Cracked aluminium downtube, advice needed

[mercalia]

It's interesting that so many similar defects seem to arise near the bottom bracket or around the down-tube/head-tube joint. It's not surprising though, considering the stress diagram that some knowledgeable member put on a thread on the forum some time ago. I wish I could find it now.

From observation I believe that the most common frame material over the last 10-ish years has been alu, probably because it's the cheapest way of achieving lightness. I've never had an alu frame, but from the evidence of the experience of others I don't think I ever would.

However, a few of my cycling colleagues have bikes built around 'Tiffosi alu' frames - either built by themselves or by their LBS, and I've personally not heard of any them having these problems.

Well, I've recently broken my 7 year old steel audax frame - it's the 6th steel frame that I've broken - 3 of those had a design fault (badly positioned braze-on causing a stress point), but the others have just broken after a similar amount of time. I did have one steel frame that lasted over 20 years, but for me, that is the exception rather than the rule.

I've also broken one Alu frame, but that was a design fault too.

The moral for me is that things don't last forever, not when you are 6ft 2 and weigh over 100kg, anyway.

you are not by chance "the Bionic Man", that would explain you experience?

[Vorpal]

...Steel has an indefinite fatigue life, which means it can withstand repeated load cycles without breaking.

Strictly speaking, steel does have a fatigue life. It differs from aluminium in that at low stress values the life is infinite. Above a critical stress value, the life becomes finite and then reduces with higher stresses. The critical stress value varies with grade of steel but is typically around half the failure stress. I don't know for sure, but I imagine that many steel bike frames will experience stresses above the critical value.

Strictly speaking, steel has an indefinite fatigue life within the endurance limit; I presume that you meant endurance limit, since this is typically about half of the ultimate tensile strength.

A steel frame should be designed so that the endurance limit is not normally exceeded in use. I don't expect, honestly, that it would be a problem for a frame to occasionally exceed the endurance limit, so long as it did not exceed the yield strength.

When steel frames fail it is typically because of an extraordinary stress riser (quality problem, rust, modifications outside of the design, etc.), or an extraordinary strain, such as a very heavy rider, extreme riding conditions, carrying heavy loads, or another strain beyone design intent. Obviously frames have to be designed with limitations, and I would suppose that most steel frames are designed for some range of weight / height ratio, plus some additional carried load, depending on the design. So there will always be some overlap between the specification and the usage. Otherwise they would be way over-designed.

That said, I would still consider a long-cycle fatigue failure in a steel bicycle frame to be extremely unusual.

[Ben@Forest]

Well, I've recently broken my 7 year old steel audax frame - it's the 6th steel frame that I've broken - 3 of those had a design fault (badly positioned braze-on causing a stress point), but the others have just broken after a similar amount of time. I did have one steel frame that lasted over 20 years, but for me, that is the exception rather than the rule.

I've also broken one Alu frame, but that was a design fault too.

The moral for me is that things don't last forever, not when you are 6ft 2 and weigh over 100kg, anyway.

Surely that last sentence must be part of the reason? It seems logical that a heavier rider is going to cause more stress on any frame? I read(?) or was told(?) once that though frames/wheels have different tolerances on a 'normal' road bike 22 stone (or 139kg) is about the limit - so the closer you get to that the more likely a frame will fail?

[TrevA]

Not bionic, but I'm a big, powerful rider who occasionally gives it some welly when riding. If a bike frame has a weakness then I will find it.

I have had one steel frame that lasted 20 years, built by a local artisan builder, rather than knocked out on a production line. But I also had a Mercian that cracked after 6 years, a Holdsworth Professional frame that failed after a similar length of time.

I've only had 3 aluminium bikes, my current Trek Domane failed at a weld on the seat tube, but was replaced under warranty.

I've only been this heavy for about 15 years, I was about 80 kg or less for the previous 20 years, but still managed to break frames. So I don't think it's the weight, it is the power that causes it. I can generate 16-1800 watts in a sprint.

