Technical FAQ: Derailleurs, brakes, and frame fatigue

Lennard Zinn takes a variety of questions for this week's tech news column.

Have a question for Lennard? Please email us to be included in Technical FAQ.

Front derailleur adjustment

Dear Lennard,
I keep having to readjust my Ultegra Di2 front derailleur. How do I get it to hold its adjustment?
— George

Dear George,
The limit screws on some Ultegra Di2 front derailleurs from a few years ago tend to unscrew over time. We have had success with just putting Loctite on the threads of both limit screws. It completely made the difference on my personal Ultegra Di2 bike.
― Lennard

Brake lifespan

Dear Lennard,
I have been following with interest your recent conversations about the fatigue life of aluminum and its inevitable failure at some point due to the material’s characteristics. This made me wonder about caliper brakes. In all my experience of riding and repairing bikes over 30 years or more, I have never encountered or heard of a broken caliper arm. My first thoughts were that maybe I’d been lucky, but could this apparent strength be due to them being forged, or is it the amount of material and nature of the stresses put on them that makes them so resilient?

In the past, I had heard stories of CNC’d components failing due to imperfections in the billet they were machined from and the proper way of manufacture was to rough forge the shape and finish with CNC. Obviously, this would not be applicable to frame tubes, so are there different rules for the same material having been treated differently?
— Peter

Dear Peter,
I forwarded your question to Craig Edwards, the designer of the eeBrake and many other bicycle components. Here is his answer:

“Forces that brakes encounter in use are limited by hand strength, lever configuration, cable stretch, and coefficients of friction at the pad/rim as well as the tire/ground. All of these effectively limit brake stresses to be relatively low.

This, combined with the number of times a brake receives the maximum stress the “system” can deliver, is relatively few, conservatively speaking: 10 years use x 1 all out panic stop 3 times per ride (to max. stress) x ride 365 days a year = 11,000 cycles). Sub-maximal braking creates very low stress and has little effect on fatigue. Realistically, I would say if you had one panic stop a week, you would probably find a new sport within a year, so getting to 11,000 maximum stress cycles is hard to do.
— Craig

More on widening gear range

Dear Lennard,
I climb a lot of mountains and ride off road, so I need a bailout gear.  I recently set up this configuration. It shifts well — unlike my old bike!

I would have tried a subcompact crank but worried about front derailleur clearance.

Cervelo C3
— Shimano Ultegra ST-RS685 Disc Brake Dual Control Lever 2×11
— Praxis Zayante Carbon 165mm 48/32 crankset
— Ultegra 11sp Direct Mount FD-R8000-F
— Shimano XT M8000 Shadow RD+ Long Cage 11sp RD-M8000-SGS
— Wolf Tooth Tanpan Shimano 11 INLINE
— Shimano XT M8000 11 speed 11-42
— Norman

Dear Lennard,
We (Wolf Tooth) actually have a RoadLink optimized around road derailleur geometry and more moderately sized cassettes. The RoadLink DM would be Kristian’s best bet for an R8000 with 10-42 (SRAM), 11-40 (XTR), or 11-42 cassettes. A GoatLink wouldn’t improve things nearly as much (the GL10 is better than the GL11, though).
— Marc Basiliere
Lindarets, LLC

More on fatigue life of carbon frames

Here are some more answers from engineers in the bike industry regarding this topic.

From Felt Bicycles:
Like aluminum, carbon fiber will always start decreasing in strength as the number of stress cycles continue. Loss in strength over time is measurable as loss in stiffness. But we can work with this and achieve light bikes that last a long time.

We used to have problems with aluminum frames developing fatigue cracks at the weld joints, but when we went to butted tubes, the frames became lighter and the fatigue problems disappeared. A butted tube didn’t move all of the stress to the weld joints the way stiffer, straight-gauge ones did, and that’s why it improved the fatigue life of aluminum frames.

Similarly, different carbon fibers and orientations will make a big difference in fatigue life. If there is not a “balanced laminate” around the plies — for instance, the carbon plies are too stiff on the outside relative to those inside, stresses will break down the layers that are stiffer first.

If you have the same mold and the same carbon but have an unbalanced layup or voids or imperfections in the laminate, then fatigue life will be reduced. Since carbon does not have standard, always exactly repeatable mechanical characteristics, you can’t make a blanket statement about fatigue life like you can with metal frames. Carbon is not like a metal alloy tube that has standard mechanical properties for that alloy and thus can have a predictable lifetime. This is why you don’t have as much carbon in aerospace applications as you might expect (given its light weight), because you can’t make a 100 percent guarantee of how the tube will perform like you can with a metal tube.

Even though the strength of a carbon frame starts decreasing as soon as it starts going through stress cycles, 100,000 stress cycles on our frames with a layup balance only result in a 1-2 percent loss in stiffness.

A premium frame from a reputable manufacturer will have a balanced layup and will be well-tested, whereas a counterfeit frame won’t. Multiple thin plies is better for fatigue life than the same amount of carbon in fewer, thicker layers.

While it is technically correct that carbon has no fatigue limit, it is practically correct to say that it behaves like a steel or titanium frame with a fatigue limit because the carbon frame is so overbuilt. You can keep running the fatigue test almost indefinitely because you will never approach a stress level in regular use that starts to break it down significantly.

