With regards to the Conti Attack, I thought those came in a set with the Force for a rear tire. Does the force run similar or slower?

I previously bought two Attacks individually, but now the twin packs of Force/Comp are considerably cheaper than either one alone. I don't know if the Force has anything different in terms of tread compound and thickness. If not, it should have a bit lower rolling resistance than the Attack. I'll have one to test next time. For racing I think I'd use the Attack front and rear even with a wide rim. My main road wheels are 23mm and it makes a nice profile on those. You could always use the Force for training.

Nice price here: http://www.probikekit.com/...-tyre-twin-pack.html

Does anyone run Hutchinsons? I like the Fusions for training and have run the Atoms for a few races. The Atoms are very light and narrow. I am sure they offer little flat protection so I use them for sprints and oly's only. You can get them for almost nothing at Performance most of the time. Both the Fusions and the Atoms have a decent road feel to me when run with latex tubes (not as good as Corsas but about the same as Pro3s)
Am I the only one that likes them? Can't find much Crr data on them.

Tom, thoughts on the following:

http://www.artscyclery.com/...age-VTOCESCTS23.html

This wouldnt happen to be the same as the SL would it?

Just a week ago, based on some perusing of the forum, I decided to go with the attack/force combo on my race wheels this year. Nice to see the attack validated by Rons work.

---

"One Line Robert"

Yup. That's the SL.

---

http://bikeblather.blogspot.com/

Does anyone run Hutchinsons?

Tour's tests on the ones that take tubes (Top Speed, Intensive, Fusion) show an awful Crr. The tubeless are much better... but the tubeless Fusion is still worse than the GP4k with a latex tube. I've never seen a test of the Atom, but if you are running them tubeless they are probably not bad.

I'd be interested to see what I think could quite possibly be the worst CRR 23mm pneumatic tire tested ever: Continental Super Sport Plus . Why? I ride with these pieces of shit feeling, terribly slow tires all winter in order to avoid flats when it is in the 20s, no chance in hell I'm stopping if I do either. (Thankfully these are stiff enough I can probably ride them without air) Anyways it would be awesome to quantify just how slow these are. I have a tire I would be willing to send you to test, probably 150 miles on it.

---

---------------------
Jordan Oroshiba --- Roadie invading Triathlete space for knowledge access

rruff,

I'm sure it is somewhere... but does the gains you give on smooth roads as indication are for one wheel or for one bike ? A difference in Crr of .0005 or 0,17mph is quite big so if it's only for one wheel then it is a lot for the two wheels.
Something else interesting in your chart : the height of the GP4000S, 49mm, quite higher than the others (the corsa CX is slightly wider but not as high), it probably tells a lot about why it's so aero (elliptical shape with a high height to width ratio).

Thanks for this great work !

Looks like it was resized to 650px Ron, cropping the empty rows and rotating to portrait orientation might make it more legible.

thanks for this. i'd love to do some testing and offer up some comparison data -- i have access to many tubulars.

are you basically just looking at average speed while holding a (relatively) constant power?

i see you have a "power" column--is that the power you were holding? how consistent do you strive to be? for what duration?

or are you aiming to ~35mph and trying to hold speed constant?

my head unit only displays speed to the nearest 0.1kph -- but perhaps if one rides a set distance (or time) then a more precise speed could be calculated.

any tips for someone looking to contribute useful data that may still be valid for comparison's sake?

i have a trainer and rollers available to use.

Could you outline the protocol a bit? It sounds like it isn't the most complicated thing in the world to do. I for one have aluminum rollers and would test and post up data if I could, and am sure a number of other slowtwitchers would as well.

You said it is similar to Al Morrison's protocol but I did some searching for that and failed.

---

---------------------
Jordan Oroshiba --- Roadie invading Triathlete space for knowledge access

I'm sure it is somewhere... but does the gains you give on smooth roads as indication are for one wheel or for one bike ?

Both wheels. Assuming that you are riding a TT bike with a half decent position. It doesn't change that much with weight or speed over a reasonable range.

But I highly recommend you set up a calculator with your personal parameters to see for yourself. That number is intended only to show a ballpark value. Yes, it isn't small. Remember the Jack Watts story about the guy who suddenly got dropped in crits because he mounted Gatorskins. The wrong (or right) tires can have a bigger effect than thousands of $ spent on the best aero equipment.

are you basically just looking at average speed while holding a (relatively) constant power?

I try to keep the speed fairly steady, and look to see if the power appears stable. Initially it will drop as the tire warms up. Then I ride a bit more, and average the last few minutes of output.

my head unit only displays speed to the nearest 0.1kph -- but perhaps if one rides a set distance (or time) then a more precise speed could be calculated.

0.1 kph is plenty good. My Powertap only reads in whole watts, so that is a much less precise measurement.

any tips for someone looking to contribute useful data that may still be valid for comparison's sake?