[rjb]

, a Holdsworth Professional frame that failed after a similar length of time.

Me too although i am a climber not a sprinter but it failed when the seat tube snapped a couple of inches above the bottom bracket during the gallop at the finish of a road race during the early 1970's. I still managed a 4th place. I assume it failed at the butted joint.

[TrevA]

, a Holdsworth Professional frame that failed after a similar length of time.

Me too although i am a climber not a sprinter but it failed when the seat tube snapped a couple of inches above the bottom bracket during the gallop at the finish of a road race during the early 1970's. I still managed a 4th place. I assume it failed at the butted joint.

Mine was exactly the same and also cracked in the gallop of a road race, I think I finished about 10th. I didn't actually notice until I tried to ride it the next day.

[PBA]

...Steel has an indefinite fatigue life, which means it can withstand repeated load cycles without breaking.

Strictly speaking, steel does have a fatigue life. It differs from aluminium in that at low stress values the life is infinite. Above a critical stress value, the life becomes finite and then reduces with higher stresses. The critical stress value varies with grade of steel but is typically around half the failure stress. I don't know for sure, but I imagine that many steel bike frames will experience stresses above the critical value.

Strictly speaking, steel has an indefinite fatigue life within the endurance limit; I presume that you meant endurance limit, since this is typically about half of the ultimate tensile strength.

A steel frame should be designed so that the endurance limit is not normally exceeded in use. I don't expect, honestly, that it would be a problem for a frame to occasionally exceed the endurance limit, so long as it did not exceed the yield strength.

When steel frames fail it is typically because of an extraordinary stress riser (quality problem, rust, modifications outside of the design, etc.), or an extraordinary strain, such as a very heavy rider, extreme riding conditions, carrying heavy loads, or another strain beyone design intent. Obviously frames have to be designed with limitations, and I would suppose that most steel frames are designed for some range of weight / height ratio, plus some additional carried load, depending on the design. So there will always be some overlap between the specification and the usage. Otherwise they would be way over-designed.

That said, I would still consider a long-cycle fatigue failure in a steel bicycle frame to be extremely unusual.

You may call it whatever you like. I always try to avoid specific technical terms when discussing simple material properties in a non-technical manner. Regardless of the name; I suggest that steel bike frames may well exceed the value at which fatigue effects become measurable. How often, if at all, is a matter for speculation unless you have verified it experimentally. I have no such data but I speculate that a steel framed bike would be (unnecessarily) very heavy if designed to avoid fatigue limits under all reasonable riding conditions.

I would also contend that steel framed bikes would have been unlikely to have been designed to any actual stress limitations. Rather, experience gained from failures would have lead progressively to more efficient frames. For this to occur - there MUST be failures.

[Brucey]

I don't think that anyone is claiming that steel frames operate below the fatigue limit stresses 'under all circumstances' more that excursions into the upper 50% of the operating range are somewhat unusual for most riders and this means that the fatigue damage so caused isn't likely to be that bad for most such riders.

A convenient analogy (which may also correspond well with the actual usage stress pattern ) is to consider how much time on the bike one spends riding anaerobically. For most folk the answer is 'not much' but during this time power outputs of about 1.5kW are likely with a strong rider, and stresses above the fatigue limit are very likely. Contrast this with the likely aerobic power output; it (and the accompanying stresses, more or less) are liable to be about 1/3rd of that.

So arguably a strong rider (who has a frame on the verge of failure when sprinting) is not so very likely to generate stresses enough to cause fatigue damage (in a well-built steel frame) when riding aerobically. The actual fatigue damage to any given frame will vary with the proportion of such high intensity activity, and how well built the frame is for its intended task.

