Gear
A Dura-Ace 9000 crank can be compatible with a...

Technical FAQ: Bottom brackets, carbon fiber fatigue, and more

Lennard Zinn addresses reader questions about bottom bracket compatibility, the life of carbon forks, and more.

Dura-Ace 9000 with a BB30 bottom bracket

Dear Lennard,
If I have a BB30 and want to use Dura-Ace 9000 cranks, is it better to use, for example, Wheels MFC BB30 to 24mm adapters or an FSA 68mm adapter sleeve and threaded cups?
— Serge

Dear Serge,
Any time you use a sleeve or pressed-in cups, you are increasing the likelihood of having a creaking bottom bracket.

A great solution for this situation would be to use a Wheels Manufacturing BB30 outboard bottom bracket. As it says in the description, “The two machined aluminum cups thread together, ensuring the bottom bracket remains stable in the frame, increasing bearing life, and simplifying maintenance. This means no more creaking, smoother operation, longer bearing life, and really quick bearing replacement.”

Because Shimano’s method of removing axle side play depends on the installer to just snug up the left arm without making it too tight, the bearing life of standard cartridge-bearing bottom brackets depends on the absence of human error during installation. But in the event of such an error resulting in a side load on the bearings, the BB30-OUT bottom bracket with angular contact bearings can take both radial and axial loading and ensure long bearing life, in contrast to the versions with standard ABEC-3 cartridge bearings.

You just tighten these cups into each other and slap in the Shimano cranks; installation couldn’t be much easier.
— Lennard

Fatigue limit of carbon fiber forks

Dear Lennard,
I was reading an article in our local newspaper concerning the finding of a coronial inquest into the death of a local cyclist following the failure of the Al steerer tube on his fork.

This got me thinking about the life of full CF forks. I know that steel has a fatigue limit, whereas aluminum (and its alloys) do not, but do carbon fiber composites have a fatigue limit? In the absence of crashes and/or impact damage, should one consider an all-CF fork to be something that should be replaced at regular intervals?
— Nick

Dear Nick,
No. Carbon fiber composites have no fatigue limit, but that is not really the question to ask, since they do not behave like isotropic, hard (i.e., brittle) metals. Steel (titanium, too) has a fatigue limit, so fatigue failure can be avoided with it. As you say, aluminum has no fatigue limit and can thus be more difficult to predict when it might fail under a given cyclic loading. But being a matrix, carbon is less predictable in its failure modes under fatigue.

If a material has a fatigue limit or “endurance limit,” it means that if the cyclic stresses applied to it are below a certain magnitude (measured in units like pounds per square inch), the structure will not fail due to fatigue. Look at the graph; if the stresses on the steel structure depicted in the graph are below about 28ksi (kilopounds per square inch) – or 28,000psi, the steel will never fail due to fatigue (of course assuming no rusting or impact damage). The fatigue curve for titanium looks similar. But you can see that the fatigue curve for aluminum never flattens out, so even small cyclic loads, if they go on long enough, will eventually cause it to fail.

Carbon fiber composites can indeed fatigue, but they do so very differently from metals. Metals tend to be isotropic (uniform in any direction), even though they can be mildly anisotropic in that they can have a predominant grain direction and, like wood, can be stronger along the grain than across it. (According to the coronial report you sent, this particular aluminum steering tube was not uniform in that it had an “inclusion flaw” in it.) By contrast, carbon fiber composites are extremely anisotropic, depending on how the layers of carbon fabric (usually unidirectional fibers) are laid across each other. And this can vary from fork to fork in the same batch, since the fabric pieces are laid up by hand.

Metals tend to eventually fail under fatigue at a single crack. Carbon fiber composites do not do this — they tend to degrade under fatigue throughout the entire volume of the structure. Composite materials fail due to fatigue in four basic ways: cracking of the matrix (i.e., the resin that holds the fibers together), delamination (peeling apart of one layer of fabric from another), breakage of fibers, and debonding of individual fibers from the resin. You may be able to hear delamination, debonding, and cracking by tapping a coin along the fork; where the matrix is intact, it will make a nice “clack” sound, whereas the sound will be deadened if layers have peeled away from each other or if fibers are cracked or no longer stuck together. Cracking of the fibers and/or of the matrix can be deep within the layers and may not be visible from the surface. Fatigue cracks in metal structures always propagate from the surface (the trick is finding the cracks when they are small).

Long answer to your question, but carbon forks can be subject to fatigue. However, in the absence of crashes or impacts, high-quality carbon forks tend to be highly resistant to fatigue. If your fork goes through a hard crash or receives a hard impact, you ought to replace it right away. In the absence of crashes and/or impact damage, it still makes sense to replace full-carbon forks every now and then. Unless you weigh over 200 pounds, however, you can probably figure a good fork will last a long time (10 years or so?) before it needs replacing. The heavier you are, the shorter the interval between replacements. A sub-130-pound rider can probably assume a full-carbon fork will last them a lifetime unless they go through a crash or suffer an impact.

The issue of course is how a bike is used over the years, since 10 years of use for different riders will be different mileage and riding surfaces and tire pressures. To truly monitor fatigue, you would need have an odometer on it and a way to measure the magnitude of the stresses it receives. I don’t see this appearing on a smartphone app anytime soon.

I have been talking about forks from trusted sources, too. As you can see from this article, if you get a super deal on a brand name, it may be too good to be true, and you could be endangering your life.

When it comes to forks, handlebars and stems, breakage due to fatigue is catastrophic, as it was in this instance, since there is no way to control the bike. When it comes to replacement of bike parts, it’s always advisable to err on the side of caution.
— Lennard

Feedback on a rear derailleur explosion

Dear Lennard,
Greetings from India.

I’m a long-time reader of your column and there are always issues that interest this bike-nerd and ex-racer.

I can’t understand how the cage of the Campagnolo rear mech is fitted upside down. In image “G” (top and lower left) the cage looks absolutely correct. What am I missing?
— Ashok

Dear Ashok,
You’re correct. Thanks. I also hadn’t picked up another thing, namely misrouting of the chain. I’ve updated last week’s column.

Fortunately, in the original post last week I did go on to describe many ways that a rider can unwittingly put a properly assembled derailleur into the spokes.
— Lennard

Dear Lennard,
I’m sorry, but you are incorrect about the inner pulley cage plate being upside down. The lower pulley is shrouded most of the way around and there is a tab behind it, as it is supposed to be. The upper one is mostly exposed, and there is a tab below it, as it should be. That said, you are correct in that there are many reasons a derailleur can go into the spokes.

The real mysteries to me, and I’ve been seeing them with some regularity on particular bikes, is when the derailleur goes into the spokes at a medium speed with the chain on a middle cog and little torque. Even happened to a good mechanic friend. Still can’t explain it.
— Gregory

Dear Gregory,
Thanks. I cannot explain how a rear derailleur goes into the spokes at a medium speed with the chain on a middle cog and little torque, either.
— Lennard