Measuring material fatigue...

[dd74]

Is there a way - any way at all - to measure life expectancy on certain materials? Say, for example, one has a 25-year-old steel frame: as has been said, every material fatigues. So with that in mind, is there a way to measure a material's remaining strength, which in this concern is steel? And can a similar measurement be applied to carbon fiber, Ti, etc.?

[pitcrew]

My dad's career was in non-destructive testing. Essentially testing materials (metals) for tiny hairline type cracks/stress. That's my discription so it's probably far from the actual definition/truth. He mostly worked with nuclear reactors, etc, testing to i.d. problems before they get larger.

He always mentioned testing one of my frames to see what would show up.

Essentially, a piece of material is tested and a color coded photograph showing various stress areas and cracks would be the end result.....

It wouldn't tell you how much life is left, but I guess it could give you some type of idea of what the frame "looked" like.

[AndrewS]

I think, using the above methods, you could find a crack. But I don't think there is any reliable way of detecting the work hardening that leads to cracking. You'd have to strip the paint, too.

Carbon can be X-rayed to view internal cracking. They do that with composite wings all the time.

[dd74]

Stripping the frame. Yes, and that's the rub. One has to pay money and go through effort only to chance that the frame is bad to begin with. Which is why, I imagine, just buying new is a safer bet.

[AndrewS]

New is always safe, that's why it costs so much.

I wouldn't hesitate to buy a 25 year old frame if the original paint looks good. It takes mileage to where out steel, and that's hard to do while keeping the finish perfect. Plus, tubing wasn't so thin back then.

[OtayBW]

XRD/residual stress measurement - measurement of stress at the crystallite level, way before any cracking occurs - but that's kind of academic overkill for this application, I think...

[Mark McM]

Is there a way - any way at all - to measure life expectancy on certain materials? Say, for example, one has a 25-year-old steel frame: as has been said, every material fatigues. So with that in mind, is there a way to measure a material's remaining strength, which in this concern is steel? And can a similar measurement be applied to carbon fiber, Ti, etc.?

Well, yes and no. You can cut the frame into thin slices and examine them under a microscope for micro-fractures in the crystal structure - but that won't tell you how long the frame will last, it will tell you how long the frame would have lastest if you hadn't slice it up.

Apart from that, there isn't an easy way to tell how much longer a frame will last until the fatigue cracks become visible, at which point the frame has a limited amount of time left.

Some machinery maintenance plans stipulate component replacement on fixed schedules, based on expected life of the component. Other maintenance plans are based on inspection for fatigue cracks (which may include microscopy with or without a dye, electric current induction, etc.), but these tests are usually used to determine that a component needs to be replaced, not how much life it has remaining.

Steel frames that have fatigued almost always have visible cracks before final failure. Many times, there are also noticeable changes in stiffness as well, although this generally doesn't happen until shortly before final failure. I've seen a lot of (metal) frames that have suffered fatigue failure, but I've never seen one that failed with no prior indication of some kind. In the cases when final failure occured while riding, the rider either didn't look for or ignored the symptoms of fatugue cracks. (And even in those cases where the final failure of a metal frame occured while riding, I have seen very few that actually resulted in a crash - usually the remaining intact frame tubes will hold the frame together enough for the rider to stop).

Fatigue in forks is a little different. A broken fork can often result in a crash - while a rider can sometimes control a bike with a single broken fork leg, a broken steerer almost always results in a crash. Steerers are also unfortunately hidden from view, so they can't be easily inspected. I've seen a few broken (metal) steerers, and although the failure appears to happen "suddenly", inspection afterward indicated that fatigue cracks had been present for some time before final failure. Had the steerer been in plain view, the cracks may have been discovered before failure.

[auto_rock]

Is there a way - any way at all - to measure life expectancy on certain materials? Say, for example, one has a 25-year-old steel frame: as has been said, every material fatigues. So with that in mind, is there a way to measure a material's remaining strength, which in this concern is steel? And can a similar measurement be applied to carbon fiber, Ti, etc.?

An important thing here is that not every material fatigues. Steel alloys have an "endurance limit", for a given alloy there is a stress below which the frame will effectively never fail in fatigue. Aluminum and Titanium will, eventually, fail in fatigue but steel is just that much more real (sorry, couldn't resist).

