Ever wonder how all those marketing and braking performance claims from carbon wheel manufacturers would hold up if everyone were thrown into a giant test, all variables removed, and one winner could be announced?

Alto Cycling did, too. And they did something about it.

Partnering with Spark Wheelworks to create a test, because they wanted to know how much of what was being claimed would hold up in the lab. So, they called in rims from the following brands:

  • Alto Cycling (CC52)
  • Bontrager (Aelous 5 TLR D3)
  • Boyd Cycling (60mm Clincher)
  • ENVE (SES 4.5)
  • FSE (EVO 55C)
  • Knight (65 Clincher)
  • Mavic (Cosmic Pro Carbon SLC)
  • Roval (CLX 50)
  • Zipp (404 Firecrest)

Then they designed the study to remove as many variables as possible. If you want to see the videos, skip to the bottom. If you care about how they designed the experiment and why some rim brands weren’t included, keep reading…


2018 Alto Cycling filament wound carbon rim brake wheels
Alto’s new rim after phase one of the test shows no damage.

To be clear, this is a test dreamt up by Alto Cycling’s team and performed by Spark Wheelworks. Other than Alto providing their own wheelset for the test, all others were procured at wholesale by Spark as part of their sponsorship of the test.

Alto co-founder Bobby Sweeting says they wanted to include the best wheels, and they planned to publish the results regardless of who won and how their own rims performed. But they weren’t going into it blind. They’d tested prior generations of their rim brake carbon clinchers and new they could withstand a lot of heat. And then they made them better:

“We began processing different resins with different additives (which are now proprietary and will not be disclosed), and researching ways in which we could increase the onset temperature of the resin. Essentially, that is the point at which the resin begins to soften. We then decided to filament wind the brake track while infusing these additives, which gives the fibers much tighter compaction and allowed the resin to flow more evenly throughout the structure. In our first round of prototype testing we were failing tubes and rim strips (more on that later), but were never able to fail one of the new rims!”

I asked why Reynolds wasn’t included, and Sweeting says they reached out, but didn’t get a response in time for the experiment, and they weren’t available through Spark’s accounts. The other two that Sweeting says he really wanted to test against were Shimano and Campagnolo, but their budget didn’t allow for retail purchases of any wheels (hence Spark’s sponsorship support). He told us Campagnolo was willing to send their wheels on the condition that they be tested with Campagnolo pads, but as you’ll soon see, that would have introduced a variable that they couldn’t accept. Sweeting says they’d like to a second round to include more brands in the future, and they’re using the feedback from this first round to hear which brands all of you want to see tested. So, let ’em know in the comments.


2018 Alto Cycling carbon rim brake performance comparison test against Zipp Knight Boyd Bontrager Mavic and more wheel brands

New rims in hand, it was time to design the experiment so all brands would face the same test and results could be trusted. Here’s how they did it:

“We wanted to be absolutely sure that we were testing the heat capacity/dissipation of the rim, and not the tube, rim tape, etc. We didn’t want any premature failures other than the rim itself. In the prototype testing of our own rim we found two extremely interesting bits of information: a nylon rim strip would melt, and a small tube would fail pretty quickly! So the first thing we did was develop our new high temperature tubeless tape, which has a melting point of 915ºF. That solved that issue completely. For the tubes, we saw a massive difference when we switched to a 25-32c Continental tube, as opposed to their 18-25c option. Even though we were testing with 25mm tires, the wider rims allow that tire to have an actual width around 27mm. So the 18-25c tube was too small, and the wall of the tube was failing at sustained temperatures over 200ºF. So, as two side notes for customers to take away from this test: make sure you’re using quality rim tape and that you’re using the largest tube possible for your setup! Every rim in this test used our rim tape and the same sized Conti tube, of course.

“There are A HUNDRED ways that we could do this test, but we wanted to focus on eliminating as many variables as possible so that we could focus on the rim. That meant standardizing the brake pads, and using the same generic carbon pad on every rim. In this case, we used SwissStop Black Prince pads. Of course the times and temperatures will be different with brand specific pads, but we found that as long as the pads are the same across the entire test (regardless of what is used), the actual results of how brands compared with one another remained the same.

“We also decided to let the speed of the wheel be a floating variable, and to set the power input from the motor at 1200 watts. This way, the total energy into the rim would be constant, and the speed would simply be affected by whatever the rim manufacturer was doing to their brake track. We felt that this would be more accurate than controlling wheel speed and forcing the motor to do more work on some rims than others.”

The tires were inflated to 100psi. In addition to the heating and braking, they also tested how quickly the rims would return to 120ºF, which is a normal between-braking-efforts temp. Since few of us actually drag our brakes continuously for an entire descent, an important measure of performance is how quickly they can shed heat. The faster it cools, the more likely it is to make it to the bottom intact. When you’re watching the video below with the cooling times listed, keep in mind that in real world conditions you’d have a lot of airflow over the rim surface to aid in cooling, which would almost certainly reduce the times listed.


The variation in results is dramatic, some definitely fared better than others. Here’s the “after” photos, as provided by Alto Cycling. In some, you’ll see some minor delamination, in others the resin softened enough to blow the sidewall apart. In both cases, failure ultimately happens because the resin softens to a point where the tire pressure is stronger than the resin’s ability to keep the structure intact. HOW the rim fails, as Sweeting explained it to us, depends on variables like onset temperature, fiber orientation and layup schedule, molding technique, etc.

ENVE (left) and FSE (right)

…and how did Alto fair? Here’s their rim after Phase 1, which used 7lbs of brake lever force for all of the rims tested:

Alto Cycling experiment shows how different brand carbon clincher rims compare for braking heat tolerance

After 20 minutes, it hadn’t failed and showed no signs of damage. So they increased the brake lever force to 9lbs and ran it again…

this test shows how does braking heat affect carbon rim brakes

After another 20 minutes, there are initial signs of delamination, but the tire hadn’t popped yet. Full results are:

alto cycling carbon rim brake heat tolerance test results showing comparison between bontrager boyd enve fse zipp roval and mavic carbon rimsDoes this prove who makes the best carbon clincher rim? No, but it’s a start. More brands will need to be tested, and we applaud Alto and Spark for putting all the information out there for us to see and process. Adding to the appeal for other brands to be involved, Sweeting added:

“I’d also like to be clear that this was simply one way to do this test in order to collect a specific type of data. We haven’t tested every rim with their brand specific pad, and we only tested one rim of each brand. So results in the field could vary, but we think this test is the first publication that gives a snapshot into the reality of carbon clincher heat transfer and braking performance. And we want to test as many brands as possible, or more rims from the brands that we’ve already included! So if any company feels left out or if anyone wants to be retested, please send them our way (or fly to Sarasota!) and we will do it all again. The more open and all-encompassing this can be, the better it is for the customer. At the end of the day, that’s what it’s all about. Please check out Spark Wheelworks and you can order any of the products we tested in any custom build that you choose!”

