Inconsistent cdA values with iAero

[peteriley]

Hi Everyone:

I'm trying to iron out the bugs in setting up my iAero. After calibrating the tilt, coast down, and wind I made a series of runs to compute my cdA in the aero position and upright (with my hands resting on the aeropads....very un-aero!). The numbers don't reflect what I was expecting. Initially, I had expected a cdA of perhaps 0.25 in the aero position and say 0.4+ in the upright position. But, I seem to be getting 0.19-ish in the aero and often the same in the upright position! To add to that, I sometimes got larger values in the aero position than in the upright position. I tried tests starting at different initial speeds, into headwinds/tailwinds, and on inclines and declines (the course was a relatively flat, smooth, and semi-closed loop). Below, I've included the notes I took if you find them useful. Can anyone suggest what I'm doing wrong?

Some of the questions I have are:

1. Why are there more than one number for a single coast down? Which one(s) do I use?
2. Why are there such large fluctuations in the values from one test to another and even within a single test?
3. How can my upright position be as aero as the aero position? Obviously it's not.
4. Are there optimum conditions for performing the test? Best initial speed? Incline/decline? headwind/tailwind?
5. What does a 3-digit number, like 0.123, mean? Is that 1.23 or an error code?

Thanks in advance for any pointers, Pete

--------------------------------------------------------------
Notes from Fiesta Island Interval Training 08/08/08:
--------------------------------------------------------------

Performed tilt, wind, and coast down calibrations.

Pre-test:
Ride up to 35km/hr in my standard aero position then coast. I see the following numbers on the display:

0.21, 0.25, 0.24, 0.23 flat/vmax=32km/h/aero

why are the so many numbers? Is it the first, the last or some average?

Test 1
0.19, 0.18 flat/downhill/vmax=32km/h/aero

test 2
0.20, 0.19 flat/downhill/vmax=32km/h/aero

test 3
0.19, 0.18 into a headwind. /vmax=32km/h/aero

test 4
Didn't work - Uphill into wind. Speed not high enough? Deceleration too quick?

test 5
0.19/vmax=32km/h/aero

test 7
0.22, 0.21/vmax=40km/h/flat w/ headwind/aero

test 8
0.18/40km/s/slight headwind/aero

test 9
0.2,0.19/50 km/h/aero

test 10
0.2, 0.19/40 km/h/tailwind-cross?

back to start of loop

test 11
0.19, 0.22, 0.24/upright position (palms of hands resting on aero pads)/34 km/h

The aerodynamic drag from being sat upright should be WAY more than the aero position. Why is the first number so low? And why the drift?

test 12
0.25, 0.24, 0.23, 0.22/41.5 km/h/upright

test 13
0.123, 0.122 going into a hill/corner/headwind/upright

Why 3 digits? Does this mean it's really 1.23? If so, why such a huge change from the previous ones?

test 14
0.21, 0.25, 0.24, 0.23/41 km/h/upright

test 15
0.19, 0.21, 0.20/43.5km/h/upright

test 16
0.32, 0.30/41.5km/h/aero/slight headwind, slight incline/slight curve

test 17
0.21, 0.20,0.19/aero/41.0km/h

test 18
0.124/aero/41.7 km/h

Why 3 digits? This was in the aero position. Why such a large number?

test 19
0.30, 0.31, 0.32/aero/37.7km/h/hit a bump during coast down

Possible these measurements messed up by bump in road

Questions:

What affects cdA calculation? flat/up/downhill? wind: tail/head? curves: straight/bends? road surface - very smooth on Fiesta Island except for a few bumps (one noted). What's the optimum speed to perform test at?

test 20
0.25, 0.23, 025/aero/50.5km/h

test 21
0.122, 0.124, 0.123/37.5km/h,upright

test 22
0.25, 0.27, 0.30/34.7

test 23
0.19/aero/40km/hr
This was after one lap (about 5-6 minutes) at 350-400 watts. Power fluctuated when changed direction (at a bend); presumably because the wind changed direction. Typically the power dropped. I certainly didn't reduce my effort.

half a lap to recover, then second interval....

First lap felt relatively easy. So average power of 350-400 would be about right given my current fitness.

test 24
0.132, 0.18, 0.30, 0.31, 0.32/aero/46km/h
This is at end of second interval. Is the 3-digit number an error signal? Why is second number only 0.18 and the remaining two are 0.31 and 0.32?