There is no non-destructive way to test for present fatigue cracks inside metal, unless you use in some sort of imagery (usu. x-ray) which will allow you to detect cracks above a certain size, which requires you have the resources to x-ray your bike. This technique is used in routine inspection of some airline parts.

As was said previously, routine inspection is the way to go, steel will generally show signs it has failed (visible cracks)

[dd74]

As was said previously, routine inspection is the way to go, steel will generally show signs it has failed (visible cracks)
Got it. Thanks.

It's very interesting to me how steel still requires some care. More so, I imagine, than Ti.

[Ahneida Ride]

An important thing here is that not every material fatigues. Steel alloys have an "endurance limit", for a given alloy there is a stress below which the frame will effectively never fail in fatigue. Aluminum and Titanium will, eventually, fail in fatigue but steel is just that much more real (sorry, couldn't resist).



Is this true ? I thought Ti was the metal that would last indefinitely if
kept below it's elastic limit.

An Al spring will fail catastrophically.
A steel spring will fail with warning.
A Ti spring has infinite cycles.

Well at least that is what I thought I read somewhere. :confused:

[dd74]

Is this true ? I thought Ti was the metal that would last indefinitely if
kept below it's elastic limit.

An Al spring will fail catastrophically.
A steel spring will fail with warning.
A Ti spring has infinite cycles.

Well at least that is what I thought I read somewhere. :confused:
I've heard the same as well. I've also heard Ti can be repaired of damaged. Much like steel.

[Louis]

I'm pretty sure that like steels, Ti alloys do have endurance limits.

[AndrewS]

I'm more in line with Ms. Ride. However, the truth is always more complicated.

If you start out with perfect welds, a ti bike will probably never wear out, no matter how many cycles. Steel will eventually work harden and crack, but only after a lot of heavy cycles.

Aluminum numbers always always looked bad, but the truth is that aluminum frames are, by necessity, so stiff that they barely fatigue at all, because most riders can't flex them far enough to matter. I've seen some fatigue tests that bear this out. Of course, something that rides like it's made of armor plate isn't worth putting that much wear on in the first place.

I think carbon fiber can take a tremendous amount of flex - the question is if the epoxy substrate can take as many. I certainly put a lot of trust in composites when I pole-vaulted - but I did break a pole; painfully.

[Brian Smith]

Is this true ? I thought Ti was the metal that would last indefinitely if
kept below it's elastic limit.

An Al spring will fail catastrophically.
A steel spring will fail with warning.
A Ti spring has infinite cycles.

Well at least that is what I thought I read somewhere. :confused:

Almost.
Steel also has an elastic limit, as with ti, and ti will also fatigue, like steel, when the elastic limit is exceeded. Of those 3, only aluminium and its alloys have no elastic limit.
Ti failures, in most products discussed here, actually progress more quickly, in my experience, than those in steel, once initiated. The repairs are also more difficult/problematic.
Probably the rhetoric about titanium frames being "lifetime" frames gave rise to the assumption that ti has no elastic limit, but whatever the source, it isn't true. Ti simply corrodes less, so perhaps those who equate corrosion with failure birthed the "lifetime" frame notion.

[cadence90]

This doesn't answer the OP in any way, but...

...almost 20 years old and still an excellent read.

"Metallurgy for Cyclists, by Scot Nicol" (http://www.ibiscycles.com/support/technical_articles/metallurgy_for_cyclists/)

[dd74]

I think carbon fiber can take a tremendous amount of flex - the question is if the epoxy substrate can take as many. I certainly put a lot of trust in composites when I pole-vaulted - but I did break a pole; painfully.
What do you think of multi-material frames? By that, I mean steel or Ti with carbon stays and forks.

I know several people who refuse to use anything but one material with (at least) steel bikes. That means steel forks, steel stays. The fear is a breakdown of epoxies, deeply scratched or gouged surfaces, compromising the fibers, etc.

It's for the same reason, I imagine, that these same people don't ride carbon frames.

I've never seen, but have heard of Ti forks. I also hear they flex like crazy in comparison to c/f forks.

[Mark McM]

Hmm… So many misuses of technical terms in this thread. Hopefully this will help clear things up …

Is this true ? I thought Ti was the metal that would last indefinitely if
kept below it's elastic limit.

