AeroPod, PowerPod and Newton

I have been using a PowerPod V4 for some time now. I like the principles of operation and find it a pretty good tool for not overcooking it, when climbing. If I compare my heartrate at particular powers between the powerpod when climbing and my turbo, there is a reasonable correlation.

I'm not surprised that the calculation works well for climbing as the energy used to climb is easy calculated and overwhelms any other losses such as drag and friction.

When on the flat or downhill, however, I find the power numbers are very often low compared to what my heart rate is saying and how it feels. It seems to me that the coefficients are wrong somehow.

Looking at the file using Isaac I can see the wind sensor is working.

Attached are two files. One from the PowerPod and one from my Garmin.

If you look at the period between 23km and 24km that from 23.4km my cadence and heart rate are still quite high but the PP is showing very few Watts (sometimes 0W).

I have done out and back rides quite often and the coefficients come out the same each time but don't seem to align with the defaults coefficients. I'm surprised there isn't a coast to standstill required to complete the calculation of coefficients.

Please could you have a look at the files and see if you concur with my thoughts and advise how can I correct for it?

Many thanks

Tyres are 25mmx700c Gatorskins rear 76psi front 70psi. My weight 59kg and about 11kg bike and kit.

This actually is a very tricky situation.

Downhills present very challenging conditions for opposing force measurements, because there are strong beneficial force components (gravity), offset by strong opposing force components (aero). Basically, you are subtracting two BIG numbers: aero watts and gravity watts, to get a small number (pedal watts).

At the 22km mark you are descending a steep hill: about -6%. You are pedaling (cadence is not zero) but your watts are zero. HR is somewhat elevated, suggesting that you are applying power.
What's going on?

The place to start is by looking at the power components of the descent. Gravity accelerates you down the hill (negative power). How much power? About -482W!
Against the "negative" power of gravity is the "positive" oppositional power required to overcome the wind. What is that value? +243W.
Finally, there is rolling resistance, another opposing power, about 46W
So, you have roughly 300W of opposing power (wind and friction) vs 482W of propulsive power (gravity), and another x watts of power from your pedaling.

How big is "x"? PP says x = 0, but is that correct?

A third party calculator says that, for the measured conditions, your power should be zero:
During your descent your HR is 131; at an earlier point in your ride (km 12) HR is also 131; at that location, your applied pedal power is about 72W.

So, let's suppose that your you were pedaling at 72W on the downhill, or x = 72. PP says x = 0.

There are three possible sources of PP measurement error:

1) measured downhill slope is too negative. This could happen if, when leaning forward on the handlebars, your bike flexes, which angles down, very slightly, your PP

2) Your measured opposing wind speed is too low. On the downhill you show a slight tailwind, -5KM. Since aero watts vary as the cube of speed, that error could result in a significant problem, because a tailwind would REDUCE the amount of opposing watts required

3) Your estimated CdA is too low. If CdA is low then aero watts will be low.

In looking at these three factors I find the following:

A. Your mounting angle DID shift very slightly during your descent. I don’t think the actual shift is enough to be a problem; PP automatically corrects things after a short period of time.
B. Your wind speed might be low during the descent. Because your ride file is long and not a true out and back, I can't determine if the tailwind is correct or not. For laughs, I adjusted your profile so that wind speed on the descent is the same as bike speed.
C. I adjusted your CdA up to 0.36. This, in conjunction with the wind correction, is enough to put about 60W into your descent. My only concern about adjusting CdA by this much is that you appear to be a small person, so I worry this is too large a CdA for your body size and ride position.

Note that the effect of the wind correction and the CdA correction is to add about 10W to your overall. See ride file attached.

I’m attaching a profile with changes to your CdA and wind correction. You can try this profile and see if it helps.

PS: Frankly, I would leave your current profile unchanged. When you're pedaling downhill and gravity is giving you about a 500W+ boost, why not just stop pedaling and enjoy the recovery? Also, in terms of overall results any error in this particular situation is truly minimal.

Thanks for the reply and very thorough explanation. From this I will be able to investigate further to tweek the values a little.

The purpose of that ride was to stay in zone 2 but I chose too hilly a route, really, and quickly ended up going with HR instead of power as the power wasn't making a lot of sense except up hill. And that's the point really. If I want to use the PPV3 for a zone 2 power session it needs to be able to read moderately accurately uphill, downhill and on the flat. With this understanding I may be able to get somewhere near that.

Thank you very much.

DaveR62 wrote: ↑Sun Feb 12, 2023 2:15 pm Thanks for the reply and very thorough explanation. From this I will be able to investigate further to tweek the values a little.

The purpose of that ride was to stay in zone 2 but I chose too hilly a route, really, and quickly ended up going with HR instead of power as the power wasn't making a lot of sense except up hill. And that's the point really. If I want to use the PPV3 for a zone 2 power session it needs to be able to read moderately accurately uphill, downhill and on the flat. With this understanding I may be able to get somewhere near that.

Thank you very much.

I agree that a course that does not have steep downhills will be much easier to handle. Holding constant power levels on steep downhills is not easy, for the reasons described in my prior post.

HR is less variable because it is much less sensitive.

I went out for a ride with the new profile today. Clearly the new CdA is too high so it should be set somewhere in between. I have a bit of relatively flat road between two roundabouts which I'll do some testing on on Sunday evening as there should be little traffic or wind. I have also put a little spreadsheet together to calculate the numbers. I'll try riding at constant speed and hopefully constant heart rate and then try and emulate the same on the turbo. From that I should be able to estimate some numbers.