We’d add that, if you’ve switched to disc brakes, this test isn’t very applicable as all of the braking forces and heat are at the rotor. The main braking stresses applied to the rim in a disc brake set up are at the spoke holes, and we’re betting your wheel is more likely to lockup and skid the tire before you start ripping spokes out of the rim. It’s also worth pointing out that all of these wheel brands have been on the market for years and have solid reputations and that this test was intentionally a test to failure that, for many riders, may not replicate real world conditions. But it sure was cool.

So who’s Spark Wheelworks? Founded in October 2016 by Ryan Mason and Jonathan McKenzie, they’re a custom wheel building company out of Long Beach, CA, that picks and chooses from premium component brands to build wheels tailored specifically to the rider as well as pre-built wheels from curated parts.




  1. Thermal analysis on specimens taken from the rims should have been performed as well.

    “new they could withstand…” – please fix, editor.

  2. based on my knowledge the weight will be make performance different, more material make rim stronger. but rider loves the wheel are balance in weight and performance. So what’s the point if you compare heat resistance for those wheels are in very different weight?

    • The type of resin matters, not weight. In fact more weight will always be worse, since carbon dissipates heat poorly (thicker traps heat).
      All this is kind of pointless as the only rim brakes will belong to retrogrouch riders and vintage collectors in a couple years. Look at how many MTB still come with rim brakes 🙂

  3. It would be good to get some kind of idea which had a more effective brake track, rather than how long until failure. Yes, that’s important. But, I know I’m unlikely to get to that point. I’d like to know which requires the least force to slow down.

  4. Super test and top marks for spark taking the time and cash to supply it customers not what it think but what it KNOWS ARE THE BEST . Knowledge is power. Still rinding Campagnolo bullet wheel set-4 years on. Carbon rim with alloy breaking surface . 150 extra grams I will save at desert

    • Armed with the results of this test, do you think you *know* which of these wheels delivers the best braking performance? I’ll help you with the answer, which is no.

      All this test shows is that when applying a frictional load in a situation that will never occur in real world use, all carbon rims will heat to a point of failure. What is not offered is how much friction did each brake track generate? How consistent was the pressure? How was their thermal performance compared to an alloy rim? More importantly to the purchaser, how strong and durable are these wheels outside of this test (impact, fatigue, flex)? How fast are they in the wind? How lightweight are they? These are the things that drive a customer to pay $2,000+ for a wheel set.

  5. I started to write a long-winded explanation of why these results are not valid or useful as they’re currently presented but then I remembered this is the internet and nothing matters.

  6. This is very interesting and not too late. A million Europeans who like to climb mountains will stay on rim brakes for a good while longer. Death before defeat on Strava. If they would have included lever force required to achieve the 1200w in the dry and wet it would have been extremely useful as well as merely interesting. Knowing my wheels haven’t delaminated is of little comfort if I’m dead because they wouldn’t stop.

  7. Looked at the protocol again. I should have written that braking force achieved from the 7lb lever pressure would have been interesting. Seems to me that a very slick rim would do best if the lever pressure is constant with this setup. Am I stupid?

    • I haven’t dug in and compared all the data myself, but I noticed some people geeking out over this test on another forum, and one of them pointed out that the Alto wheel spun at a 1mph faster speed than the others, and stayed cooler for the same 1200w input, and 7lbs braking force. That indicates that braking on the Alto rim is not quite as good (if you value the shortest stops possible) as other rims. There was debate about how this affected the test though. Some people thought that since all wheels received 1200w input, they all had to absorb the same heat input, and that the lower temps on the Alto indicated that their resin genuinely can conduct heat away from the brake track better than other rims, but it seems to me that if the wheel is spinning faster that it may allow for more opportunity for air cooling.

      • A faster spinning wheel is not going to be cooled faster than a slower one. It simply means that less off that 1200W is being converted into heat because there is less friction.

        • You sure about that?

          Convective Heat Transfer Coefficient Air
          The convective heat transfer coefficient of air is approximately equal to
          hc = 10.45 – v + 10 v1/2 (2)
          v = the relative speed of the object through the air (m/s)

          Note! – this is an empirical equation and can be used for velocities – v – from 2 to 20 m/s.

      • To bring the geek factor here just a bit, if the wheel was indeed spinning 1mph faster for the same lever force and input, “braking on the Alto rim is not quite as good” translates to LESS FRICTION. Less friction = less heat and if the air cooling from that 1mph of extra speed is just enough it could definitely allow the rim to stay just cool enough for long enough.

        Note the maximum temperatures on all the wheels in the final image. The Boyd and Bontrager wheels topped out at 332 degrees Fahrenheit!

        • One interesting thing that we noticed was regarding the increase in wheel speed throughout the tests! After sustained high temps, the pad will begin to gloss over and you’ll lose a negligible amount of power. This is true on all wheels because it’s a function of the pad and not the rim, but most other rims didn’t reach this point in the test because the rim had already let go.

          You can also see in the phase 2 Alto test that increased braking power (slower wheel speed) will still not raise the temperature of the rim. This is because the max temp that is reached has more to do with the dissipation properties of the rim/resin than anything else. With our new resin tech, it’s not possible to reach a temperature that can damage the rims regardless of brake force, duration, or pad selection.

          This was a super interesting test to run, we learned a lot!

  8. Yeah, watched the entire video now. Alto wheels have extremely impressive failure resistance but spin as fast or faster at 9lb of lever pressure as e.g. Enves at 7lb. Difference may (probably?) increase in the wet. So heavy brake-draggers (a lot of people) should chose Alto. Late breakers something else. Good test! Thank you and kudos!