An Al spring will fail catastrophically.
A steel spring will fail with warning.
A Ti spring has infinite cycles.

Firstly, all metals are subject to fatigue – if you load them high enough and often enough, they’ll break. The mechanism of fatigue involves the initiation and growth of cracks. While aluminum may not exhibit the endurance limit of steel and Ti, it does not fail with no warning – cracks form and grow over time just like any other fatigued metal.

Contrary to marketing claims, titanium is not immune to fatigue. In fact, it has a an endurance limit similar to steel’s – about 40- 50% of its ultimate strength.

By the way, here is a good page that describes metal fatigue (http://www.epi-eng.com/mechanical_engineering_basics/fatigue_in_metals.htm).



If you start out with perfect welds, a ti bike will probably never wear out, no matter how many cycles. Steel will eventually work harden and crack, but only after a lot of heavy cycles.

There is no reason that a steel frame will necessarily crack, and no guarantee that steel frame won’t. Also, it is bit misleading to link work hardening with fatigue cracks. It is true that there is some work hardening happening on a microscopic level at the root of a propagating crack; but on the other hand, there are some work hardening operations like shot peening that can actually improve resistance to cracking. (Work hardening is the re-alignment of the crystals in a metal by plastic deformation, such as forging or rolling, and primarily acts to raise the yield point - it does not increase ultimate strength or stiffness. Most metals can be work hardened, including those that can hardened by heat treatment).


Almost.
Steel also has an elastic limit, as with ti, and ti will also fatigue, like steel, when the elastic limit is exceeded. Of those 3, only aluminium and its alloys have no elastic limit.

I think you are mixing up the terms “elastic limit” and “endurance limit”. The elastic limit is the stress beyond which the metal will be permanently deformed; or in other words, below the elastic limit the metal will return to its original shape when unloaded (the metal will behave elastically).

The endurance limit is the magnitude of cyclic stress below which a metal will tend to last indefinitely (which is not to say that the metal will never fatigue below that stress, but the number of cycles to failure increases so greatly that it will likely be taken out of service for some other reason first).

[Brian Smith]

I think you are mixing up the terms “elastic limit” and “endurance limit”. The elastic limit is the stress beyond which the metal will be permanently deformed; or in other words, below the elastic limit the metal will return to its original shape when unloaded (the metal will behave elastically).

The endurance limit is the magnitude of cyclic stress below which a metal will tend to last indefinitely (which is not to say that the metal will never fatigue below that stress, but the number of cycles to failure increases so greatly that it will likely be taken out of service for some other reason first).

Yes, that's correct. I re-used the term "elastic limit" from earlier in the thread rather than the proper term - "endurance limit."
Thanks for adding to the discussion.

[Lifelover]

Assuming a **** load of "normals"

The stress a frame will see will be under the endurance limit and effectively last forever.

[AndrewS]

I think that this bit from Spectrum's website is interesting. These guys make steel bikes for a living, and design ti frames that Merlin welds for them:
TITANIUM, STEEL, ALUMINUM, CARBON
There are basically four bicycle materials on the market today. Steel has been around forever and has a unique and classic feel. Similarly, aluminum bicycles have been tried and tested and are renowned for their stiffness. Newer to the game is Titanium and its here to stay. Finally there is Carbon Fiber. This stuff can be light and stiff at the same time. So how can you choose which one is right for you? Perhaps this will help.

Steel
Steel, the original racers choice offers a predictable ride. In general, a well made steel frame can be responsive and stiff. A well made steel frame can last years but over time it will fatigue and rust. When it comes to a custom bicycle, steel enables experienced builders like Jeff Duser and myself at Spectrum, Richard Sachs and others to really dial in the ride characteristics of the frame. We also think lugged steel frame are beautiful.

Aluminum
Aluminum offers a very stiff (often harsh) ride. Lighter than a typical steel frame, aluminum offers quick handling and excellent transmission of your energy to the wheels. It does fatigue over time and is far to jarring a ride for many riders. It's relatively inexpensive and thus popular.

Titanium
Titanium happens to be one of my favorite materials to ride. Its light, won't rust or fatigue, and offers an amazing combination of stiffness and comfort. A well-built Custom Titanium bicycle will last a lifetime. Ride it and you will know what I mean.