Actual power isn't important to me but I would like to be able to compare turbo sessions with outdoor sessions.

In the profile settings, could you explain Aero and Windscaling, the relationship between them and how to choose values?

Thank you.

DaveR62 wrote: ↑Sat Feb 18, 2023 9:22 am I went out for a ride with the new profile today. Clearly the new CdA is too high so it should be set somewhere in between. I have a bit of relatively flat road between two roundabouts which I'll do some testing on on Sunday evening as there should be little traffic or wind. I have also put a little spreadsheet together to calculate the numbers. I'll try riding at constant speed and hopefully constant heart rate and then try and emulate the same on the turbo. From that I should be able to estimate some numbers.

Actual power isn't important to me but I would like to be able to compare turbo sessions with outdoor sessions.

In the profile settings, could you explain Aero and Windscaling, the relationship between them and how to choose values?

Thank you.

Please post a ride file

Hi John

Please find attached file of my 5 laps round a piece of relatively flat dual carriageway. I tried to stay at 28km/h. Gentle wind was coming from North West. The first three laps were in my usual straight arms on hoods and last two were in a bit more aero position. This was with the profile you sent me.

After the ride I did some data crunching and could see that for the first three laps going South my HR was about 128bpm and for the North section was 137bpm.

I then got on the turbo and tried to replicate a similar session. I attach the .tcx . While I was riding, I estimated that 128bpm corresponded to 125W and that 137bpm corresponded with 135W. This is less evident when looking at the file. Starting at 07:40 I tried constant 137bpm the constant 128bpm repeated two further times. 2m50s into each section, I eased off for 10s to simulate the turn. I was surprised how difficult it was to target a specific HR. My conclusion from that was the lower heart rate (128bpm) was pretty meaningless as my HR didn't drop fast enough but the attempt at 137bpm lined up roughly with 135W.

So I then tried using my excel sheet to calculate the Watts based on the data. I was amazed how closely my excel calcs lined up with the PowerPod with CdA of 0.36 as per the profile you had given, but obviously the Watts are too high for that HR. When I started to reduce CdA to try and get to the watts which correspond to my HR of 137bpm (135W) I ended up using a CdA of 0.235 which is obviously too low for a straight arms on hoods rider like me.

Please could you have a look and see if you agree with the logic I have used here and also explain the relationship between the profile vales of Aero, Windscaling and CdA? I've tried my best to understand but there's something I have missed here.

I couldn't attach the excel so I included a screenshot.

HR is a relatively insensitive indicator of cycling power effort, and it also is a variable indicator. As a training session progresses in time, HR goes up even though power output remains the same.

I'm not comfortable using HR readings from your indoor and outdoor workouts as indicators of comparative power output.

I do agree that a CdA of 0.36 is likely high for you; your weight is a very light 130lbs, from which I guesstimate your body height to be about 5' 4".

If I'm close in your weight/height then, for your hoods ride position, your CdA should be about 0.30.

I don't know where you got the CdA factor of 0.235; that would be a valid number only if you're riding in a TT position, which you are not...

I can't open the .tcx file, I don't know how the Wahoo trainer measures power, and I don't know how good its power measurements are.

The 0.36 factor was "forced" in order to achieve a slight downhill watts. But as I stated in the prior post, this value is suspect, because downhill watts are difficult to measure.

I suggest you use a CdA of 0.30, and move on…

Aero is a factor that is used internally. It is computed from CdA and wind scaling. Wind Scaling is a factor measured during the cal ride. Wind Scaling provides calibration for the wind sensor.

Hi John

You guessed my height exactly right!

To explain my maths. Taking the example below
screen shot1.png Average power was 173.1W of which 132.9W was Aero and 40.2W was everything else
HR was 136.8bpm

I know on my Turbo at that heart rate I can push around 135W at 137bpm. I know from experience there is quite a band where HR=W coincidently.

To get down to 135W, Aero would have to reduce to 94.8

Paero=0.5*CdA*Air Density*(V+Vw)*(V+Vw)*V

If I use
air density of 1.23 kg/m3
V=28.3 km/h = 7.86 m/s
Vw=2.5 km/h = 0.69 m/s
CdA=3.6 m2
I get Paero=127.2W Pretty close to the Paero value above

If I change CdA to 0.30 m2 I get Paero= 106W and total Power of 146.2W still a bit high
This is what I got when I put those values into the profile and applied it to the same section
screen shot 2.png If I change CdA to 0.27 m2 I get Paero=95.4W and total Power of 135.6W Pretty much what I was aiming for.
screen shot 3.png The 0.235 m2 came from my spreadsheet which includes wheel bearing losses and drive train losses which I think is where the difference might be in my results. As you say we're trying to work out small numbers from calculations which produce big numbers of varying accuracy.

The key here is that you corrected Wind Scaling at the beginning and this is what really explains why my power numbers were down. I have been playing catch up to try and understand the maths but I think I am there now and think my CdA is somewhere between 0.27m2 and 0.3m2. As CdA has a linear effect on Paero and windspeed has a cube effect, getting wind scaling right is key.

That was fun, wasn't it? And a good rest day challenge for a retired old engineer

Well, I am absolutely amazed at how good this is now. Just been out for a few hours with a friend and the power readings really do match very well with what I am feeling. I used the 1.4 Wind Scaling at a CdA of 2.85 and to be honest, I'm probably as close as I'm going to get.

The good news now is that I can use the data for training on the flat and not just up hills. I knew the value of knowing power when riding hills so I could pace myself well but the readings on flat roads didn't really mean anything. Now they do!

Thank you, John.