  9. The test setup „speed of the wheel be a floating variable“ looks like a strange approach to me – off real world conditions, where of course it’s about targeting a certain speed.
    Lower resulting speed at equal energy input means more energy dissipation, so „the total energy into the rim would be constant“ obviously isn’t true. What strikes me is that Alto needs more lever force to slow down to a certain speed. Enve brakes better with 7lb of braking force (ca. 19.6mph @250s) than Alto with 9lbs of braking force (ca. 20.1mph @250s)! Heck, you could build up a fantastically performing rim for this test, outlasting even more energy input and/or lever force, simply through bad braking efficiency…
    I’d wish for a test with variable lever force to slow down the wheel to a target speed at a given energy input. Alto possibly would still be a performer in terms of reliability (probably not so much in required lever force and modulation). But the results would be different in that regard too and way more meaningful and relevant.

    • I do like your test structure of measuring pad force, it’s just a difficult setup to achieve with any sort of reliable accuracy. And I believe the results you are looking for are in this test. First, you certainly shouldn’t compare how fast the rim is spinning at a specific temperature. The two rims dissipate heat completely differently, so the requirement to reach 250 F is very different. By the time the Alto rim hit 250 the pad had been running for so long above 200 F that it had glossed over and the wheel started to speed up. This is a phenomenon that would be present in all tests, as it is a function of the pad, but most rims didn’t reach this point before failure.

      I would recommend noting the brake lever force and wheel speed in the first 3 minutes of each test, as a good comparison of braking power. And keep in mind that if the Alto is running at 19.5mph at 9lb of lever force and the Enve is running at 19.5mph at 7lb of lever force, what does that mean? The average adult has a grip strength of 100lbs, which is what you would pull at if you were panic braking and locking it up. So the difference between 7lbs and 9lbs in order to achieve the same braking power would not even be noticeable by the user. That is the difference between our brake track and their textured track, but we think it is inconsequential. I am also curious to see how much faster the Enve would rotate with their brand pad, which is information we don’t have at this time. What is more important is the fact that the Alto rim seems significantly more durable in heat capacity, which provides a factor of safety that should give our customers the utmost confidence in the product.

      • Thank you for taking time to reply. First of all, I didn’t refer to temperatures – “s” stands for seconds in my book. That was just to specify a rough point in the test procedure to make my point: after 250secs the Enve rim has even better stopping power with 7lbs of brake lever force than Alto after 250secs and 9lbs of brake lever force – whether or not the pads have glossed after 4mins 10secs.
        I know that real world lever forces can be much greater, and 2lbs a hard-to-notice difference, but I don’t see how this invalidates my point: to really have comperable results in terms of reliability, the actual stopping power (i.e. dissipation) has to be the same! So the variable should be brake lever force, not wheel speed.
        This would be a far more conclusive setup to really proof that (possibly) Alto tops the board. And maybe lead to different outcomes among the competitors. I’m sure you’d find a solution for this “variable pad force”/targeted speed setup (if I recall correctly german magazine Tour has done something like that years ago). As I’m currently on DTSwiss ARC’s – it’d be great to see them on that test too.

      • Expensive, yes. Difficult, no. Enve has a $100k machine for doing real tests and never hear of issues with their rim brake wheels blowing up.
        I have see in it, and it’s a brilliant piece of custom machinary that’s very well thought out.

        This test is flawed from the get go when you don’t even use a company’s recommended pads in the test. Everything else is pretty petty as is.

        Always kinda thought Alto was neat and then this “test” happened.

        • I encourage Enve to display actual results to the world so that people who have no toured the factory get get insight into the products they are riding. We will be the first in line to submit a rim for a comparison, and they’re welcome to use any brake pad they want. If the composite structure and resin has specific properties, there is no carbon pad on the planet that can damage the rim.

  10. Pretty darn interesting! I wish they would have had enough rims to test the recommended brake pads for each rim. Would be great to see what sort of difference (if any) that made, It is also cool to see how the different rims look after failure. Makes me more comfy with my 404’s as at least you MIGHT be able to ride that to some sort of controlled stop.

  11. So. If you use a small 23-25c tube on a modern rim, it will fail due to flatting well before the carbon clincher. And it won’t fail catastrophicly.

    If you use your rear brake to scrub speed on sustained descents, your rear tube will fail well before the front.

    Therefore run small tubes and favor the rear brake, and you can learn the capacity of heat dissipation for your rims without melting them.

    Also, this test metric favors rims with less friction on their brake track as less heat would be generated. I’d wager the alto rims lack the bite of the other rims.

    • Yeah as others have pointed out, this is a (deleted) test. You can pretty much ignore everything about this test. Your 303’s are fine. Properly tested by a company that has far more to lose than this brand’s owners, staff, and their familes combined.

      Chances of you overheating any of their competitors rims to the point of failure is slim to none. Alto is also likely using the wrong resins and the assumptions these will have terrible braking performance is very correct.

      Also note 1200 watts is nothing. Their are amateur sprinters that generate that sort of power in a sprint on flat ground. Question is what are the effects of a 230lb rider doing 40-50mph down a 12% grade hill with multiple curves? On a 70 degree day? Repeatedly having to slow down then get back up to speed then slow down?

      As mentioned, Enve has a machine that does just that. Laugh as you may these are dentist/lawyer wheels, but go ride Alto’s and when you crash because they don’t brake as well you’ll be paying the Dentist you hire fo fix your teeth and lawyer you hire to sue Alto the money you saved and then some.

  12. What TheKaiser said. Using a constant brake pressure and constant power input with no external airflow means that a more slippery rim spins faster, which has the double benefit of distributing heat over more of the rim and increasing convective heat loss. The simple way to at least give some account for this would be to report rim speed for each rim, but it would be more meaningful to apply constant power and adjust brake pressure to achieve constant rim speed, and provide airflow at close to that rim speed.

  13. It wouldn’t be scientific but I would find it interesting to have the test run with each brand’s recommended brake pad just for a wheel speed and temperature comparison. Is it simply that each brand tailors the friction of the pad for rim longevity or is there something about the compound that plays nicely with that brand’s resin?

    My big takeaway here is the same as others have commented and comes from looking at the last image displaying the maximum temperatures of each rim. The Alto rim in test 1 maxed out at 263 degrees versus the Bontrager and Boyd rims at 332 degrees Fahrenheit. The Alto rim obviously has a far lower coefficient of friction than Bontrager or Boyd so I’m drawing the conclusion that the Altos will lead to poor braking performance; I’d avoid the Alto rims if braking performance is important.