Could make plots Adam Hail showed: plots two parameters against each other, the slope of which IS cdA. He computes these two parameters based on the power and speed. In theory the iAero has all this information.

[peteriley]

If I could post a reply to my own question; Adam Haile has pointed out a number of things I can do to resolve these poor numbers, ranging from how/where I mount/position the iAero to how many coast downs I perform, including the 4 mile calibration. I'll re-test on Monday, and post back the results.

[coachboyd]

How clean is the air that the front of the ibike sees? If you have any "wake" coming off your hands that would prevent air from entering the wind port this will be read as a tailwind. Since you are decelerating faster than the ibike thinks you should in a tailwind, this will be read as a higher cda. If you are using the aero bars and seeing inconsistent numbers for cda, this might suggest that you don't have clean air.

I notice that when I have my iaero mounted on my stem using aero bars I will see good wind speed data until I get about 24 mph. Above that it will be very hit and miss and my cda values will be all over the place due to the wake of the air coming off my hands, but for the most part I am showing a higher cda in the aero position than if I sit up. If I have the iaero mounted with good clean wind (for training and cda testing), then I get great and very consistent numbers. I am using this in conjunction with the Quarq crankset.

[peteriley]

Hi Coachboyd:

I think the air is clean...in fact it might be TOO clean. I set up a Cinelli stubby bar on the side of my cow horns to ensure that it had a clean view of the scene ahead. But it might be too much: It's like having it out on a long boom. I didn't see much in the way of choppy data, but the wireless mount is not as secure as I originally thought (the Cinelli bar is bolted very tightly, but I used electrical tape to thicken the bar before mounting the wireless sensor, and I can feel a little 'give' in it). So I'll try with a Topeaks extender and see if that helps. Here's a picture of the setup:

http://supermantis.blogspot.com/2008/08 ... iaero.html

How consistent are your numbers? That is, for a given setup, what's the range of values you might see if you run each coastdown carefully?

Thanks for the tips, Pete

[travispape]

Pete,

I'm thinking that you are using the snapshot mode, right? (Continuous mode is the one that is available when you have a ANT+Sport power meter sending power measurements to the iAero. Snapshot mode is when you measure CdA while coasting.) It sounds like Adam has been corresponding with you helping you get your setup worked out. Like he said, it is critical to get a profile based on multiple coast-downs + a 4 mi out & back ride.

One imporant thing to keep in mind about CdA measurements is that is takes a lot of integration time to get an accurate fix on the number. Traditional field studies typically integrate several minutes of data or more. The iAero integrates approximately 30 s of data, which should give a good trade-off between accuracy and response time. This might be a slower response time than you expect, so make sure that you hold a given position long enough for the iAero to get a fix on it.

It would probably help to pick a day when the wind isn't gusting. The iAero obviously measures the wind and factors it into the calculations, but gusts may add to the noise.

Also, pick a long gradual downhill where you have plenty of time to coast but don't go too fast. If you get going too fast, any little errors in your aero/CdA calibration in your profile will be magnified which would make the snapshot CdA numbers jumpy.

And another thing to point out just in case: the snapshot CdA mode does not factor in any braking that you do. If you apply brakes to keep your speed down, the iAero will think you have deployed a parachute or something and jack CdA way up to account for it.

Travis

[coachboyd]

Something else you could also try was posted by rruff at the wattage group. This one stuck in my mind because I am usually the one coming up with McGuyver like fixes, but I got beat to the punch on this one. I am going to give this a try using those metal supports that hold up the center of blinds for windows. Although, if you have a big distance between your bars and front wheel, this may put the computer too far away to see. I use a pretty aggressive aero position so this one should work for me well.

And you can experiment with different types of brake mounts to get the iBike placed higher and even further out. It should be pretty secure with a piece of metal like that.

[rruff]

The brake bolt mount is very solid, but my previous try with a thinner piece of metal was effected by vibration... so if you wish to try this, go with a thick piece similar to the one in the photo.

I've also been trying to determine CdA via coastdowns with an iBike Pro, and it seems that vibration is a factor there as well, since it effects the tilt readings slightly. If your mount isn't solid, this might be even worse. At any rate, when I do several coastdowns on the same stretch of road, the tilt is slightly different each time, which effects the calculated CdA. So my advice for using the snapshot feature of the iAero is to seek out a smooth stretch of road and make several coasting passes at about the same speed.

Of course you must have a soild mount and a good calibration as well.