Flo30 vs xxx lite wheelset

[mpulsiv]

I wonder whether the brake track can be re-anodized
Anyone familiar with anodization process?

[Illuminate]

Comm - thanks! I

[rijndael]

I wonder whether the brake track can be re-anodized

Sure, it can be done. You're better off buying a new rim, or a different rim, if you're that bothered by something that's considered normal wear and tear.

It's a bike rim, not a fresco painted in the early 1500s.

[BobDopolina]

During their testing Flo calculated the drag of the stand (pylons) used to hold the wheel and removed it from the results (tare). See: http://flocycling.blogspot.tw/2013/04/f ... sults.html. They state this is an industry standard practice.

To the best of my knowledge the data presented to testers at A2 doesn't include the 'tare' calculation and subsequent subtraction in the final results.

Have the results published by November also been "tared"? I looked and didn't see this stated one way or the other.

[NovemberDave]

No, we don't do this. Honestly I wasn't aware that anyone did this until just around the time I was last at A2. I asked about it, and was told that they consider it not to be a correct protocol.

For anyone who sails, the reason why the struts without a wheel would see a different load than the struts with a wheel is obvious: the leeward strut maintains a "bow out," "lee bow," or "safe leeward" (depending on where you're from you'd use one of these phrases, they all mean precisely the same thing) and thus has clear air throughout the sweep from 0* to 20*. As soon as you put a wheel in there, the leeward strut's wind load will get disturbed (decreased) once you start turning away from 0*. The leeward strut will be in the wheel's wind shadow early on in the sweep, and it only gets worse the more you turn. I got so fascinated by this that I started a diagram it, it's much easier to comprehend an illustration, but haven't had the time to finish it off.

So far as I know, no other wheel companies/tests remove the tare. I've read Bontrager's Aeolus D3 white paper about 47 times and they don't mention doing it anywhere in there. Other companies would have to clarify whether they do this or not. We don't.

The raw data that you receive from A2 absolutely does not include a tare load column. I've been there for about two dozen tests (more tomorrow) and have been there virtually for about a dozen more, and in none of this testing were the struts run alone to establish the "tare" value. It's not part of their standard protocol.

[BobDopolina]

@novemberdave.

Thanks for the post. I was indeed a little confused by this as I hadn't seen it mentioned in protocols before. Honestly, I have given the FLO guys a lot of crap recently and didn't want to seem like I was being a dick, yet again. If I'm going to be a dick I want to really mean it...I want to be a dick with purpose.

Based on your post it seems reasonable to ask if you have any idea what effect this might have on results? Would this be a significant factor or lost In the noise?

[Thuekr]

I want to be a dick with purpose.

Spilled my coffee

on another note, instead of the Flo30 rims, I went with the WU6C from Yishunbike - same internal width as Flo30 and 28mm toroidal (60mm deep).

[NovemberDave]

@novemberdave.

Thanks for the post. I was indeed a little confused by this as I hadn't seen it mentioned in protocols before. Honestly, I have given the FLO guys a lot of crap recently and didn't want to seem like I was being a dick, yet again. If I'm going to be a dick I want to really mean it...I want to be a dick with purpose.

Based on your post it seems reasonable to ask if you have any idea what effect this might have on results? Would this be a significant factor or lost In the noise?

As perhaps the internet's leading voice of "you can't compare one test on one day with one tire to another on a different day with a different tire, even if the other tire is the same make/model/size," there's no way I could quantify it accurately.

In a recent test, a Rail 52 and a Zipp 404, each with a GP4000s II that had been moved from the 404 to the Rail (404 was tested first), at 100 psi, at A2, tested at zero yaw at 167 and 170 grams of drag respectively. That is an average of 2 runs each. The 404's runs were 170 and 170, the Rail's were 167 and 166 (we got screwed by rounding up!). The difference between 167 and 170, practically speaking, is not a difference. Shallower wheels had higher zero yaw drag values - the highest was a 24 spoke Pacenti SL23 at 188 grams of drag. Taking the chart that FLO has on their site, their wheels all start at about 140, 141, 142, except for the disc which starts at roughly 125 grams of drag.

At zero degrees yaw, the struts have their lowest exposed surface area. You can get an idea for this here http://a2wt.com/BikeManufacturers.html if you let the pictures scroll. I can measure them tomorrow but I will guess that they are maybe 8mm wide by 25mm deep. So as you turn them, they expose more surface area to the wind. They are rounded rectangles, not an airfoil shape. As I said before, the leeward one stays in clear air as it turns with just the struts, but putting a wheel in the struts puts the leeward strut into obstructed air from the wheel. This same principle is why a rear wheel is aerodynamically less significant than a front. So it's a little bit like measuring a rear wheel alone, and a front wheel alone, and saying that the savings you will see on the bike is the sum of the savings from the front and rear wheel's tested individually. We tested that two weeks ago and quantified about a 50% reduction in aerodynamic savings of the rear wheel when it was put into the bike. Same principle applies to the struts as they turn through the sweep.