    However, if you’re riding in a situation where you’ll be on the brakes quite a bit regardless (descending Trail Ridge Road with traffic for instance) this may be an advantage and it is worth noting the Alto rims cool several hundred seconds faster than any rim in the test, even the ones that maxed out at similar temperatures.

    • Just wanted to jump in here to add my 2 cents! The max temp has less to do with braking power and more to do with heat dissipation properties within the carbon. If you look at the phase 2 Alto test you will notice that it is running slower (i.e more braking force and friction, due to the 9 pound weight) but still remains significantly cooler than many of the other rims. You will see improved braking power with our 2018 rims, you simply will not build up as much heat because of the advancements in our resin technology.

      • Bobby …. you can’t possibly expect anyone to put a lot of faith in this test can you? There are so many holes that can be poked and how on earth can you expect anyone to believe that your wheels are EXPONENTIALLY better than theirs? How can brands like some of the ones you tested, with so many years experience building some of the best wheels on the planet, “test” so “poorly”? Are any of them perfect? No way! but come on, this all just reeks of something foul. Now that you’ve publicly called all of these people out it’s time for you to put your money where your mouth is. Send some of your wheels to them to test using their methods … lets see what happens.

        • Absolutely! I think more manufacturers should do testing like this. If they want to buy a rim or find sponsorship from a shop, we’ll send them as many as they want. That would be excellent.

          Please keep in mind that our engineers have worked in this industry, and for many of these large corporations, for a long time. There are a lot of reasons why some of their engineering isn’t as “cutting edge” as you might expect, but I’m not getting into that here. It’s bottom line oriented, and I’ll leave it at that.

          Please reach out to Ryan at Spark Wheel Works regarding the validity of this test. There is zero foul play, and we didn’t even know we’d “win” until the end of the testing. This was done as a study so that we could answer our customers when they ask how our rims compre to X brand in braking performance. We were curious, and we figured others would be as well.

        • Enve and Mavic have far better tests. And it would be scientific to test the rims with the brake pads you are supposed to use on the rims as that is what they were designed for!

          Alto’s novice testing reminds us this is still novices trying to pretend to be better than they really are. (deleted) test passed off as science. Worse yet, marketed as science.

          • Define ‘far better.’ Or quantify it. Or qualify it…

            Also: I don’t use Mavic pads on my Mavics. I can’t afford Enves, but my buddy can, and he uses Zipp pads. Both are brutally expensive, but he scientifically makes it down the hill much faster than I do. Maybe it’s cause his pads are mismatched schitte?

            But really, what I’m most curious about is how you can get that fired up over bicycle wheels!

          • Please understand that the same pad cannot be used if you are trying to compare the composite quality of different rims. An open mold rim from ebay that comes with cork pads may last 10 minutes, and the Enve rim with Enve pads may last 8 minutes. Is the ebay rim of higher quality? Those results wouldn’t mean anything, because there is not way to compare composite structures unless the same pad is used on each test.

            This was not an act of marketing, we are simply trying to shed light on an issue that should have some standard of safety and regulation, but currently does not.

  14. Running Bikecalculator ‘in reverse’ (no idea if this is even relevant but think so) and assuming that a rear wheel could add 600w for a total of 1800w braking gets that a 70kg rider would need a 22% grade to reach 20mph. A 100kg rider would need 15%. Not easy to find those grades sustained for a number of minutes so the results are actually pretty impressive for all. Rim integrity thus is perhaps good enough for all unless you are very heavy and wet braking performance should be the main safety consideration.

    • Keep in mind that no carbon wheel goes from normal structural integrity to complete failure without a significant amount of warpage to the brake track. This is one of the more nuanced points being overlooked in this test which I think will be highlighted in the next few weeks. While most riders might not be too concerned about complete delamination and failure on a typical descent, we feel that their risk of overcooking the brake track to the point of warpage is much more common, and warpage of the brake track compromises the integrity of the brake track. So in our opinion, rim integrity should not only relate to complete failure from delamination but also rim warpage, which typically occurs much earlier on and can go unsuspected.

  15. Lost me when they said they didn’t test the others with the recommended brake pads. C’mon if that’s what they were designed for then test accordingly.

    Also, just because their rim doesn’t blow up doesn’t mean it brakes well…heat is created by friction and friction is required for a rim brake to function properly. It’s a delictae balance.

    Bad engineering at it’s finest.

    I toured Enve years ago. Saw the machine they used to test their rim brake based rims. Lasers pointed at the brake track to accurarely measure heat created by braking. The machine allowed them to plug in GPS data of any given course to simulate the whole ride on the rims. Never hear of enve rims blowing up and they brake well

    Real enginneers use real testing. Nuff said.

    • I’d be happy to clear up these details for you. Regarding brake pad selection, let me pose this question to you: If we ran the Enve rim with Enve pads and it lasted 8 minutes at 19.5mph and 1200 watts, that’s all well and good. Then we ran an open mold rim from eBay, which comes with cork pads, and it lasts 12 minutes at 19.8mph at 1200 watts. Is the eBay rim of higher quality? Do those results mean anything? Without controlling the pad variable, there is no way to accurately test and compare the quality of the composite itself.

      To expand upon that point further: There is no chemical reaction occurring between the pads and the rim, it is purely friction. So an Enve pad could certainly allow that rim to last longer in this test (let’s say 10% longer). However, if an Enve pad was used on every rim then you would see a 10% increased duration on all of them. So the results would remain the same, it would just increase the length of the test. The pad that is used is irrelevant in order to test the rims, as long as they’re all the same.

      I hope this clears up any confusion. Please let me know we can help you with anything else.

      • Ugh. I wrote a long and thoughtful comment, got a phone call, hadn’t pressed send yet, and it disappeared. Here’s the (likely less good) summary:

        First, thanks for trying to do a thoughtful and controlled experiment, but I think there are some severe flaws.

        Designing a well controlled experiment is difficult and sometimes non-obvious

        “if an Enve pad was used on every rim then you would see a 10% increased duration on all of them.”

        Not necessarily true. There’s such a thing as interactions of variables that could make this not be the case. It would be better to use the pad the rim was designed for. The variable is still controlled it’s just x = [pad + rim combo]. Tell me, if you were to test the braking effectiveness of carbon vs. aluminum rims, would you use the same pads for both rims? Obviously, this is a loaded question and meant to be an example to illustrate the rims and pads were not designed in a vacuum (in the case of companies that make/spec their own pads), so the rims shouldn’t be tested as such.

