Hi Stu,
Do you feel your position was almost identical with each helmet?
Was your speed (air speed preferably) very similar for each test run?
If it both the above are yes then I see no reason to change the calibration factor which is simply making the big (and often incorrect) assumption that your out and back runs had exactly the same wind conditions (opposite direction) and therefore net to 0.

How do the results look with a single CF of 1.365?

What is the course you tested like, is it an out and back straight strech of road?
Do all test have identical start and end points?
Was there any wind?

Hi, Thanks for the reply. To answer your questions:

1- Yes, my position was the same for each helmet. I'm pretty confident of that as ride in the TT position a fair bit.
2- My speed was virtually the same for each run. The run was a circuit on an industrial estate, no cars at all. Great new road surface.
3- The results are below (hopefully you can see - I can't seem to paste in an image or attach a decent size?) with original calibration factor, avg calibration factor and a CF of 1.365 which I got for the first and last runs.




What I am taking away from this, feel free to correct if you think I'm wrong, is:
- The large mistral is likely my fastest helmet
- The POC tempor is also likely to be fast but I should test again
- the medium mistral and the old S-works aero TT helmet are also pretty good but 2 to 3 watts slower.

Cheers,
Stuart

I cannot see the data in that image

Another way to sanity check it would be to use the Virtual elevation model. Notio used to populate the headwind column (I think they left that there).

Yes, pick a factor (or an average) and stick with that for this type of tests. Calibration factor could get us into a 4 page conversation

Thanks, how do you do virtual elevation in Notio GC? It seems to be all locked out. I don't think standard GC has the capability to look at wind data and adjust VE? If it does I'd love to know how to do it.

if you go into the data tab, raw data, look at the columns
There is one called HeadWind I believe. IF it's there, regular Aerolab will pick it up.
It used to be there

As marcag replied I think you are better off using a fixed calibration factor (CF) for all the test runs here.

It would be interesting to have that conversation about calibration factors to have a better idea of when and when not to alter it between test runs and by how much.

In essence we are trying to determine how much the rider+bike is influencing\disturbing the pressure readings at the pitot. It’s plausible in your case that the POC Tempor helmet was allowing a less obstructed initial air flow and thereby having less impact (lower CF) on the pressure reading at the pitot. Subsequently the air flow onto your shoulders and back could then have been worse to give that higher CdA. This is just speculative without any empirical data (ie from wind tunnel) of how the calibration factor is affected by these sort of changes.

If anyone has some actual data on this airflow\pressure disturbance please share so that we can be better informed about likely impact on our test data.

Given that the apparent order of the helmets change depending on if you use the individual vs averaged Cal facs for your session, and that the cal fac for your A run was repeatable, then if this was my data set, I would do a further session to also repeat B,C,D to evaluate how consistent their cal facs are. If they are consistent, then that gives some further weighting to use the individual cal fac data as the differences in the generated cal fac values may be genuine.....

As others have said, a whole discussion on Cal Facs with aerosensors would be valuable for me too given, in my limited testing so far (albeit on a different sensor to the Notio), they are the rate-limiting step for me in being able to do more outdoor testing (as I tried to discuss in Dec 23 on the best aerosensor to buy thread which died off before any real discussion took place).....

Virtual elevation is a way to assess the quality of the estimate, so if you can produce them you might be able to see how "reliable" each of those estimates is.