I’m finding the same thing... rising CdA curve but the average seems to track and the average of the averages seem to track, especially when I chop off the beginning and end of each run.

Not sure why it’s happening though.

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

Thanks, below is an example from one of the runs, they are all similar to this one:

:

So there will probably be a few things you will end up investigating, but to start

the "cda chart" is actually a rolling CDA and is calculated on a 60 sec window. So for example the value calculated at 12km is made up of the data for the previous 60 seconds.

At 60km/h it will take 60sec to do 1km. So all the points between 12km and 13km is made of diminishing data from before 12km. Why your CDA is 0.1 at 12km is probably bad data and we would have to see the data between 11km and 12km to figure that one out.

You can set the window to 30sec (or 15) and you will see the CDA stabilize quicker and it should be stable from 12.5 to 13km. It may be higher than you'd expect but you would then need to figure out what is causing it.

Ahh, thanks, this completely makes sense now. So when analyzing this kind of runs it doesn't really make any sense to look at the CDA graph when it has a 60sek delay... But when I highlight the lap between 12-13km I get an average CDA of 0,221. Will this average be based on live data, or is it calculated from the rolling 60sek CDA?

I see now that I have a few things to learn, looks like it will be a long summer;-)

It will be calculated on the data between 12 and 13km which is what I believe is what you want.

But there is something very strange in your data. If your cda is .221 between 12 and 13km there is no reason it should be rising to .3 as in the chart.

You want to get to the bottom of it if you want to be able to trust that number.

Yes, need to have a closer look. My CDA is not .3, so I will need to do some investigation on this. Hope that I will be able to some reliable and repeatable testing with the Notio, but need to get my head around the details.
Thank you for the help so far, really appreciated.

I had an inquiry with a person at Wake County Speedway a while back basically asking if there were days they were ever "open" but no cars there. I told them what I'd be doing. They replied "sure".

My sensor is on the truck for delivery. I think it is time to maybe reach back out to them.

It's no different than an athletics track distance wise, but the corners are banked 7 degrees. Maybe worth a look.

Also wonder about the track by Elon where they run the winter crit series.

I found an out/back right by work that will suit fine, I'd just prefer a loop to avoid all the u-turns.

What surface is on that athletics track? If it's not asphalt (most aren't these days), then you'll see some non-linear effects of Crr (due to the hysteretic properties of the particular surface) "bleeding" into the CdA estimate. I saw this one time when I was testing the Simkins aero brakes...I had done my usual test course runs (on asphalt) and then decided to run a few laps on the local HS running track on the way home to compare. Very interesting results...

---

http://bikeblather.blogspot.com/

Just saw something in that plot - assuming the Notio Elevation variation is correct, if your mass was not entered correctly then your CdA would float up and down in a similar manner to the elevation curve (in this case it appears to be the inverse - your potential energy is being converted into kinetic energy, but if you entered in a 'heavier' mass than your true mass, your CdA would rise I believe as the expected kinetic energy gains are not matching the potential energy transfer).

---

Chris Morton, PhD
Associate Professor, Mechanical Engineering
co-Founder and inventor of AeroLab Tech
For updates see Instagram

It's pavement. On that note, I've been playing around with extreme tire pressures to see if the breakpoint shows up, and it does, so far. I'm not 100% confident in math yet as we've discussed on FB... converting CdA at the speed I test into Crr at the speed I would race at.

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

I work it out like this.

Yes, your assumption is correct, the CDA rises if weight is increased. The elevation should be close to correct.

It only rises going downhill when you enter a "heavier mass" by accident. It would do the opposite on an uphill (CdA would decrease I think).
My thought was that perhaps some of what you are observing in the weird behavior of CdA is related to (i) incorrect mass, or (ii) incorrect elevation curve.
I always start by assuming its a problem with the elevation curve. Even a small deviation in the potential energy changes can lead to fairly large swings in CdA estimates.

