Wednesday, May 6, 2015

Modern Bike Problems: Campagnolo Ultra-Torque

Pedal, kerChunk, pedal, kerChunk, pedal, kerChunk.

I could feel and hear the "kerChunk" with each revolution -- every time the pedal rounded the 6 o'clock position. It was a little disconcerting, but I knew exactly what was happening. The curse of Ultra-Torque. I've spent lots of time talking about the problems associated with press-fit bottom brackets, I thought today I'd share another bottom bracket issue, but this time, one which even crops up on threaded systems.

The Ultra-Torque crank spindle is comprised of two half-shafts
that are joined together in the middle with a Hirth joint.
Back around 2006 or so, Campagnolo introduced its answer to Shimano's Hollowtech II BB with outboard bearings - which they dubbed the Ultra-Torque system. Comprised of two half-shafts that join together in the middle with a very cool Hirth joint and a single large-diameter bolt, the system consists of some impressive engineering. Hirth joints require complex machining to manufacture, but they are very strong and are used in things like aircraft engine shafts and turbines. Great -- so what's the problem? In those other high-tech applications, the manufacturing tolerances are much better and more controllable than they are in the bicycle industry. In this application, however, Campagnolo can make its cranks to very tight tolerances, but unfortunately they cannot control the tolerances of the bicycle frames into which they will be installed. That in itself is one of the reasons that press-fit BB systems have been so problematic. Even with threaded bottom brackets, there can be certain variables which can affect fit, but most traditional systems have a means to compensate or adjust for those variables effectively.

Here's the non-drive half shaft. It's really a beautifully machined
 piece of equipment, but there may be a problem in the application.
In frames with threaded bottom brackets, the bore of the shell and its threading can be tightly controlled to industry standards. However, the exact width of that shell can vary ever so slightly. A standard British/ISO bottom bracket shell should measure 68 mm wide, or 70 mm for an Italian threaded frame, and the faces of that shell should ideally be perfectly parallel. In reality, the width of those shells could vary by a millimeter or more, or by minute fractions of a millimeter. Even in steel frames with investment cast bottom brackets, there can be a slight variation. And it's not always due to a flaw in the manufacturing necessarily. A BB shell that starts off at the proper spec will be narrowed slightly if a facing tool is applied. But in most cases -- in practice -- that tiny difference in width is not a problem. That's why most traditional threaded bottom brackets have at least one adjustable cup. Phil Wood's excellent cartridge BBs have two adjustable cups, and work perfectly well even if the shell faces are not perfectly parallel. Even with Shimano's Hollowtech system, there is a means for adjusting for those tiny variations in width. But the Ultra-Torque has a rigidly defined shaft width with no effective means for adjustment on frames that might not have such exact tolerances.
Joined together, the two halves of the Hirth joint look like this,
held together in the middle with a large diameter bolt.
So, how does the system accommodate tiny variations in frame shell width? By use of a "wave washer."
To accommodate slight variations in BB shell width, the
Ultra-Torque system uses this "wave washer." It is less than
ideal for the task. (photo from Park Tool)

I suppose that washer is supposed to act almost like a spring to keep the system under proper load laterally. In practice, it doesn't seem to be up to the task. When installed into most frames, even those conforming to Campagnolo's defined specifications (their installation instructions listed an acceptable range from 67.2 mm to 68.8 mm for British/ISO frames), one can physically move the crank laterally in the frame. When pedaling, sometimes this movement can present itself, resulting in a "knock" or that kerChunk that I described earlier. Apparently, Campagnolo's engineers say that this lateral movement is not a problem and that it should not present itself when pedaling. Many rabid Campagnolo fans echo that sentiment, and deny there is any issue. "It's Campagnolo," they'll say -- "how could there possibly be anything wrong?" Some will point to the strength of the Hirth joint, and hail it for its high-tech turbine applications, which is actually irrelevant to this issue. It would be hard to find anybody who is a bigger Campy fan than I am, but realistically, I think there is an engineering problem here.

The issue was first brought to light in 2008 by John Satory, the bicycle mechanic and blogger known as "RogueMechanic." This guy has been on a one-man crusade about the Ultra-Torque issue, has had several conflicts with Campagnolo over it, and has even received hate mail about it.