I think there is sometimes a disconect between the engineering books and the application of principles to real world usages (usually due to oversimplification). Tom Kellogg seems to believe that a steel bike (his own!), ridden long enough, will wear out. He has been riding ti bicycles for 22 years, and thinks they won't wear out.

I have seen ti frames crack in HAV zones, but nothing that ever looked like the metal was generally fatigued. On the other hand, I met a guy who was on his third high end steel bike - they always cracked in the center of the right chainstay due to his pedal stroke and high mileage. Both are relatively unusual occurances, but I think Tom is saying that the second example is not entirely surprising, but would be on a Ti frame. The exact metalurgical reason for that is hard to nail down, but 3/2.5 ti seems to be a material that is very comfortable with the amount and type of flex that pedalling and road shock impart, and can take that kind of work indefinitely if there is no imbrittlement from the welds.

[chromopromo]

"Apart from that, there isn't an easy way to tell how much longer a frame will last until the fatigue cracks become visible, at which point the frame has a limited amount of time left."

Do rusted frames fail in the same manner i.e. will you see fatigue cracks. I have a bad habit of buying old steel frames, some have had rust that has pitted the tubes. I have put many mile some rusted frames with no problems but I would like to know if I am riding on borrowed time.

[Peter B]

<snip> but 3/2.5 ti seems to be a material that is very comfortable with the amount and type of flex that pedalling and road shock impart, and can take that kind of work indefinitely if there is no imbrittlement from the welds.


This seems like an oversimplification. What about tube diameter and wall thickness?

[AndrewS]

What do you think of multi-material frames? By that, I mean steel or Ti with carbon stays and forks.

I know several people who refuse to use anything but one material with (at least) steel bikes. That means steel forks, steel stays. The fear is a breakdown of epoxies, deeply scratched or gouged surfaces, compromising the fibers, etc.

It's for the same reason, I imagine, that these same people don't ride carbon frames.

I've never seen, but have heard of Ti forks. I also hear they flex like crazy in comparison to c/f forks.
Sorry I missed this. I think that a ti frame with carbon tubes is a ti frame missing some of the best parts. There's so little wrong with the ride or stiffness of ti to replace it with a material that lacks some of its toughness AND add some dissimilar material joints in the process. I don't think you get a better bike than you would out of a 100% carbon, ti or steel bike, but you've introduced more to go wrong. They're usually heavier, too. I thnk the liveliness of a material comes from how it acts from joint, across the middle, and back to the next joint. Interruptions aren't going to improve matters.

While epoxy works on lots of different materials, only composites (stuff made of fiber in resin substrate) really love to stick to it. Metals are best joined in metalic ways, epoxy based products are best joined via epoxies. So I think gluing metal is only a little better than welding plastic. Epoxy on epoxy is a lot like welding metal - you end up with one continuous piece of reinforced resin.

I had thought about designing a lug system that would allow you to mix and match frame tubes - ti, steel, aluminum and carbon - using aluminum lugs. Ultimately, I decided it would be a waste - you'd never get anything but novelty from such a set up. The lugs always use up whatever weight savings you attempted to get by multimaterial, and the joins could never be as good as weld, braze, or resin on resin. Raleigh used to have a frame with aluminum lugs and straight, thin guage 531 tubing. It was kinda light, but not near as light as many tig'd all steel frames are now.

The ultimate "multimaterial" frame, I think, would be to use a polymer to reinforce a ridiculously thin section of a butted steel tube against denting and tearing. As long as the interior coating doesn't ruin the ride, it could allow unprecedented wall thicknesses that aren't fragile. Maybe boron?

As far as forks go, I don't think anyone has ever seriously applied themselves to the task of designing a proper Ti road fork. To have the stiffness needed at a low enough weight, a ti fork would need fairly oversized blades. And if highly shaped, tapered tubing isn't readily available, you end up with an ugly monster of a fork. Done correctly it would look like one of those super-aero forks, but on steriods with a long front-to-back crown. Had molded carbon processes, like Kestrel's, not debuted at exactly the same time as the first Ti bikes, we might have seen some fork development, but now it's not going to happen. Hydroforming might be a solution, but no one is interested (I would be, though).