---

Chris Morton, PhD
Associate Professor, Mechanical Engineering
co-Founder and inventor of AeroLab Tech
For updates see Instagram

I'm seeing quite a variation in elevation curve, despite being on a flat test route. I'm not sure if it's a shade vs. sun thing affecting Rho or density/pressure.

E

---

Eric Reid AeroFit | Instagram Portfolio
Aerodynamic Retul Bike Fitting

“You are experiencing the criminal coverup of a foreign backed fascist hostile takeover of a mafia shakedown of an authoritarian religious slow motion coup. Persuade people to vote for Democracy.”

Also remember he's testing at 60km/h. A slight error in air pressure measurement would leave to really large CDA errors.

I think he has some debugging to do :-)

Can you post the elevation curve ? There could be a few reasons for it.

Very valid point. I overlooked the speed being upwards of 60+kph!
Generally a tailwind upwards of 10kph, so a relative air speed that is still very high.

---

Chris Morton, PhD
Associate Professor, Mechanical Engineering
co-Founder and inventor of AeroLab Tech
For updates see Instagram

Thanks for all the comments. Can't see any obvious errors in weight, elevation and air pressure, but as you point out only small error could lead to variations. There is not many completely flat roads where I live, so would be good to be able to get reliable results on roads with some elevation. I think I will have to try to do some testing first on flat roads to sort out elevation, just need to find a decent one first...

The virtual elevation curve is the best tool for diagnosing what is going on.

You should be able to figure it out on a rolling hills course so don't worry too much about flat. If anything rolling hills are easier to find problems.

Ya’ll measuring rollout when bike is rider weight loaded with tire flex or with it unloaded?

I would think that may make a cm or so difference in rollout.

I can confirm after only one ride and guessing weight versus using a scale the accel decel idea is true. So old non smart scale goes into the car with rest of kit!

I feel that is also why instructions specify holding constant speed versus power. Accel factors in the math a good bit and is probably tough to deal with.

I am just learning the environment and software side and searching for a good spot.

Needs to be weighted! It's ~20mm IME. For example, 22mm tall tire with 15% sink x Pi x 2= 20.7mm.

I used to worry about this. Then after a few tries I found that GPS distance/wheel count gave me as much or even more precision and one less thing to worry about.



I am not sure what the "accel decel" idea is



You should not have to hold constant speed.

Agreed, not necessary. It is just that if you dont enter a good total weight, accel and decel in elevation induced speed changes is miscalculated due to wrong weight. If speed is constant, obviously accel or decel is irrelevant to the calc of cda. So, if you hold constant speed you eliminate a source of error. Way I see it.

Good to know about the rollout for that speed thing.

I loaded up GC tonight and results look much more constant. App is meh without GC. I’m on my way. Pretty good cda for flappy kit and road helmet with training wheels. Plenty to test in coming months.

A 1kg error on 1% incline will introduce about a watt of error at 36km/h. If you are doing out and backs will introduce no error for the whole of the segment (for altitude)

For accel/decel it will introduce a bit of error, but actually less than not including the moment of intertia of the wheels (I am not sure your device allows to configure). If you accelerate from 28.8 to 36km/h 1kg error will introduce 0.3watts over 60seconds.

Accurate weight is important, but rarely the major source of error in my runs. No need to re-weigh after you pee.

If Robert finds an error in my math, I'll blame it on the beer :-)

If it was weight, it would likely need to be a massive error (like entering mass in kg instead of lbs conversion error, that sort of magnitude). Hopefully that would be obvious!
For elevation and air pressure, this would require more detailed debugging for sure. If you are doing a full out and back section, then at least you can have a reasonable assessment of CdA without the elevation issues (not major issues anyway since the net elevation change is zero on an out and back). But it you are trying to measure "real-time" CdA, then you need good instantaneous elevation measurement (technically you just need the changes in elevation, the absolute value is less important). For air pressure, the full out and back data is important to see, as it will give a better understanding if there is an offset or scaling in the wind that needs to be better taken care of.

---

Chris Morton, PhD
Associate Professor, Mechanical Engineering
co-Founder and inventor of AeroLab Tech
For updates see Instagram