The thing is, it's quite possible that some people would never notice a problem when riding. If a frame's actual width measures at the wider end of the acceptable specifications, it's possible that the lateral movement would be negated. Maybe some people have such unbelievably smooth pedal strokes that it never presents itself on the road. It's also possible that some people aren't as sensitive to the knocking and therefore don't notice it. Or perhaps some notice it, but don't know to what to attribute it (I've seen where people have attributed the problem to bad bearings, pedals, shoes/cleats, and more).

I installed the Ultra-Torque crank and bottom bracket onto one of my bikes some time back. I was impressed by the engineering, and what seemed to be an unbelievably simple installation (no bearing adjustment whatsoever). After a couple thousand miles, I started to detect that kerChunk. I asked a local mechanic about it, and learned that Campy was aware that some people might have that issue (though apparently, they still consider it a non-issue) and suggested using a second wave washer to solve it. We tried that, and it seemed to settle the problem. I could still produce the lateral movement of the crank when the bike was on the stand, but it did not seem to present itself when riding -- for a while, that is. A couple more thousand miles, and I started to notice it again. Faintly at first. Gradually, it got worse. Pedal, kerChunk, pedal, kerChunk.

The Ultra-Torque shim kit, from RogueMechanic.
Researching it online, I discovered the RogueMechanic site and figured out the reason for what I was sensing. Luckily, Satory had come up with a good, solid solution to the problem. Get rid of the wave washer, and instead use shims to properly space the bearing cups to eliminate the lateral play. He has had shim kits made that he will install for his clients, or which he sells online for home mechanics. The shim kit consists of 2 shims each in 1.0 mm, 0.5 mm, 0.2 mm, and 0.1 mm.  I purchased the shim kit for $45 and watched his video instructions on how to best install them. I'm glad I watched the video, as there is a definite system to determining the best combination of shims that would eliminate any lateral play, but at the same time, did not cause any binding. Installing the shims wasn't difficult, but it was time consuming, and I'll add that having a pair of digital calipers is really helpful. Even with the calipers, there is still a fair amount of "trial and error" to determining the best number and combination of shims. Ultimately, I ended up using about 1.2 mm worth of spacers to eliminate play, which is just a shade (not a technical term) thicker than the pair of wave washers that I removed from the setup. Since the operation, out on the road, the crank feels like it should, with no knocking. No clicking. No kerChunks. We'll see how it holds up over time, but I have to say that RogueMechanic's assessment of the problem seems spot on to me, and I'm convinced the solution is correct as well.
It's a little tough to see, but that's the left-side Ultra-Torque
cup with about 1.2 mm of shims.

It's worth pointing out that press-fit bottom brackets aren't the only ones that present problems that shouldn't be problems, though it's also notable that Shimano's Hollowtech BB system has more adjustment designed into it. But luckily, with the threaded frame and cups, it seems to me that fixing the problem is at least fairly straightforward. I can also say that I'm glad I don't have the press-fit version of the Ultra-Torque. Perhaps one could still make adjustments for width using shims, but I can only imagine that press-fit would only compound the issues. Still, when I think about the fact that I've got a Phil Wood bottom bracket on my Rivendell with about 15 years and I-don't-even-know-how-many miles on it (still feels as slick as butter), could somebody remind me exactly what was wrong with a good old fashioned square-taper bottom bracket?

12 comments:

  1. There are a number of things both right and wrong here.

    When you buy a Campagnolo, SRAM or Shimano product, you are buying a technology that requires a level of care, expertise and possibly tooling that is normally found in a team garage or a reasonably expert shop - these products are made for racing and need to be treated as such.

    In assembly, threads that are aligned / concentric and faced BB shells have always been required - it is simple geometry that placing the bearings further apart and on a bigger annulus, these tolerances need to be more finely controlled.

    Having said that, provided that the BB shell is in tolerance, few threaded BB systems provide a problem. The stated shell width tolerance for Campagnolo is quite wide - plus or minus 0.7 mm, the same tolerance as Shimano and SRAM recommend.

    Provided the BB faces are square and the threads are correctly aligned, the wavy washer works fine in the Campagnolo application within this tolerance - it should be noted that it has an additional role in providing correct pre-load in the bearing.

    The problems usually come if:

    1. The spring clip o the RHS of the BB assembly is missed out - this will allow up to 1.5mm (ish) of lateral float instead of limiting it to less than 0.25 mm - the exact degree being a function of the exact width of the BB shell.

    2. The BB shell is narrower than the -0.7mm tolerance allowed, as the wave washer will not be in compression in this case. The BB axle can still only move 0.25mm laterally as it will prevented from greater movement by the spring clip but without the wave washer being in compression, there is only the friction of the bearings on the inside of the BB cup limiting the propensity to move - hence the noise as the RH bearing comes up against the spring clip or the base of the RH cup.