Tom put up a spreadsheet many years ago for doing roller calculations. Mine's a little different, but I'm sure we can figure it out if you have the rollers and the motivation for tedious testing.

Look for AFM on BikeTechReview.

I'll outline the procedure when I get a chance.

Okay, I have a 2010 Zipp 808, dimpled not Firecrest front wheel. I've been running the Continental 20mm SS for my race tire in combination with the Michelin latex tube. Considering aerodynamics would I be better off using the Supersonic 23mm?

This is great information. Thanks for putting this together and sharing this.

Since the Attack is only sold as a set with the Force (at least from what I've found), do you have the Force to test so it's possible to see the total difference of Crr?

---

---

Fraser Bicycle | First Endurance

Check out my blog here | Twitter:@tmalis3

Similar here. I have a Flashpoint 80 rim. My guess is that the better aerodynamics of the 20mm tire more than makes up for its higher Crr... in most conditions anyway.

I bought the Attacks separately, but will have a Force for next time. I suspect that the Force will be a little fat for optimal aerodynamics... with most rims anyway.

I've found that concentrating on a set cadence value is easier than speed. For my testing, I'll run the setup at 95rpm in a 53x13 for at least 5 min. to adequately warm up the tire, and then I'll do a 4 min. span at 90 rpm. I then take the average power and average speed from the final 2 minutes of the interval. Oh yeah...using km/hr instead of mph for the speed values typically gives better resolution.



That brings up a good point. Getting the "ground speed" measurement accurate is VERY important in the Crr measurement since it's DIRECTLY proportional to the Crr calculated. That's the main reason why I mounted a magnet to one of the rollers and mounted a separate ANT+ speed sensor. That way I know, with the least amount of error, what the "ground speed" is during the test and I don't have to change it from tire to tire. IME, trying to measure the rollout, or even estimate it, is fraught with difficulty and error possibilites, especially since you need to know the rollout ON the roller (which is different than it would be on a flat surface).

As an aside to those who have been following along in this, I'm pretty sure that Ron may not be applying his temp compensation to the Crr values correctly. He applied it to the temp rise of the tires above ambient, while I'm fairly certain that (i.e. the differences in energy dissipation between roller and flat) is already taken care of in the "roller to flat" calculations. Therefore, the temp compensation only needs to be applied using the difference of the test ambient temp to whatever reference temp he wants to normalize the Crr values to...we've been swapping PMs in an attempt to get that worked out. I'm just mentioning this since it most likely means that the absolute values, spread, and possibly even the ranking order of the Crrs in his original chart might change a bit...

---

http://bikeblather.blogspot.com/

The way I'm compensating for temperature changes is way better than not doing it at all...

And regarding the speed. I'm using the tire dimensions and guessing on the amount of compression. It wouldn't be unreasonable to have a relative error +- 1mm in the radius measurement, which would manifest as a +-0.3% error in the speed. IMO that isn't enough to be concerned about... considering that I'd need to buy a speed sensor to measure the roller. At any rate I'll refine the tire compression factor when I can get my wife to take measurements while I'm sitting on the bike.

Heres a good trick: Measure the diameter of your wheels, but set the value of your power meter to the smallest possible diameter. This will give you a too high speed, therefore reduces the procentual error. Transfer the data to your computer and calculate the actual speed with actualSpeed=measuredSpeed* setDiameter/actualDiameter

Actually...I've got data (admittedly limited) that implies it may be worse than doing nothing. See my PM ;-)



I bet if you say "Honey, either I can buy this $40 speed sensor...or, every time I want to do this silly thing of testing tires I'm going to bug you to come over here and measure the distance from the axle of my bike to the roller...while I'm sitting on it...and oh yeah, you're going to have to consistently make that measurement to within +/-1mm, EVERY TIME. What do you want to do?" , she'll say: "Buy the sensor." ;-)

Seriously, that $40 is going to save you a LOT of time and hassle, and make your results more consistent to boot.

---

http://bikeblather.blogspot.com/

I like that...it ties in well with my method of taking the "ground speed" off of the roller itself, which is a much smaller diameter and thus you also get more "triggers" per given unit of time :-)

---

http://bikeblather.blogspot.com/

The used R4 Aero is getting tossed in the post today but I have a feeling it is slightly different than the new model. Bontrager has changed the part number; however, other than the graphics I am not sure what has changed. I don't plan on mounting it up until next month so I won't be able to compare until then.

---

My YouTubes

The Michelin Pro4 Comp seems interesting too! See the article below:

http://www.bikeradar.com/...ew-pro4-tires-36376/

I had Al test a Maxxis a few years ago, along with some Challenge Criteriums that did great. The Maxxis (it was their top of the line racing tire, forget the name) did terribly.