Hondo6

Sorry for the delayed reply here. Got tied up yesterday, then zonked early last night.

It's not really a “sudden drop-off of durability and performance”. It’s more like a slow internal degradation due to material fatigue starting on day 1 of use. That slow degradation eventually manifests itself in a crack or break - which I guess could qualify as a “sudden drop-off in durability” if you want to consider it that. Until that happens, performance and strength remain essentially unaffected.

Or, as I've read elsewhere (paraphrasing): "They're great - until they aren't." (smile)

The passages you quoted from commenters here seemingly implying that CF has an infinite life expectancy indicate they appear to have misunderstood the meaning of fatigue limit and have it exactly backwards - or are perhaps misreading the definition. My guess is that they’re confusing a material’s fatigue limit (some sources instead use the term endurance limit) and a material’s fatigue strength.

The definition of a material's fatigue limit is “the maximum stress a material can undergo an infinite number of times without failure.” Thus, if a given material does not have a fatigue limit then repeated stresses - no matter how large or small - will eventually cause that material to fail. In contrast, the term fatigue strength has a very different definition and meaning; it’s defined as the amount of stress at which a failure can be expected to occur after a specified number of repeated stress cycles. Unfortunately, the terms are similar enough that it would be easy for someone to confuse the two, even though they refer to two different concepts entirely..

https://www.merriam-webster.com/dict...atigue%20limit

https://material-properties.org/what...th-definition/

https://en.wikipedia.org/wiki/Fatigue_limit

https://extrudesign.com/fatigue-limi...durance-limit/

https://fractory.com/material-fatigu...#Fatigue_Limit

The last source above uses endurance limit in place of fatigue limit, and defines fatigue limit slightly differently (e.g., in terms of a specific but very large number of cycles vice an infinite number of cycles). I wish everyone would use those definitions, as I personally think they're a bit more understandable as well as being more precise. But that usage seems to be less common.

Again, for emphasis: any material without a fatigue limit (as the term is generally used) will eventually fail after it has been repeatedly stressed enough times. That's true regardless of the magnitude of those stresses – and pretty much everything you do while riding causes cyclic stresses in a bike frame. Larger stresses will cause the material to fail with fewer repetitions, while smaller stresses will take more repetitions. But a material without a fatigue limit will eventually experience fatigue failure (e.g., it will crack or break). It will just take longer when the repeated stresses are smaller. Exact prediction of precisely how long this will take is not really feasible for a specific item, but it can be predicted statistically from material properties and expected stress levels experienced by the item during use. And a fabrication or material flaw (or a design flaw that concentrates stresses) can grossly shorten the time required, particularly if the flaw happens to be in a high-stress area.

Designers account for this when using materials without a fatigue limit by designing in a large safety factor - generally by using more material and building a structure that is stronger than absolutely necessary, especially if reducing those stresses isn't practicable. In this case, additional material used translates not only into increased strength but also into a longer useful lifespan. Internal stresses (in simple terms, stress is the force experienced divided by the cross-sectional area) are reduced by having more material over which to "spread" the force.

Materials with a fatigue limit (or if you prefer the term, with an endurance limit) behave differently. Absent some type of damage, they appear immune to fatigue failure - provided that all stresses experienced within the structure remain below the material's fatigue limit. (As I noted previously, anything can be damaged if stressed enough.) They thus behave very differently in this regard than materials without a fatigue limit. As a result, a properly designed and constructed structure made from a material with a fatigue limit theoretically has an infinite lifetime provided it is never over-stressed. Material or fabrication flaws, or poor design that concentrates stress in a particular area, can of course reduce the overall force on the structure required to break the structure at that point; and any type of damage (including corrosion) can also obviously change the structure’s lifetime.

AL and CF composites are generally regarded as not having a fatigue limit per the above standard definitions, while in contrast steel and titanium indeed appear to have fatigue limits. (I say "appear to" as it's somewhat problematic to verify definitively that steel and titanium have such a limit via stressing a test sample an infinite number of times at a level below the material’s apparent fatigue limit.) Regardless, in practice all four can be used to make excellent and long-lasting structures - including bike frames. The designer simply has to take the varying properties of each into account and design accordingly.

Bottom line: absent damage or a design/fabrication/materials flaw, a properly-designed AL or CF frame will IMO likely have a longer useful lifetime than its rider, even if it won’t last forever. That’s because I’d I’m reasonably certain that they’re designed to be far stronger than absolutely necessary in order to extend their useful lifetimes to account for the absence of a fatigue limit – if for no other reason than legal liability and manufacturer reputation. Precisely how long will it be safe? Dunno. But I wouldn’t sweat it overmuch.

However, I’d still look over a CF or AL frame or fork carefully on a regular basis – which reminds me, I need to do that to mine this coming weekend. And if I had a thousand-year life expectancy, I might be hesitant to ride an AL or CF frame or fork made today 500 years or so in the future without knowing a helluva lot of details about its design and the quality of its fabrication and materials - unless I could verify that it was NOS. (smile)