        As for the temperature of the test itself. It seems there are some errors. In the video, the room is stated to be temp controlled. I’m guessing this isn’t perfect, but assuming you’re using an IR thermometer, one source of variation could be the emissivity of the rims. This is a material property that needs to be calibrated to properly measure the temperature with IR.

        The Alto rim’s temperature decreases over time, but its speed does not continue to increase. Energy is conserved. You’re putting 1200W in, and quickly getting to a roughly constant speed out. This 1200W goes into friction in the motor, friction between the roller and the tire, friction in the hub bearing, and the vast majority (I assume??) friction between the pads and the rims. The power that’s dissipated by the brake pads turns into heat. If the temperature is going down with the same input, then the speed must increase so energy can be conserved. Otherwise, where is that 1200W going? Since the speed isn’t increasing (that much?), it could be the case that the pads glazing over the rims are changing the emissivity of the surface and it’s only appearing that the temperature is decreasing.

        There is obviously a buncha speculation on my part because I’m just trying to glean what I can of the experimental setup from the this article and the video. But there’s some screwy stuff going on with the temperature measurement, at least with it jumping around so much in the video. I don’t really understand the cause of that.

        Again, thanks for the attempt and the willingness to have a reasonable conversation in the comments. That said, I think there’s a lot to be improved upon here.

    • Nice ad. I didn’t even have to read it to know what rim was going to come out on top. Bad move for the Alto brand for sure. Sevo, yeah – at least have to be on the same playing field.

  16. Useless real life information… Everything fails that’s just the reality of it. Even Disk brakes… Who in the hell is applying there brakes for 300 seconds at any given time???????? 5 to 10 at most. any more and you should be walking down the hill:)

  17. Did the Zipp rim actually fail or did the tire/tube alone fail and blow up? The zip rim (other than the Alto) is the one that still looked rideable. Boyd, Knight, Roval, not so much

    • It did, but the photo we took while the wheel was on the jig wasn’t a great one. We just didn’t want to change anything, that way we could publish what we saw directly after the test finished. Once we got the tire off it was clear how it had failed. The delamination was actually on the inside of the hook, and it opened up just enough to allow the tire/tube to blow out. The Zipp rim was impressive and the delamination was very minor.

  18. Moronic. Does anyone notice the wheel speed variance between models??? A given positive load, a given negative resistance, among lesser variables, and thus the resultant wheel speed. Assuming consistent inputs, these differences indicate a difference in brake force required (and brake surface drag) to stop or slow any given wheel. One wheel averages 19.5 for the test duration, another approaches 21…do these wheels therefore require the same brake load to slow or stop? Of course not. In the real world we care about how much brake pressure is required to slow to a certain, “safe” speed and, in that construct, is the wheel capable of withstanding the heat resulting from this deceleration.

    Come on guys, while a self-sponsored test full of bias, you can do better (or at least be less transparent in your objective)

    Shall we repeat this exact test tomorrow on wheels with brake tracks composed of a near frictionless surface vs one made of sand paper??

    • Please see my previous comments regarding brake pad glossing and lever pull weight vs human grip strength. You will see that a wheel speed variance of 1mph and pull weight difference of 2lbs becomes negligible when compared with a max grip strength of 100lbs. Also watch the phase 2 Alto test if you want to see the wheel spin more slowly, and compare the heat dissipation from that test with the others. No matter how you look at it, the resin properties speak for themself. Please let me know if you have any other questions!

  19. This test is completely moot. This is like saying we did a test on the entire world’s engines using 0W-20, because thats what our engine runs, and we wanted to be consistent. Doesnt matter if the other engines were completely engineered around different oils.

    • Brake pad selection has been discussed at length. In order to compare composite quality, the test must be set up with the same pad. SwissStop is not “our” pad, it is the highest quality generic pad and is spec’d by half of the wheels in this test.

    • Well, maybe. Or it’s not completely moot. It adds to a growing body of metrics evidence that supports potential solutions to problems with braking which, when added to said body, can help improve performance across a variety of riding situations. Or, if you prefer, “do not block the way of inquiry.”

  20. Interesting test. Nicely executed I’d say. Seems like a cool design (pun intended).
    However, this is just one factor in a good non-disc design. How about impact resistance of these rims? And of course, their weight?
    I guess the question becomes: Is there any additional weight required to achieve these heat transfer properties, and does it potentially come with extra brittleness?

  21. Seems as though all things equil Alto has better heat dissipation resulting from superior resin technology, not hard to believe. Fairly clear to me.

  22. I think the brkae force part is uestionable. Why not just heat up the rim and see when it fails… Anyway osnt carbon rim failure the whole reason why the industry is pushing disc brakes. So big deal if your rim is so great, there is no reason to buy them when we will all be on dics soon anyway.

    Wheels should also be tested with the manufacturers specified pads. It makes sense to test a product as it is intended to be used. This is a big fail for the test if you ask me.

    Also it sounds fishy that all wheels were not tested. Like campy and reynolds. If your buying the wheel wholesale why do you have to tell them your testing it without their brake pads. Or tell them your even testing them at all?

  23. The test is interesting, but what I find most interesting is how many chat room experts seem to have their own ideas and opinions about results that aren’t based on any tests. They just feel the way they feel, and no test–run correctly (in their opinion) or incorrectly will provide enough evidence to correct or change their emotions-based opinion. And they will voice their opinions with no data to back them up strongly, using foul language.

    At least this company ran a test. Other companies might too, but I haven’t seen them transparently post their testing on line… maybe this post will motivate them.

    In other words: Ecce l’America!

    Also: brake pads schmake pads. Who the devil wants to run $50 Zipp pads when $10 buys you cork pads that work just fine? I climb more than you do anyways, and I descend faster too.

    Go kiss your kids, or go for a ride. Who has time for this much vitriol over bicycle wheels? Sad!

  24. I’m just glad those who are trying to make progress in the world, however that may be, aren’t discouraged by the constant barrage of naysayers no matter what advancement is made. I say, flip the script and let’s see what they can accomplish!!!

  25. Your mom says, “Unnecessary anger and unsubstantiated opinions…. wonder why…”

    And then, “I climb more than you do anyways, and I descend faster too.”

    You’re not my mom!