I think this tallies well with my own experience; I've broken a few frames but most of them broke because there was a pre-existing flaw of some kind. Had the frame been built properly (or had otherwise remained flaw -free) the local stresses need not have been anything like so high. By contrast one of my riding buddies had a pedalling style that wasn't in any way smooth; he went about the same speed as me but he broke loads more stuff, presumably because his pedalling style caused higher peak loads at any given power output.

cheers

[Vorpal]

I was writing up a reply, and then Brucey came along and said more or less most of what I was going to say

I will add that bicycle frames are indeed designed to meet particular stress criteria. In fact for carbon fibre, because the material handles stresses directionally, the designers need to have have very good idea how the frame will react to loads. The same tools can be and are used to analyse other frame materials.

What they cannot predict accurately is every single incident or condition that a variety of riders will inflict upon a frame. They can map usage, and a range that they want to design for, etc. And then they can test like that usually in accelerated testing, where the loads exceed by some amount, the normal design. These tests are developed statistically to represent typical 80th or 85th percentile usage.

I don't have test data, but I am aware that the testing occurs, and I know people who do stuff like that. There was a German test standard that included fatigue testing. I don't think all of the tests in the German standard were included in the EN standard, but I not have read it, so I cannot say.

http://www.efbe.de/news/2007_04_Newsletter_english.pdf has some information about it and a nice little picture of a bicycle endurance test from 1912.

[PBA]

I took issue with the statement, "Steel has an indefinite fatigue life, which means it can withstand repeated load cycles without breaking". It is demonstrably untrue and perpetuates the myth that steel bike frames will last for ever.

There is certainly clear evidence that steel frames typically last significantly longer than aluminium. I would be concerned if people asking questions about cracked aluminium frames were given the impression that all their problems would be solved simply by acquiring a steel frame.

That is an interesting article on German bike standards. Note that it describes manufacturers either having to redesign bike parts to achieve code compliance or choosing to comply with a competing standard!

[Ad.ads]

Hi all,
Apologies for the late reply, been busy at work and setting up the new bicycle (a felt f65 with full ultegra components by the by)....I'm trying to post pics as attachment but this site is telling me their too large however they're only 1Mb!

[Ad.ads]

Ok so the pictures are loaded, the cracks are to the underside of the downtube lugs...I'm anticipating it's a write off

[Abradable Chin]

I would still consider a long-cycle fatigue failure in a steel bicycle frame to be extremely unusual.

Does this occur after a LEJOG, or a RTW?
Could your frame die from HCF if you go over the Pennines? Or would it need to be the Alps?

[Ad.ads]

Been trying to catch up on the above comments, thanks again to all.

I believe it's a 90s era bike... Super retro! Shogun team issue aluminium racer bought second or third hand from a rugby player about 6 years ago.
Cable lugs aren't welded, they appear to be bolted, which I presume creates weak spots.
It's lacquered aluminium so can clearly see the tube splits, no need to remove paint.

I'm 6,2" and around 75kg. I mostly blame London roads for the issue.

[Brucey]

Ok so the pictures are loaded, the cracks are to the underside of the downtube lugs...I'm anticipating it's a write off

Well, IIRC that frame is in a 6000 series alloy, so any part that you weld ought to be artificially aged after welding if it is to be reasonably strong. This leads to various possibilities such that other parts of the frame are then overaged, that the finish is destroyed, and so forth.

I think the frame is anodised; you will lose the anodising in and near any weld repair areas.

IIRC there are standard lever bosses (under the adaptors you have) which are riveted/bonded onto the frame. If you did repair the frame these would have to be removed and then (something) refitted afterwards. Given that the cracks appear to have started under the lever bosses (probably where the mounting holes were not deburred properly) a band-on fitting of some kind seems like a better idea.

BTW that frame looks a lot like a Kinesis one to me. I had one like that: It rode OK mainly because of the fork, which appeared to be a straight knock-off of a Vitus fork, with no welds in it, but the frame (with all its dodgy welds/HAZs/drilled and riveted fittings in)...

..that cracked....

So yeah, you could have that repaired, but it probably just ain't worth it; it won't look the same afterwards and it is quite likely to break again anyway, if not at the repair, then someplace else...

cheers