We measured the rotational drag various wheels in wind tunnel validation of a couple of decades ago. The differences are too small to be reliably detected using the approach you suggest.

Interesting... For whatever reason, from all the slowtwitch reading people always throw around an extra 3-5w or more in the "disc is always faster" threads.

You're saying you observed much smaller than that?

I am saying that the difference between wheelsets (pairs are what we tested) is smaller than that. (Then there is the whole propogation-of-errors issue that would arise from comparing/combining wind tunnel and field test data, but that's a somewhat different issue.)

There is a lot of contradictory information about the loss/gain for watts to spin. Bike TechReview had some reports for sale on differences in single wheel performance. You basically sign an NDA when you purchase the reports, so all I can say is “ya there is a difference”. How these results generalize to actually riding a bike is unknown.

My best guess at getting a decent handle on this number for rider on testing would be to test very similar wheels on the track. The HED Jet 6+, 3+, share the same rim width and profile. I am also assuming the Jet disc+ shares the same rim and airfoil shape as the 6 until is curves toward the hub ( but it might be closer to the 9). If you track test with the same tire/tube on all three wheels you would probably have a decent estimate of watts to spin at zero yaw, though this isn’t perfect either.

Every study that I can recall has found that translational drag is 2-4 times greater than the rotational drag.

Swisside have stated that the measured rotational drag is about 25%. I don't know if that's a quarter of total wheel drag or a quarter of translational drag.

I did some basic calculations and rotational drag came out as about a third of translational. So 25% seems to be a realistic number.

Doesn't the negative watts to spin of the HED 3 at low yaw angles invalidate that as a general rule?

I also think one of the original articles on the subject had the rotational drag of an aero spoked wheel being about 80% (let's say it was about 18 watts) higher than a disc, which they gave translational drag of 10 watts. The 10 watt figure was more or less confirmed by Specialized.

I guess my point would be that I have seen estimates that put the rotation drag difference of a disc and spoked wheel at somewhere between 2 and 12 watts at 30mph. I have NO idea where in this range you might find the "typical" value. On the other hand, I think there are many tests that give you a pretty good idea of typical drag differences of the various aero frames on the market and the range seems to be only 5-6 watts between a P5 and just about any other decent frame.

If you can dig up the studies you mention, we can discuss them.

How did you measure rotational drag in the windtunnel

Method described in this paper:

https://www.academia.edu/...76-291?auto=download

My reading of the paper is :
Spin cranks with no wheel but somehow with chain attached to measure the drive train loss.
Then add the wheel in...

Spin the back wheel with the bike suspended in air by rotating cranks. The only losses you should have are the drive train and the spokes through the air. Since you know the drive train loss from one, the difference can be ascribed up the rotational drag. I think he mentions also measuring the drag on the wheel bearings to take that out of the equation as well.

I think the results of this paper
http://journals.sagepub.com/....1243/17543371JSET17

Are posted here :
https://docs.google.com/...UliYVotdVc2d2c#gid=0

(authors match, but I didn't pay for the articles).

They seem to have an old 808 @ 18w of rotational drag with a disc around 8.5

But this ^ measures the drag of a rotating but not translating wheel. The forces on a wheel which also translates (so the wheel of a moving bike) should be different. The upper part has the greatest velocity and the lower part the lowest. And drag is proportional to the square speed. So I do not think that the rotational drag which was measured here above corresponds to the drag of a wheel in a moving bike. Anyway, here:

http://forum.slowtwitch.com/...ring=theory#p6306742

I tried once to understand the measuring of the drag of a wheel, because I hear often contradictory statements on this topic. I'm still not convinced that my assumptions are correct, but I did not get any reasoning up to date that they're not.