    3. The rider has a style that side-loads the crank during the pedal stroke by more than the combination of the compression force required on the wave washer (around 10kg) and the friction of the bearing on the inside of the bearing cup can withstand - this is unusual. In this case, Ultra-Torque is not a suitable chainset system but fortunately this occurs only in a tiny percentage of cases.

    It should be noted that UT was designed before frame makers went down the channel of press fit (widely acknowledged to be a poor idea) and although a workable reverse engineering solution is provided, it’s seldom correctly implemented as it requires the near-permanent bonding of the adaptor cups into the frame and many shops / home assemblers either don't RTFM or do so then ignore it.

    On the question of shims - the thinnest shim that RBM makes is 0.1 mm which is a huge distance in bearing assembly terms and the use of his shim technique is not necessary - in our testing, we have seen it cause accelerated bearing wear in some cases and Hirth Joint failure in at least two others.

    Although we work closely with Campagnolo on technical education, service / warranty issues, so might be regarded as biased, we are an independent business working with all manufacturers in the area of technical education for cycle technicians and mechanics. We see every system and are aware of the shortcomings and problems that affect all the systems from all the major manufacturers and also a high percentage of the more niche / boutique brands.

    Through the work that we do with Campagnolo, we see many Campagnolo installations more than even the busiest bike shops - we also see the worst installations as by definition, dealing with warranty, we will always see the "problem cases". In around 95% of such cases, we can either show a shorcoming in the BB shell or the assembly that is causing the problem expereinced by the user. In the remaining 5% of cases (totalling maybe 5 in the UK in the last 7 years, of those bought to our attention), we have only been able to attribute the problem to user pedalling style or some other factor outside of mis-assembly (including tolerance issues) or maintenance.

    ReplyDelete
    Replies
    1. Thanks for taking the time to comment, Graeme.

      Delete
  2. the problem with a properly installed and well maintained square taper BB is that you only buy it once, as it lasts your whole life..

    ReplyDelete
  3. I've built up many bikes over the last few years but mostly Shimano and Sram with just a few Campy UltraTorques. My first experience with the PowerTorque BB being discussed was on a used Colnago. It had been ridden by a young hipster and while not abused, it was well-used.

    I pretty much needed a hammer to disassemble the BB. Rust everywhere, bearings almost seized. Not Campy's fault but I've never seen another BB in such bad shape - even on 20-year-old beaters. Weather protection seemed almost non-existent. I sent the reusable parts off for refurbishing but elected to install a lower-end group as replacement. Partly as they look better (Campy's high-end stuff is fugly, IMHO) and partly just to avoid more issues.

    Overall, I prefer Sram's solution. Shimao's alternately torqued, retightened after 100 miles approach has its problems (like how many people will check when they get a new bike?) and it's not unknown for mechs to forget to tighten the non-drive crank and ruin things. Campy has the fiddly C clip and Shimano has the plastic "spacer/locator thingie" with little obvious purpose.

    And not to start an argument, but these components are used on a wide variety of cycles - although Campy's pricing and distribution means it doesn't show up in the lower price ranges.

    I'm well aware of the RBM's website, and tend to share his opinion. You have to think "If this was the answer - what was the question?" And I must say Campagnolo provides very poor documentation (with component purchase) compared to Shimano and Sram - perhaps this is to maintain the mystery.

    Now maybe we can talk about the shifter clamp bolts made of cheese?

    ReplyDelete
    Replies
    1. Your response mentions Power-Torque, which raises another question for me. Power-Torque was introduced a few years after Ultra-Torque, and looks (at first glance, anyhow) more like the Shimano and SRAM systems. But Ultra-Torque is still being used by Campy on the top-end groups, while Power-Torque is on the mid-range stuff (they don't really have a low-end group). Why have 2 completely different systems? And if Ultra-Torque is really the best thing, why did they come up with a different system? Strange.

      Delete
  4. Having never used Power-Torque or Ultra-Torque, I can't comment on either.

    I've used only one cotterless crank that didn't fit onto a square tapered axle: Shimano's Dura-Ace Octalink. The crank was nice. But I kept on wearing out the bottom bracket because it used smaller bearings in order to allow the extra axle diameter necessary for the system. Also, it wasn't very well-sealed. Interestingly, I got more miles out of the Ultegra Octalink bottom bracket that replaced the Dura-Ace. But neither came close to Phil Wood's reliability or longevity.