[AndrewS]

This seems like an oversimplification. What about tube diameter and wall thickness?
Well, I was speaking of the kind of diameters and wall thicknesses you'd want to use. And there really are just so many. Build the tubes too wide and the ride gets too harsh. Build the walls too thin and they dent easily. Either way, you need a certain amount of material at the welds, and the ends of the tubes is where the biggest loads are so butting delivers both enough material to join, disipate heat and take the load.

With 3/2.5 butted tubing, respected companies like Serotta and Seven make frames as heavy as 3.4 pounds and as light as 2.2, and none of them are breaking. The weight spread represents stiffness, not durability. So a heavy bike is probably both oversized to be stiff and has conservatively thin center sections, while the light ones are narrow AND close to the limit thin. The rest have a different balance, which is why the majority of butted ti frames fall between 2.7 to 3.1.

Ti is something of a fortunate material because its density and strength dictate wall thicknesses that are quite durable while still being very light.

[Louis]

I would actually say this for Al, Ti and CF frames before I said it for an Fe frame.

Assuming a **** load of "normals"

The stress a frame will see will be under the endurance limit and effectively last forever.

[Mark McM]

Titanium
Titanium happens to be one of my favorite materials to ride. Its light, won't rust or fatigue, and offers an amazing combination of stiffness and comfort. A well-built Custom Titanium bicycle will last a lifetime. Ride it and you will know what I mean.


I think there is sometimes a disconect between the engineering books and the application of principles to real world usages (usually due to oversimplification). Tom Kellogg seems to believe that a steel bike (his own!), ridden long enough, will wear out. He has been riding ti bicycles for 22 years, and thinks they won't wear out.

I think this Tom Kellogg quote can be scored up to marketing puffery, and not some magical quality of titanium. I’ve seen enough fatigued and cracked titanium frames (including several Kellogg designed Merlins) to know better than to believe that.

I have seen ti frames crack in HAV zones, but nothing that ever looked like the metal was generally fatigued. On the other hand, I met a guy who was on his third high end steel bike - they always cracked in the center of the right chainstay due to his pedal stroke and high mileage. Both are relatively unusual occurances, but I think Tom is saying that the second example is not entirely surprising, but would be on a Ti frame. The exact metalurgical reason for that is hard to nail down, but 3/2.5 ti seems to be a material that is very comfortable with the amount and type of flex that pedalling and road shock impart, and can take that kind of work indefinitely if there is no imbrittlement from the welds.

I’m not sure why you draw a distinction about cracks near the Heat Affected Zones and what you mean by a metal being “generally fatigued”. The metal in the HAZ didn’t fail during the loading cycle, and took many load cycles (all below the yield point) to form cracks, which means that it was a fatigue phenomenon. Also keep in mind that the highest stress concentrations occur at or near the joints, so fatigue is likely to occur first at or near the joints (regardless of the quality of the joinery). Fatigue cracks in steel and aluminum frames also primarily occur at or near the joints - if these were discluded, then it might seem that steel and aluminum frames could last indefinitely as well.

With 3/2.5 butted tubing, respected companies like Serotta and Seven make frames as heavy as 3.4 pounds and as light as 2.2, and none of them are breaking. The weight spread represents stiffness, not durability. So a heavy bike is probably both oversized to be stiff and has conservatively thin center sections, while the light ones are narrow AND close to the limit thin. The rest have a different balance, which is why the majority of butted ti frames fall between 2.7 to 3.1.

Having seen a fatigue cracked Serotta Ti frame or two, plus many fatigue cracked titanium frames from other makers (most using 3/2.5 tubes), I know that there is nothing magical about titanium that keeps it from suffering from the types of failures that other metals are prone to, including fatigue. Unfortunately, frame manufacturers don’t release failure data on their products so we can’t know with certainty whether lighter frames are more prone to fatigue than heavier frames, but from personal experience I have seen enough cracked frames (of all materials) to know that lighter frames tend to be more prone to fatigue than heavier frames, and I haven’t seen anything to indicate that this isn’t true for titanium as well.

[dimsy]

there is no definitive way to guage how long a frame of a given material will last.

how long its' ridden, how far and how hard are unpredictable factors that will play into when, if at all it will ultimately fail.

that being said. for a straight gauge section of tubing, cycles to failure is predictable both experimentally and mathematically.

in addition, i've only ever seen one ti frame crack and it was near a weld at the bottom bracket cluster.