  26. They should test Giant bikes wheels. they claim to be good until about 245 Celsius or 475F according to their marketing material. the industry standard TG rating for wheels is 160C or 320F. according to them anyway. funny not a lot of people think of Giant as a wheel brand unlike specialized with their Roval wheels. but If you look at the numbers not a lot of manufacturers are making 55mm carbon clinchers around 1400 grams. and if they hold up to their temperature ratings like they say you would be hard pressed to find a better wheel.

    • I’d be interested in this I know a few a few giant wheel owners who attest to their “built like brick shithouse-ness” I only ride on carbon for racing and consequently use tubulars. I’d be interested in a test on CC’s brake track fragility I’ve seen so many cracked because the rider got a flat tire in the bunch and had to roll out silightly chipping the brake track.

  27. Well…with all of the comments and explanations above, Bobby is slowly winning me over.

    It would certainly be interesting to run a 2nd test to see how each rim performs with its recommended pads, but his argument that even if Enve lasts 10% with Enve pads, then so would every other rim too if used with Enve pads, so this test is a more pure reflection of heat capacity of the rim.

    He is also winning me over with the pad glazing explanation for the wheel speed creeping up later in the test on the Altos. If there are rims that don’t make it past 3min, then comparing wheel speed at a time point all rims reached is more of a like/like comparison of the braking power of a rim.

    And last, he is winning me over with his focus on how rapidly the Altos cool even compared to rims that reached a similar peak temp. Even if the braking is inferior on the Altos and that prevented them from hitting as high peak temps, they sure do seem to cool off quicker. Similarly the Phase 2 test at similar wheel speed to others, but with more braking force applied supports this.

    I’d like to see some braking force tests like some Euro mags have done, both wet and dry, to establish more clearly how Altos stack up in panic stop situations, as that has traditionally been a weak area for carbon, and is an area where Zipp, Campy, and Mavic seem to be leading the pack now. I’d also like to see some impact tests to see if the Altos better heat transfer comes at the cost of brittleness.

    • Thanks man, you’re spot on with a lot of these points. We spent about 4 months mulling the test procedure to ensure that we left no stone unturned. At the end of the day we felt that controlling the total amount of energy into the rim and displaying wheel speed variance would be the most accurate and easily understood test.

      We would love to do more testing with different protocol, it just comes down to cost. Spark spent a fortune on rims so that we could show this to people, so it’s not something that’s very easy to pull off! But hopefully we can continue to add to this test and share different types of data.

      You bring up a great point regarding brittleness. We had to make sure that we struck a balance in our resin and laminate schedule to ensure that we would be equally as strong as our 2017 rim models. If you visit our facebook page video section, you’ll find a video that shows our UCI Impact Test (of our 2017 rims). We’re UCI certified, but it’s honestly a very easy test to pass (only 40 Joules of energy). Our internal standard to sell a new rim is 90 Joules, which our new 2018 model passes with flying colors.

      Please let me know if you have any other questions at all, and thank you for the constructive feedback!

  28. I work in the trade in the and we see lots of carbon rims failing from no name rims to big brands. For those that have not seen issues Enve we have, so many we wont stock them any more. We dont have long climbs here bit we do have steep rough roads and there is an element of pilot error too. Bottom line-Im very careful about recommending carbon clinchers, to anyone.

  29. Very cool testing. Kudos to Bobby and others who allowed their rims to be tested. As a test engineer, what we tell people is we can testing anything with enough money and time. When you design a test protocol, compromises have to be made. The test looks pretty solid.

    • Thanks Collin! You’re absolutely right. We felt like controlling the amount of energy put into each rim and displaying wheel speed variance would be the most accurate and easily understood test protocol. Spark Wheel Works also spent a fortune to get us all of these rims! If we had unlimited rims and nothing else to do but test them, I bet we could do some pretty cool stuff! Haha. Hopefully this test will be a nice starting point to shed light on the reality of carbon clincher braking performance.

      • Just watched the whole video. Again good work. One question is did you guys monitor the temperature of the drum, or more importantly, did you start it at the same temp each time? If not, the first sample will have an advantage as the increase in tire pressure will take longer to reach its steady state pressure at the increase temps which causes the hoopstress on the rim sidewalls to go up much higher. From what I’ve seen in similar testing (I test wheel bearings for cars) once the rig becomes heat soaked, it will pretty much maintain a constant temp for all remaining samples when you go back to back testing.

        One interesting thing would be if there was a way to monitor the tire pressure throughout the test. For example if you could find a TPM system that threads onto a valve stem. I’m sure you could find a schrader version, use an adapter, then figure out the signal its sending out. It should follow the overall brake track temperature. I remember reading something with zipp saying that they were seeing tire pressures over 200PSI.

        As a side note, one of the things that Enve uses for their test rig are Laser dial indicators which have crazy sampling and accuracy. They are looking at sidewall growth throughout the test. Where this would be useful is determining the start of a failure. It appears that you get some pulsation in breaking prior to failure in many of the samples, which I would assume is where variances in the wall thicknesses is leading to some areas to reach the glass transition temp (potentially point of heat deflection temperature). In reality, this is the point that your wheels are starting to fail and would be deemed unsafe.

        Another interesting addition to the test samples would be a sample from Nextie and Light bicycle rims. Both have advertise some sort of high temp brake surface, although I doubt there is much to it.

        • That’s a good point, Collin. We didn’t have any equipment set up to measure the drum temp (other than a manual laser sensor), but we also weren’t testing any of the rims on the same day. The rims came in bit by bit from Spark, and we had to build most of them up. So it took us at least a month to get through all of the testing, especially since November is a very busy month for us otherwise! We did ensure that the ambient room temp and drum temp were the same at the start of each test, of course.

          I have no idea how the tire pressure would change throughout the test due to temp increases. That’s definitely something I’d be curious about in the future. It must be fairly substantial, because tube choice (particularly size) was much more critical that we expected! The initial prototype tests of our own rims were popping tubes all over the place, either because they were a lower quality brand or they were 18-25c size. That was pretty interesting, because none of us thought it would matter as much as it did.

          You’re right about the exact failure point of each rim! Some of the rims exploded with no warning signs. Other, however, would start to audibly “tick” after just a short amount of time. The Mavic, for example, sounded like it developed a hop very early on in the test but it held out for a while before completely letting go. This is information that we want to share, we just didn’t have a gauge to measure brake track width during the test itself. We also could display ambient sound throughout the whole test video because it’s at 5X speed, and it just sounds like dying mice (and that video is already super long and boring, haha). What we might do is post each individual wheel test on YouTube in its entirety, with real time and without music. That was you will be able to hear exactly when the rim fails, because it’s very obvious when you’re able to listen to it.