    ReplyDelete
  5. So if the BB shell width is narrower than the .7mm tolerance then shims added would be a proper remedy provided that they are installed correctly?

    ReplyDelete
    Replies
    1. I would assume that to be the case. John Satory, aka Rogue Mechanic, could say more definitively.

      Delete
    2. I'm curious about the use of the word 'investment' as above in the sentence:

      "Even in steel frames with investment cast bottom brackets, there can be a slight variation."

      Would appreciate a clarification, certainly the word is known but not the way it's been used here. I've run across this once before, it seems to be ill-suited, at least in a non-academic setting, to use investment this way.

      Thanks
      Steve A

      Delete
    3. "Investment Casting" is simply a method for casting metal - using wax models or patterns to make a mold - then the wax is melted away, leaving a very precise void to be filled with molten metal. The method is also sometimes known as "lost wax casting" and some companies have called it "microfusion casting." For bicycles, it is often used for making steel lugs and bottom bracket shells. The method yields more consistent results - more precise - than some other casting methods, and usually needs less finishing machine work after casting.

      Delete
  6. As a comment to Graeme's post:

    I appreciate the time he has taken to post both his experience and opinion.

    Graeme posted his experience which appears extensive with Campagnolo. I respect that.

    Personally, I have slightly over 16,000km on Campy Super Record Ultra Torque installed on a Colnago C60. My background is an Automotive Engineer for Mercedes Benz, and besides college education Mercedes put me through their training programs, from basics on the shop bench to the design office and failure analysis and reliability teams. It follows that the Ultra Torque design is a simple concept striving towards light weight and low rolling resistance.
    Here is my experience for what it is worth:
    1. The Colnago C60 comes with thread fit, which is precise and UT is pressed into an aluminum fixture with an exact fit. Never the less, measurements were taken prior to installation and all was within specification.

    2.The cups were pressed in using Campy's tool and have provided no issues.

    3. After about 600km, the infamous click started and it was analysed as movement between the ceramic bearing outer race cup and the frame retaining cup. In other words where the bearing outer mates with the frame cup. Thus, lateral movement with the wave washer not able to prevent the movement.

    3. A high viscosity lubricant coated lightly where the bearing outer race mates with the frame cups solves the noise, however it returns around 500km after re-installation.

    4. Removing the wave washer and eliminating the clearance or side play in the BB was done by placing dimensioned stainless washers behind the non-drive side bearing. That means measuring the side play, pulling the bearing, placing a precision machined washer behind the bearing and re-assembly. This is basically the same concept as what the Rogue Mechanic used, however using precision measuring equipment and a machine shop. The result is a crank with no pre-load on the bearings, no side to side play and almost zero friction due to the Ceramic bearings.This solved the noise issue 99%. The C60 has wide diameter thin carbon fiber tubes, meaning any noise is multiplied, kind of similar to a drum beat.

    7. At around 10,000km a dull clunk started developing in the BB. It was not audible to a cyclist riding next to me, however I could hear it and feel it. High torque low RPM. This has progressively gotten worse. At 16,000km the Cult Ceramic bearings show no wear whatsoever. There is still no side clearance in the assembly.

    8. Disassembly and inspection shows that the dull clunk is produced by slight movement between the bearing inner race where it mates on the crank shaft. This movement has progressively caused wear on the shaft and a looser fit on the bearing shaft than what is desired.

    9. Both bearings were pulled and re-assembled on the crank shafts using Loctite 641 with a Loctite 7649 curing agent and given 24 hours to cure.

    10. For the first time the BB is 100% quiet.

    IMHO, I believe Campagnolo Super Record is great for a pro-team with a bunch of mechanics preparing equipment for a stage race and who knows what happens after that?

    I believe the bearing fit tolerances and/or clearance dimensions deserve a bit more attention from Campagnolo, and further, they should recognize that the attempt to limit side to side clearance with a wave washer is not a reliable design solution. My gruppo was number 096 off the production line of the new 2015 series. Perhaps something has changed since, we shall find out?

    I have always been a Campy fan, however IMHO, they are no longer at the top of the list when it comes to quality and reliability. In fairness, I plan to put a few more miles on the setup as it is now, to see how the latest solution works out. Then contact Campy and debate this with them before deciding to abort the brand. Stay tuned.

    ReplyDelete