I did a little digging after our conversation and found a nice article that summarizes the same ideas we spoke about.
— Jeff Soucek
Felt Bicycles Director of Research and Development

From Kappius Components and Broken Carbon (repair service):
Here is what I have to say about fatigue on carbon fiber composites. Unfortunately, it’s not as cut and dry as we wish it would be with a composite frame. In general, the fatigue strength of composites is very high, but you are right, they do not exhibit a fatigue limit like steel alloys do. But the counterpoint is that their cycles to failure is so high that ultimately their time to failure exceeds any standard lifetime of a frame on the market. I could easily see a well manufactured and maintained frame last 100 years.

The assumptions in this model are the issue though. There are a huge number of variables on the manufacturing side, including but not limited to the volume fraction of fibers, voids, inclusions, misoriented layup, curing rate mistakes, and poor compaction. Failure could also be associated with bonded-in components that fail at the interface between the composite and insert. Those in combination of potential issues on the consumer side with impact damage, high stress events (crash for example) and UV degradation, the lifetime of carbon fiber composite frames can be significantly reduced.

Ultimately, with proper care and a frame from a reputable manufacturer with a good warranty policy (most have lifetime warranties for the original owner now), a composite frame should last a very long amount of time.
— Brady Kappius
President of Kappius Components
Owner of Broken Carbon repair service

From Specialized:
Generally speaking, carbon/epoxy composites are not practically affected by fatigue testing. Composites, if designed and manufactured properly, should have a near-infinite fatigue life. This is true, assuming that the product is kept in good condition without impact damage. If damage is introduced into a higher-stressed area on a frame, fork, or component, then fatigue can become an issue, depending on the damage and location.

Bottom line, we fatigue test all our composites. Generally, impact and ultimate force testing are higher priorities at Specialized than fatigue testing but when you get cocky, you are likely to be on the wrong side of safety.
— Mark Schroeder
Specialized Bicycle Engineering Director

From CarboFibretec:
To answer your question concerning carbon fiber fatigue.  Compared to metals, this topic is much more complex with composite materials and not yet as well-researched. The fatigue of composites has much more influence than only the type of material and the load. It depends on the materials of fiber and resin, if the fibers are unidirectional, a woven fabric, braided, filament wound … and also the process parameters while curing. Due to its anisotropic properties (different stiffness and strength in different directions) it is also not possible to determine only one fatigue strength.

But in short conclusion, the fatigue strength of carbon fiber is much better than fatigue of aluminum and can also be better than metals like steel, depending on the use case. 
One example: if you have a part with loads just in fiber direction, the fatigue strength will be almost at the level of the static strength.
— Michael Hübner
Head of R&D – Sporting Goods
CarboFibretec GmbH

From a bike testing engineer:
As a design concern, fatigue is less of a factor for carbon. Certainly, everything will break eventually. There are notable exceptions for fatigue in carbon frames, mostly involved with bonding at lugs aluminum joints, etc., which is why they are still tested in fatigue to make sure the attachments are sound and that layups have not pushed the limits of the material or bonding.
— Mark Rhomberg
Test Engineer, formerly for Schwinn, GT, SRAM, and Bike Testing, Inc.

From a former Kestrel Bicycles engineer:
As is usually the case with composites, it is really difficult to generalize because of the virtually unlimited combinations of matrix materials, fibers, fiber orientations, manufacturing processes, and environments.

But the short answer is yes, composites are subject to fatigue.

If you ran a fatigue test of neat resin you would find its strength degrades under cyclic fatigue loading much like aluminum does. Fibers behave in more or less the same way.

What’s interesting is when you create a multi-ply composite laminate. Failure in one layer (either fiber or matrix) will not necessarily propagate through the entire structure since the adjacent ply is likely oriented in a different direction. Of course, aluminum, once a crack develops, will ultimately fail.

Composite materials are also extremely strong — so even though they do not have an “endurance limit,” they are generally operating at a relatively low level of stress. Toray T700 (carbon) is the workhorse of the industry and has a tensile strength of 4900 MPa. Given the log scale of the S-N curve, this can put the failure point out many lifetimes or hundreds of lifetimes.

I can’t say I have ever seen a classic fatigue failure of a composite bicycle frame or component. Rather, I have seen failures that fall into two buckets; impact damage and manufacturing defect.

After being acquired by Schwinn, Kestrel sent a 4000 road bike to be tested on Schwinn’s fatigue rig. After exceeding several lifetimes of cyclic fatigue loading with no failure in sight, they gave up and shut it down.

Another interesting tidbit: there is speculation that one of the main economic drivers in support of the 787 composite fuselage will be reduced maintenance costs. Aircraft go through a periodic heavy maintenance check where the airframe is inspected for corrosion and fatigue and repaired. It is almost a certainty that fatigue cracks will be discovered in an aluminum airframe during this maintenance check. The expectation is that the 787 will show little or no degradation from fatigue loading (and corrosion), which may compel the FAA to extend this very costly maintenance interval.

This is an area of research that could keep PhD candidates busy for decades.
— Kevin Kenney, formerly with Kestrel Bicycles