          I absolutely want to include more rims, so we will keep those two in mind as we look around! I’m hoping that any confident manufacturers will simply send them to us, or fly to our facility, to be included. If it’s a good product, it would be great publicity for them. We’ll see what happens!

  30. Not scientific AT ALL! This trick is possible because of no brake treatment. Meaning very little braking friction = Horrible braking, horrific collisions & flying off hairpin curves.

    Same factory Alto buys from also makes rims for, Giant, Scope, Mercury, Prime and many, many more… yet the other brands chose to design and pay for the rim brake treatment for safety and performance sake.

    Sad and dangerous marketing stunt.

    • Well, for what it’s worth, Alto rims, while made overseas, are manufactured by of the leading companies for carbon production.

      Moreover, they’re also made using moulds that Alto produces here in Florida, and the new resin and filament winding they discuss, does mean that there really aren’t any other rims out there exactly like these…

      • Yeah, and for that matter, Giant claims far higher tg point for their rims vs. other commonly seen brands, so they claim that the rims can safely reach temps that would cause failure of others. From what I understand, Giant’s focus was making the rims able to handle high temps, whereas Altos focus was to improve heat dissipation throughout the rim, so those seem to be 2 different ways of approaching the problem. My point being that even if those 5 brands all roll out of the same factory, that doesn’t mean they use the same tech.

    • Not scientific? We followed the scientific method to a T, but we are open to suggestions. The energy input into each rim was identical (motor power, caliper force, brake pad, rim kinetic energy), the rim speed was variable and displayed in real time, and the results were shown in their entirety. I don’t know what else we could have done.

      We have always been very open regarding our manufacturing. All of our metal working is done in Sarasota, Florida. This includes hubs, hub internals, and rim molds. The rim molds and the resin are then shipped to Topkey for construction, and rims are shipped back to us for assembly. I worked closely with Topkey while designing frames at Cannondale, and there simply aren’t any facilities in North America that can replicate what their equipment can do. Trust me, I’ve looked! Haha. I’m not sure who else Topkey manufactures for, but it’s likely that they make a large majority of the frame and rims in the industry. There aren’t that many composites factories in the world that can put out quality work, so many of the brands work with the same facilities.

      • You are making your molds in Florida?

        Then how come the new Pacenti rim and Hunt rims have the EXACT same rim shape (and even drawings) as your rims?

        • They literally have different widths, depths, and shapes… You can also check out the videos on our website that show our machine shop in Sarasota where the molds are made. I’m amazed that you could watch this test video and think that we’re buying open mold rims without any design work going into them at all…

  31. At the end of the day, regardless of which carbon wheel you have and how you brake (dragger vs late), braking on carbon rims is an issue you need to consider. taking the brake track off the rim entirely happened with mtb’s over 15 years ago and nobody would ever go back to rim brakes there for obviously reasons. There are zero negatives for disks on road bikes and a whole lot of positives. If you only select you bike components using a gram scale and wind tunnel data, then you are sacrificing performance over other factors that at the end of the day, probably are not actually making you faster.

    • IMO, comparing discs on mtb’s to road is apples to oranges. Mud clearance and the benefits of modulation because of constantly changing degrees of traction aren’t practical concerns on road bikes. Disc brakes do have disadvantages: Cost, complexity, the hassle setting up and increased maintenance for starters. Then there’s that godawful squealing that sometimes happens. If you race forget about neutral support. Disc brakes require beefier axles, spokes and hubs and along with the rotors and calipers is for me aesthetically distracting. Aluminum rims coupled with caliper brakes is simple and elegant and under most conditions is proven to have virtually the same braking distance as discs, and is lighter, less expensive and more aerodynamic to boot. I’m not against technology; I just don’t see the point.

      • Almost none of your negatives are real or an actual concern. About the only one I’ll give you is initial set-up complexity. Thats about it. Aesthetics vary but I bet most younger riders will prefer discs so its not a set in stone thing (like sloping top tubes when they first came out).

  32. Bobby, your patience and demeanor in answering som pretty hostile questions has been fantastic. Not slandering the competition… Rare. I’m a sceptical person but think this is a valid test. You don’t claim to test all relevant parameters, only one very important one. I think you should take away that a lot of people are like me. They have a great fear of not being able to stop. You can’t unsee the vision of flying into a ravine. I am a skinny guy and I can close a 165lb gripper so your point of 7 vs 9lbs of pressure is understood. Even so – saying more friction is not necessarily better is probably just not effective (even if true) from a marketing perspective. However, seeing that e.g. the Enves do speed up as temperature rises maybe more friction just leads to faster glassing of the pads? Maybe this decrease in friction is permanent until pads exchanged and not due to rim temperature as one would perhaps assume. Suggestion: if you ever find resources to do another test. Run for 1 min, stop, 2 min, etc until failure. The last run will then be on glassed pads and perhaps the speed difference between your rims and the competition will be less or even disappear. Perhaps maintaining the integrity of not only the rims, but also the pads is more important than anybody thinks. Good test that made me think, thanks.

    • Thanks man! We went into this publication know that people would have a ton of questions, so we really wanted to put ourselves out there and do our best to explain why we did what we did. The pad glassing issue is one that we didn’t even know would happen when we went into the test. But it does seem to be a function of the pad being a sustained high temperatures. Regardless of the coefficient of friction of the rim, if the pads are over 200 degrees for more than a few minutes, they’ll begin to glass over. It could happy sooner or later on some rims, depending on how quickly it reaches the temperature that initiates this process.

      Your test criteria is very similar to what we do in our internal safety tests. Our “real” test is 1min on, 30sec off, or 20 minutes. This is how we qualify our prototypes. We simply chose to do the “zero to failure” test for the video because it was more straight forward and easier to understand/control. But if we can do more testing in the near future I’d like to use brand specific pads to see how that data compares, and I’d like to do interval testing to see exactly how the cooling curves effects the final results. There are plenty more tests to do, but it’s tough to get other brands to send us free product! Haha. And Spark spent a fortune on rims to allow us to do this. But we’ll keep our eyes open and will do everything we can to provide more data.