[AndrewS]

Mark,

Why do you think TK would engage in "puffery" to undercut his own steel products? He seems to believe more in the ti made for him than the products of his own hands. It is an odd strategy.

As you acknowledge, all we have are our anecdotes and impressions. I tried to explain mine with the example of the guy who kept breaking chainstays well away from the joint. I have seen a cracked Serotta ti as well. I still end up with a strong impression, from both engineers and bicycle makers, that titanium wears less from riding and that the majority of failures were unrelated to mileage and power. That's the best I can do.

[Louis]

Simple way to avoid all these issues:

1) Buy a frame that strikes your fancy.

2) Ride it. If you like it keep riding it. Chances that it will fail in a manner that will harm you are next to nill.

3) If you don't like it don't ride it, go back to (1).

Louis

[Brian Smith]

While epoxy works on lots of different materials, only composites (stuff made of fiber in resin substrate) really love to stick to it. Metals are best joined in metalic ways, epoxy based products are best joined via epoxies. So I think gluing metal is only a little better than welding plastic. Epoxy on epoxy is a lot like welding metal - you end up with one continuous piece of reinforced resin.


Andrew -

Just a couple of data points here for you. In the years of Serotta bonding carbon fiber composites to metal, I have formed no impression of any process or material related reluctance of metals and carbon fiber composites to be joined with epoxy. I think there's absolutely no distinction on the carbon fiber's part about what it was being bonded to. I would love a simplistic perspective on metal bikes being metal and carbon fiber bikes being carbon fiber only, but really there is no supporting data, in my experience. I also think that envisioning joining carbon fiber parts together with epoxy necessarily imparts equal structural continuity as a single carbon part is flawed simplicity. There may be other reasons to not prefer a bike with carbon tubes bonded to metal parts, and I might even agree with you there, but they certainly can't be said to stem from a necessary sacrifice to the bond being anything less than absolutely suitable and reliable.

Titanium makes a great bike frame material, but it also requires careful engineering, specification, and process to yield desired performance. I wouldn't take what's on Spectrum's website as the complete embodiment of everything that Tom and the entire frame manufacturing world's knowledge about titanium frames. They, too, suffer fatigue failures, and not at a rate that proves the material to be decidedly superior to other materials.

[AndrewS]

Hi Brian,

My own impressions of what happens when ductile and imporous metals a bonded comes from my time as a pilot, and watching aluminum composite decking delaminate and rotor blade pockets detach. Certainly, not the same thing as a bike.

Years later when I started working with micarta in my knife making shop I was amazed at its tendency to stick to itself so well.

Tom's quote is certainly not offered as the end all, be all. But he's been around, and makes a rather bold comparison. It doesn't seem like puffery is his style, and I've never seen any other manufacturer so boldly state the life limitations on their product. I don't know what to make of his statement, otherwise.

I realize that all the different ways of making bikes are legitimate, safe, and ride well if designed well. DD74 asked me my opinion, and I posted it. It isn't scientific, just my feel for the materials and biases based on my experiences. Consider an "IMO" right at the beginning.


If Serotta has kept track of failures and likely causes, I'm sure everyone here would enjoy hearing about it!

[Lifelover]

I would actually say this for Al, Ti and CF frames before I said it for an Fe frame.

and fixed it
Assuming a **** load of "normals"

The stress any frame will see will be under the endurance limit and effectively last forever.

There is a lot of over thinking going on in this thread.

[firerescuefin]

Maybe....but I do find the engineering Wonks breaking it down interesting. I certainly would have nothing to add to the conversation, but am glad for those with the background giving their input.

[Louis]

and fixed it

There is a lot of over thinking going on in this thread.

And I agree with both of these statements. :beer:

[auto_rock]

Simple way to avoid all these issues:

1) Buy a frame that strikes your fancy.

2) Ride it. If you like it keep riding it. Chances that it will fail in a manner that will harm you are next to nill.

3) If you don't like it don't ride it, go back to (1).

Louis

+1, and I am a chronic material science over-thinker.

[Mark McM]

Mark,

Why do you think TK would engage in "puffery" to undercut his own steel products? He seems to believe more in the ti made for him than the products of his own hands. It is an odd strategy.

