a much higher chance of getting repeatable data, but still, I think you’ll have trouble controlling/accounting for all the variables…

Concerning tire Crr, I don’t think you’ve fully grasped the implications of The Shen Method (http://nyvelocity.com/node/3350). I’d suggest you have a good read before writing it off completely.

17

Or you could fly yourself, a coach, and all your gear thousands of kms to the nearest wind tunnel… :wink: (I’m not dissing wind tunnel testing, here, btw. It is the gold standard, and permits quick equipment changes. But it’s not what you’d call more convenient or accessible to the average Joe, by any stretch. )

Or you could do neither…that’s what I’m saying. Seriously, something as minute as a visor on the helmet is not worth this much trouble (at least not to me). Fitting yourself on a brand new bike might be worth the wind tunnel time and money, but something as minute as this certainly isn’t, particularly when you consider the magnitude of the potential gains relative to the time and money expended.

That’s the mystery that is aerodynamics. Surprisingly, visors don’t always improve aero efficiency.

hmm…proof please?!?

You might be surprised at the accuracy of a carefully-controlled VE test on a track. Modern sensors can do wonders…

Carefully controlling everything else can’t fix a +/-5% variability in power output readings. I doubt that any of the other things this guy is testing will have a significant enough effect that anyone could definitively say that the improvements were a result of the change in equipment and not simply variability in power readings. I’m sorry, but I just don’t think that this will even give you a good first approximation of anything but the effects of the largest changes to equipment or position. I tried searching for the results of tests conducted with the Chung method to try to find the smallest difference in drag that they could discern, but my search was fruitless. I could possibly be convinced otherwise if you send me some examples. I feel like I should have mentioned this before, but, supposing that a carefully controlled VE test is accurate enough, this doesn’t sound like it’s going to be carefully controlled enough to be accurate. Perhaps that is the rub.

I doubt the sensors as much as I doubt the rider. In static wind tunnel testing the rider can throw off the test with even the slightest positioning change (knees in/out, back arch, head position, etc.) This is why companies that actually care about test integrity (like cervelo) use a dummy rider. I just don’t see how a rider could maintain EXACTLY the same position for miles on end on a velodrome.

18

Or you could fly yourself, a coach, and all your gear thousands of kms to the nearest wind tunnel… :wink: (I’m not dissing wind tunnel testing, here, btw. It is the gold standard, and permits quick equipment changes. But it’s not what you’d call more convenient or accessible to the average Joe, by any stretch. )

Or you could do neither…that’s what I’m saying. Seriously, something as minute as a visor on the helmet is not worth this much trouble (at least not to me). Fitting yourself on a brand new bike might be worth the wind tunnel time and money, but something as minute as this certainly isn’t, particularly when you consider the magnitude of the potential gains relative to the time and money expended.

If you mean by this that aero testing is not for everybody, then I will wholeheartedly agree with you. I mis-understood your direction on this. However, for those who want to know these kinds of things, there are ways to find them out.

That’s the mystery that is aerodynamics. Surprisingly, visors don’t always improve aero efficiency.

hmm…proof please?!?

Look through this Slowtwitch post and you’ll get some firsthand opinions:
http://forum.slowtwitch.com/...um.cgi?post=2566897;

You might be surprised at the accuracy of a carefully-controlled VE test on a track. Modern sensors can do wonders…

Carefully controlling everything else can’t fix a +/-5% variability in power output readings.

You can get better-than-manufacturer specs out of a power meter by doing a careful calibration. Manufacturers quote specs that they can support for the masses. This doesn’t mean you can’t carefully re-calibrate your power meter to get much better than 5%. Again, it’s a question of wanting to do it…

I doubt that any of the other things this guy is testing will have a significant enough effect that anyone could definitively say that the improvements were a result of the change in equipment and not simply variability in power readings. I’m sorry, but I just don’t think that this will even give you a good first approximation of anything but the effects of the largest changes to equipment or position. I tried searching for the results of tests conducted with the Chung method to try to find the smallest difference in drag that they could discern, but my search was fruitless. I could possibly be convinced otherwise if you send me some examples. I feel like I should have mentioned this before, but, supposing that a carefully controlled VE test is accurate enough, this doesn’t sound like it’s going to be carefully controlled enough to be accurate. Perhaps that is the rub.

Again, let’s not confuse whether the average Joe can get it to work vs., say, someone like Tom Anhalt. There are those out there that want to be Joes and others that want to be… well… er… Tom Anhalts. :slight_smile:

Look up the Simkins Egg brake stuff. It’s extremely easy to find. Aw, heck… let me do it for you. :slight_smile: Here:
http://www.google.com/...iDmgryLg&cad=rja

I doubt the sensors as much as I doubt the rider. In static wind tunnel testing the rider can throw off the test with even the slightest positioning change (knees in/out, back arch, head position, etc.) This is why companies that actually care about test integrity (like cervelo) use a dummy rider. I just don’t see how a rider could maintain EXACTLY the same position for miles on end on a velodrome.

You’re totally right about this. Dummy riders cost a lot less, too. :slight_smile: Aero testing of any kind requires that you care about all the details. Again, for those who care, though … it is possible.

19

Make sure your power meter is zeroed on each set, and that your tire rolllout, weight, and air density is accurately measured. One of these is highly recommended:
http://www.nkhome.com/kestrel/kestrel-4250/
It will also help to calibrate your power meter. Absolute CdAs are not essential to answering your questions, but it’s always nice to double-check that you are measuring things that are in the right range.

Thanks for your guidance! Kestrel makes a bunch of products. Does the 4250 have something the others don’t? Will the 2000, 2500, 3000, 3500, 4000 or 4500 get the measurements needed? Would wind direction measured in the center of an outdoor velodrome be a valuable measurement? What measurements are needed?

Do you have a preferred method of measuring tire rollout?

20

Make sure your power meter is zeroed on each set, and that your tire rolllout, weight, and air density is accurately measured. One of these is highly recommended:
http://www.nkhome.com/kestrel/kestrel-4250/
It will also help to calibrate your power meter. Absolute CdAs are not essential to answering your questions, but it’s always nice to double-check that you are measuring things that are in the right range.

Thanks for your guidance! Kestrel makes a bunch of products. Does the 4250 have something the others don’t? Will the 2000, 2500, 3000, 3500 or 4000 get the measurements needed? What measurements are needed?

Yes, the 4250 computes air density. There might be an upgrade from the 4200 to convert it to a 4250. But I find it extremely convenient to get air density right away.

Do you have a preferred method of measuring tire rollout?

Hmm, that’s a good one. There are people far more experienced than I am at this. But as a first cut, do at least 10 laps at speed on the black line. Readjust tire diameter so that you get the correct distance.

There are 3 sensitive data fields: speed, velocity, and mass. Get those as close as possible. In our case it’s meant getting proper calibration equipment for checking scales.