Using PowerPod to Optimize CdA

[DavidK16]

I’m trying to use the PowerPod to optimize my position on my TT bike using the CdA feature. I have a Powertap power meter so I’m less concerned with the accuracy of the PowerPod power readings. My primary interest is CdA.

I’m uncertain though about what I need to do. I looked to the Newton manual for guidance and it seems as though, for the Newton at least, you have to do a calibration ride with a series of coast downs. And to get good data those need to be done on an out and back ride and in full race set-up: wheels, helmet, skinsuit, shoe covers etc.

Does anyone know if that’s the same process for the PowerPod? Is a calibration ride still required if the unit is not going to be used for power but rather for getting wind and CdA readings? And if the calibration ride and coast downs are necessary, can they be done on an outdoor velodrome or will the turns spoil the data?

If anyone has had experience with this and could describe the sequence of steps I need to go through to get to where I want I'd very much appreciate it.

[Velocomp]

This is a fun but tricky measurement. Make sure to check on the "CdA instructions", located in the Help menu of Isaac.

[DavidK16]

John,

I’ve now had an opportunity to read the CdA instructions along with the Isaac Software and Newton instructions. I have also done an out and back ride with a series of 10 coast-downs based on those instructions. The file is attached. The coast-downs were done immediately after the out and back. If you could take a look at the file and give me feedback, I’d really appreciate it. Just fyi, it is a TT bike and the PowerPod is attached to the base bar next to the stem using the mount you supplied. It’s very secure and I take care to position it properly.

A few questions for you:

• Is the Wind Scaling too low at 1.15? Right after the Cal and coast downs I did a series of intervals and the Wind Scaling for the entire ride (Cal, coast downs and intervals) was 1.3. The intervals were done on the same stretch of road.

• The Newton instructions ask you to do things apparently designed for the Newton and not the PowerPod – tilt calibration and cal wind. Since, unless I’m overlooking something, they can’t be done using a PowerPod, is that an issue?

• After doing the above I found a note in the topic on “Anticipated Troublesome Ride Conditions” where you said, in reply to a question, that “coast downs aren’t available in PowerPod”. Is that in fact correct?

• Having played around with the software I can’t see a way of solving for CdA, getting it as output using PowerPod. You can change it by adjusting the Wind Scaling but that feels a little arbitrary. Or perhaps by fine-tuning the calibration using data from a DFPM. Is my understanding right – CdA cannot be directly derived as an output?

I’m having a blast playing around with the PowerPod and look forward to seeing how you develop its capabilities going forward. As someone else on the forum commented, it’s a really great tool to have in the training arsenal.

[DavidK16]

John, I'd really appreciate your input/suggestions regarding my last post. I'd like to be sure I've done what I can to get my PowerPod properly calibrated and that I'm not overlooking anything. I'd also like to know that I understand and am tapping into its full CdA capabilities. Thx. DK.

[Velocomp]

You have entered the land of opposing force esoterica...

PowerPod does not display CdA; that is a feature reserved exclusively for Newton.

However, you can do some tricks with Isaac:

1) Highlight the entirety of the coast down section of your ride, and select the function "Tools/Repair/Repair For Missing Cadence". You don't have missing cadence, but this tool removes the automatic function "when cadence = 0 then power = 0

2) After applying this command you'll see any residual power reported during the coast down. If the CdA is correct then you'll see nearly-zero power after you stop pedaling. Save this file with a different name.

3) Look at the power for the first 10 seconds after you stop pedaling; this is where CdA is most important. Set filtering to "0" so you can see the detail.

4) You'll notice that, in your coast downs, there is a bit of residual power after you start coasting. This could mean your CdA stored in PowerPod is a bit too high (the higher the CdA the more the aero watts will be). There is very little residual power so your CdA value is pretty close...

5) Next, use the command "Analyze/Tweak CdA..." to adjust manually the value of CdA. Try to pick a number that gets you within 30W of zero, just after you start the coast down. For you, on this ride, the number is around 0.2 (which I think is too low).

6) However, you did your coast downs on an uphill. For best results the road needs to be as flat as possible.

You need to ride the same path each time you do a coast down, and the road needs to be as flat as possible. You are starting your coast down on an upward slope of 1.99%. This is causing a bit of additional error. Try to pick a really flat place for the coast downs; this will give you more distance and time as you coast down. This will lead to better data.

I remember doing a ton of coast downs in Dallas TX, in 2005. We had very primitive software and our sensors were anything but optimized. It will be fun to see what comes out of your work...

[benjamin]

Just a thought... would a long descent with a nice steady gradient make a suitable candidate for this sort of analysis? To my mind--and assuming the PowerPod is good with estimating gradient and the contribution from gravity--this would give some nice long samples at a more steady state on which to perform this residual power trickery with? Pedal up to speed, get into normal riding position, coast as long as possible, turn around, climb and repeat?

[Velocomp]

If you're going to use the Coast Down procedure with PowerPod, then no...because on a downhill you don't slow down...

That said, you are correct that our technology does correct for gravity, so coasting on downhills does work with the Newton (the Newton displays CdA on its screen).

As I said previously, you are pushing PowerPod into a regime where it can work, but where its UI has not been designed for the kind of measurements you want to do.

[benjamin]

I'm not suggesting David attempts to get the same results as a traditional coast-down where an absolute CdA figure is derived from delta-V. My thought was in response to your suggestion regarding the removal of the "if C=0, P=0" function in order to display residual power. As he is interested in optimising his position, a way of measuring the relative benefits of a new position against his current position should be good enough. The sort of procedure I had in mind was something along the lines of doing hill repeats with coasting descents. If he were to alternate between his current and test position/helmet/shoes/etc. on each descent, then perform the Missing Cadence repair on the file, a consistently higher residual power on every even or odd descent would seem to indicate that the position used for those descents was the more 'optimised' one. By doing descents a high speed is maintained, thus the inevitable 'noise' is hopefully averaged out over a long sample, and a more measurable amount of aerodynamic drag is achieved at the high speed. I may be completely off here but this procedure, to my mind, appears to be a better way of measuring small changes in CdA than a normal coast down, assuming the PowerPod is consistent with its gradient measurements--greater magnitudes of difference in drag at higher speeds, less noise contributing to errors... and a hill repeat workout at the same time.