  33. Although I understand why, if Campy specifically recommends their pads with their rims, I think that combo should have been allowed as you are testing the system.

    As for the test…
    I’d prefer a constant power/speed. Power would ultimately be modulated via the brakes and one could check power via the motor amp reading. Brake force could be measured as a metric for brake effectiveness, although this is a side trait as the brake power and passing speed at the time would be the same.
    As for the power and speed setting, some assumptions on average grade, weight and drag speed will be needed.

    • Yes, this is absolutely one way to do it. We decided to keep the energy input into each rim the same, because we think it is the most “fair” setup. If you regulate brake force then some rims will receive more energy input than others. I think it’s a great test, but I know that we would have people saying “Yeah, but XYZ rim only failed because there was more brake force and power!” At the end of the day, we can’t satisfy everyone unless we run the test 10 times, haha. Hopefully this will be a good starting point for the industry!

      • See my question below (6:19pm)…how did you standardized energy input?

        I don’t think its a bad test, but I am not sure every rim is getting 1200W

  34. Another interesting idea would be a “repeat” style brake test instead of a drag test. A large flywheel driven to a preset start speed, then allowed to freewheel while brakes are applied to bring the flywheel down to an end speed, upon release, the flywheel is accelerated then brake test repeated until failure. Flywheel mass, start/stop speed, and accel rate assumed for averages (8% grade, 200lbs, start braking at 45 mph, stop at 15mph???)
    Not sure how to accommodate differences in braking performance though (time from start to end speed). It could be done with brake force modulated with a controller tied to flywheel acceleration (decel) but this would be somewhat expensive to set-up and require some specialty feedback controls.

    • Its too different metrics. One could in theory make a rim with amazing braking potential from say 40mph to zero, but have to maintain 30 mph down a steep incline for 2 min and it fails while another rim that brakes poorly has no problem with the drag test.

      I love descending and won’t scrub speed unless for corners IF there are no cars, riders or gravel….and that is the real issue. I have had to drag/pulse brake more than I’d like down considerable descents due to those real world conditions…which is also why I won’t run carbon clincher for non-race riding.

  35. After viewing the video, I have a question.

    How did they establish 1200 W was delivered to the rim? A 1200 W motor will not produce 1200 W just because that is what it is rated at. It’ll produce that (and a little more till its thermal protection cuts power) only if asked. Therefore, a rim will poor braking characteristics is at an advantage as the same W of thermal heat (friction) will not be delivered to the rim and it’ll last longer.

    Unless the test rig was equipped with an amp meter, the test is kind of useless unless someone wants to clarify the above.

  36. Tests like this are welcome – comapring different high end products.
    I didnt read all comments.
    Test will be best as said above:
    -each rim with their recomanded brake pads
    -each wheel powered with motor that have AMP meter – speed must stay same troug all test also force of brake pads- from reading how much current motor have you have data how efficient braking is
    – time to destroy wheel – at which temperature and tire pressure (is damage from heat or too high tire pressure)
    – all rims must have same tire and inner tube

    that is only part of test – for more detailed test you must run them under rain conditions and see how things work there

    OR just test in real world from steep long hill (20% or more) with constant braking to full stop – there will be noticed also if brake pads pulsing, making sound, loosing power after time, modulation…..
    Messauring temperature on braketrack (max temp) – digital or analog with “temperature stickers” (i saw this on one test)

  37. so if i made a rim out off teflon coated carbon would i win the internet? hey it dosent explode but your head might when it hits a tree coz you cant stop! we need to know the mu of the rim/pad of each sample and the volts/amps of the motor during the test. thats where the truth lies

    • In this particular test, it would display the speed as being much higher but the failure mode would be the same, as the total amount of energy into the system would still be 1200 watts. A torque meter, amp meter, and frequency modulator were used to ensure that the motor remained at this power and that each rim was seeing an identical amount of energy.

      • Each wheel might be seeing the same amount of energy but how that energy is dissipated is different. Not all the energy goes into braking which converts the energy into heat. Teflon would make much less energy be dissipated by heat from the brakes so a teflon coated wheel won’t get at hot

        • You are correct, not all energy goes into braking. It also goes into the angular velocity of the rim. If you look at the total energy equation, the velocity terms ensure that the system reaches equilibrium at 1200 W. The amount of energy then converted to heat and dissipated relates to the conductive properties of the resin and composite structure.

  38. How does the rim temp decrease during the test without the speed increasing? If the energy going into the system is constant then the energy must be conserved so has to go somewhere

    • The speed does increase through the test due to pad glassing. After sustained high temperatures the brake pad builds a film that results in decreased coefficient of friction and increased wheel speed. You can see this happening at the end of the Zipp and Enve tests, but none of the other rims reach the point where this phenomenon begins to take effect.

  39. Perhaps it’s a bit silly since they’re not carbon clinchers, but I’d love to see a future test include HED Jet wheels for primarily illustrative purposes.

  40. i dunno, but something tells me that no matter what the tests show that a 270 lb old track sprinter (me) should never ride carbon wheels, disc or rim!!!!

  41. Well it seems that since Alta has proven that they have very duarble construction under the given circumstances, then perhaps they are ready to do a more representative test of real world conditions like some brake to stop tests. that way we can all see if there is a brake distance ‘price’ in exchange for their duarbility. that would certainly put all of our comments to rest.

    • Absolutely, Ed! We do all of these tests internally already, because we would never clear a prototype for production without doing our due diligence! We will be sure to publish results/videos of some of our other tests so that there can be total transparency regarding all of our performance and safety standards.

  42. Sloppy test, if not just the pad selection. There are millions of racing hours in the auto industry to demonstrate that pad pairing with rotor surface is critical. Depending on application, some break pads work better than others on a given surface. For example, formula 1 uses Carbon-Carbon Ceramic and not semi-metallic-carbon ceramic. Even then there are specific formulations within brake pad varieties. Choosing one set of pads, universally, was the first way that eschewed the test… For example, both Zipp and campagnolo used F1 resins, no surprise, campy worked with Ferrari. So the resins they use match the pads they use. Zipp Platinum works well as does Yellow.

COMMENT HERE: (For best results, log in through Wordpress or your social media account. Anonymous/fake email comments may be unapproved or deleted. ALL first-time commenter's posts are held for moderation. Check our Comment Policy for full details.)

This site uses Akismet to reduce spam. Learn how your comment data is processed.