On the contrary . It is entirely very common strategy for an advertiser to claim some advantage (real or exaggerated) for a more expensive product. Kellogg wasn’t the first and he won’t be last. Besides, it is plainly untrue to claim that titanium can’t fatigue – if not puffery, what would you call it?

I still end up with a strong impression, from both engineers and bicycle makers, that titanium wears less from riding and that the majority of failures were unrelated to mileage and power. That's the best I can do.

I’m not sure what you are trying to say here - are you saying that if these frames were never ridden, they would still have cracked, while just hanging on the wall? I think you still don’t understand metal fatigue.

[dd74]

There is a lot of over thinking going on in this thread.
Really? Explain that to a guy who might or already has plopped down over $3K on a custom frame.

[AndrewS]

On the contrary . It is entirely very common strategy for an advertiser to claim some advantage (real or exaggerated) for a more expensive product. Kellogg wasn’t the first and he won’t be last. Besides, it is plainly untrue to claim that titanium can’t fatigue – if not puffery, what would you call it?Oversimplification? He could have used softer language about the steel, and chose not to. That's all.


I’m not sure what you are trying to say here - are you saying that if these frames were never ridden, they would still have cracked, while just hanging on the wall? I think you still don’t understand metal fatigue.I understand metal fatigue. I understand it enough to know that just because something is made of ti or steel, that doesn't tell you anything about its resistance to fatigue. Ti tubes come hard, and some claim cold working increases that hardening. Complete annealing takes place at 1292 F, and the weld was close to 3000 F. It is not hard to imagine a weld that either got hot enough to anneal too much of the surrounding thinner metal, or an area that was not clean or shielded enough and got oxygen embrittlement. Those areas no longer have the full properties of hard 3/2.5 and would fatigue much faster.

As I said earlier, I have always seen ti failures near the joints only, and while I have seen the same for steel - I've also seen steel fatigue mid tube. That is where I get my incredibly unscientific (and probably unnecessary) conclusion that a good ti bike may never fatigue, IF every joint was perfect executed and the its built for the rider's weight.

[Lifelover]

Really? Explain that to a guy who might or already has plopped down over $3K on a custom frame.


There is a lot of over paying going on as well!

[Charles M]

Oversimplification? He could have used softer language about the steel, and chose not to. That's all.


I understand metal fatigue. I understand it enough to know that just because something is made of ti or steel, that doesn't tell you anything about its resistance to fatigue. Ti tubes come hard, and some claim cold working increases that hardening. Complete annealing takes place at 1292 F, and the weld was close to 3000 F. It is not hard to imagine a weld that either got hot enough to anneal too much of the surrounding thinner metal, or an area that was not clean or shielded enough and got oxygen embrittlement. Those areas no longer have the full properties of hard 3/2.5 and would fatigue much faster.

As I said earlier, I have always seen ti failures near the joints only, and while I have seen the same for steel - I've also seen steel fatigue mid tube. That is where I get my incredibly unscientific (and probably unnecessary) conclusion that a good ti bike may never fatigue, IF every joint was perfect executed and the its built for the rider's weight.



That's really well said...

I've seen good Ti and steel builders toss things they find unacceptable. And I have seen the same mistakes painted over and sold.

[pbjbike]

I think this Tom Kellogg quote can be scored up to marketing puffery, and not some magical quality of titanium. I’ve seen enough fatigued and cracked titanium frames (including several Kellogg designed Merlins) to know better than to believe that.


Several? Where did these failed frames show up? Which tube/s showed a problem? Years of manufacture? I'm sincerely interested.

As a custom builder at Merlin, I also handled frame repair for two years. I don't think I saw more than 12 broken frames, (from a fabrication problem, NOT bike-meets-garage-on-the-roof rack issue, or tubing failure), during that time. Almost all had purge issues: inadequate purge hole size between the chainstay and bridge, from early batches--Right chainstay crack, near the BB.The issue was addressed immediately when a small pattern emerged. All were straight guage 3/2.5.

There were two or three with cracks near the DT shifter bosses caused by inadequate purge between the boss and the exterior of the down tube, (this was an achilles heel for many Ti builders. The advent of STI/Ergo and cable stops welded to the head tube did away with the need for shifter bosses). Double butting started towards the end of my stint there, so I'm not sure if tubing fatigue was an issue with those bikes. If it was, how it was modified to attain a desired weight was the problem, not the material itself.