FLO does not clarify whether the Mavic Open Pro they show on their chart for baseline comparison has had the tare removed. It's not immediately clear whether the values for the Open Pro were taken from tests published otherwise or if it was tested at the same time as the wheels to which it's compared.

So, from a purely factual standpoint, here's what I can say:
1. The differences between the starting zero yaw values of our test of Zipp 404 and Rail 52 to what FLO publishes are on the order of 25 to 28 grams of drag, which is about 3 to 4 watts. Their values exhibit much less variation among the different depths than our test showed, excluding their disc. We didn't test any discs so let's just exclude that as a data point in this comparison. Their suite of measured wheels includes their 30, 60, 90, and disc wheels. Our suite was Rail 52, Rail 34, Enve 3.4 (front rim), Zipp 404, Kinlin XC279, and Pacenti SL23.
2. The struts are rounded rectangles and show their least surface area to the wind at 0 degrees yaw.

Anything beyond that would be conjecture.

We have no desire to run the struts alone when I'm there tomorrow, knowing that A2 considers this to be an error in protocol I don't want to even touch it.

[BobDopolina]

25 to 28 grams of drag at zero yaw? Scrolling back to the graphs posted that changes things considerably.

I really would like to see the TARE data and how it is handled by various brands in their aero claims as it seems to be significant. Any chance you would be willing to quantify this?

[BobDopolina]

Double post

[NovemberDave]

25 to 28 grams of drag at zero yaw? Scrolling back to the graphs posted that changes things considerably.

I really would like to see the TARE data and how it is handled by various brands in their aero claims as it seems to be significant. Any chance you would be willing to quantify this?

The only valid way to quantify it would be to add back what was removed by the person who removed it. If I was to run the struts alone and add what I got to whatever data has had the tare removed, it would be useless.

I did, however, look closely at where the wheel starts to produce a wind shadow over the leeward strut while I was in the tunnel today. This is the point where I would say that the amount of drag removed from running the struts alone starts to significantly diverge from the amount experienced by the leeward strut when a wheel is installed. It clearly starts to happen by 7.5, and by 12.5 you can not draw a line from the front wall of the tunnel to the leeward strut without first running into the wheel.

I also overestimated the size of the struts before - they are closer to 6mm by 20mm.

[coachboyd]

Dave,

Did you ask about when they tare the wind tunnel before the wheels go in? That in essential already removes the tare from stands. Granted it's not at every yaw that they do this but I swear the tunnel gets zeroed out before wheels go in.

Subtracting the tare again would basically be double subtracting the stands and would definitely give false lower drag numbers.

If anything if you REALLY wanted to remove the tare you should do the yaw sweep measuring the stands. Then when you do the test you subtract the drag at each yaw from the drag at 0 yaw. So if at 12.5 degrees the stands had 53 grams of drag, but at 0 yaw the stands at 43 grams of drag (and already get zeroed out) then you should only subtract 10 grams.
But still if everybody else is reporting get the drag numbers that A2 give us, then maybe that's what we should all report. I think that vast majority to almost all of us do that.

[NovemberDave]

They didn't run the sticks immediately prior to our runs today, which started off with bikes. They could have done it before I got there this morning, but they didn't mention it if they did. The sticks definitely weren't run standalone after changing setup from bikes with wheels to standalone wheels, which were our last tests.

What we use is the normalized straight data they give us, with no further modification. I know that A2 considers that proper protocol, because we talked about it again today. They say point blank that removing the tare is incorrect protocol. It's not industry standard. In A2's words, the only way to back into a standard comparison would be to add back in what's been subtracted out, returning the data to the state in which A2 presented it.

[BobDopolina]

What we use is the normalized straight data they give us, with no further modification. I know that A2 considers that proper protocol, because we talked about it again today. They say point blank that removing the tare is incorrect protocol. It's not industry standard.

Uh-oh!

[pushstart]

In a divergence from the aero topic (I love November's work!), I will throw in that I am on the process of building up a set of Flo30 rims for the commuter bike. I wanted wide rims and despite being a little disappointed that they were 550g (which I guess is on the low end of their published weights), I know that aero benefits -- even if the claims can't be taken at face value -- will out trump the weight. At $80/rim (but $30+ shipping for 2 rims) price was not bad -- and I would get something with 19+mm internal width. Since this is a disc-brake wheelset the anodized brake track is perfect.

Anyway, I just wanted to comment that the finish on the rims is nice, but the published ERD was not consistent with what I measured (I measured 580mm for both of my rims -- and used that number in my build). Also in the stand there is a noticeable hop at the weld of the wheel I just finished; I gave up worrying about it since I am sure it won't be a problem on the bike -- and I would prefer even spoke tensions -- but it bugs me. I might expect that on a pinned rim, but have not seen this on a welded rim. My current wheelset is built on H+ Son Archetype rims and those were really flawless and easily built up round/true.

Anyway, I am sure these will be great wheels, but wanted to mention a caveat about the ERD (measure yours before you order spokes!).