I still think a well made Ti frame will outlast bikes built from other metals. It's not forgiving towards mistakes, either in preparation or proper welding.

As far as steel goes, I've been riding TK's own Prestige silver brazed fork for the last 19 years, still going strong. It was gifted to me off his former road bike, that hung on the wall in the shop as an inspiration to what perfection could look like. :)

His frame had a cracked DT, near the BB cluster. Tom just wore it out IMO. Maybe he has wattage that would fatigue any frame material eventually. :beer: He was always competitive in Masters races, and put on tons of miles. That bike had 50K miles on it at least.

[AndrewS]

Now that is the horse's mouth.

Thank you.

[jbay]

As I said earlier, I have always seen ti failures near the joints only, and while I have seen the same for steel - I've also seen steel fatigue mid tube. That is where I get my incredibly unscientific (and probably unnecessary) conclusion that a good ti bike may never fatigue, IF every joint was perfect executed and the its built for the rider's weight.My friend's Carl Strong Ti frame died emphatically a couple of weeks ago. I presume it cracked initially near the bottom bracket welds, but there are so many cracks it's hard to tell:

Dead Carl Strong Ti frame (http://picasaweb.google.com/piaw/DeadStrongFrame#)

FWIW, that frame was ridden with a Ti fork made by Black Sheep. I can't find a photo of it offhand, but my friend has toured extensively in the mountains using it and has no complaints.

-- John

[Mark McM]

Several? Where did these failed frames show up? Which tube/s showed a problem? Years of manufacture? I'm sincerely interested.

As a custom builder at Merlin, I also handled frame repair for two years. I don't think I saw more than 12 broken frames, (from a fabrication problem, NOT bike-meets-garage-on-the-roof rack issue, or tubing failure), during that time. Almost all had purge issues: inadequate purge hole size between the chainstay and bridge, from early batches--Right chainstay crack, near the BB.The issue was addressed immediately when a small pattern emerged. All were straight guage 3/2.5.

There were two or three with cracks near the DT shifter bosses caused by inadequate purge between the boss and the exterior of the down tube, (this was an achilles heel for many Ti builders. The advent of STI/Ergo and cable stops welded to the head tube did away with the need for shifter bosses). Double butting started towards the end of my stint there, so I'm not sure if tubing fatigue was an issue with those bikes. If it was, how it was modified to attain a desired weight was the problem, not the material itself. .

Actually, two of the cracked frames were Extralights from about ’92 or '93 (might have been ’94) and the cracks were on the downtubes, near the shifter bosses (both were replaced under warrantee). As I understand it, this was a common problem with the Extralight models of this era, and that it was due to the thinned middle section of the tube between the butts being just too thin for welding, and that as you say the problem was solved in subsequent years by replacing the downtube bosses with head tube mounted cable stops.

Curiously, the Merlin in the EFBe frame fatigue test (http://www.sheldonbrown.com/rinard/EFBe/frame_fatigue_test.htm) also cracked at the downtube near the shifter boss, although this particular model was the non-butted version.

A few years later I saw another Merlin (I think it was an Extralight, but it could have been a Road) that broke a chainstay during the Harvard Road Race (Harvard, MA). The chainstay broke during normal riding during the race (i.e. not during a crash), although it is likely that the chainstay had already had fatigue cracks, and merely the final rupture occurred during the race. Embarassingly, the rider was racing for a Merlin sponsored team (was it Merlin/Wheelworks?) when the failure occurred.

Finally, a member of my club had a Merlin RSR from about ’99 that cracked at one of the joints at the curious “monostay” used at the BB/chainstay junction on this model (can’t recall if it was the BB or chainstay end). This was repaired under warranty.

These are just the fatigue cracked Merlins I have personally seen. A quick google search turns up many more reports of cracked Merlins.

I still think a well made Ti frame will outlast bikes built from other metals. It's not forgiving towards mistakes, either in preparation or proper welding..

My feeling is that how a material is used (including how much or how little of it) is a bigger factor in frame life than the simple selection of the material. Even despite aluminum's so-called reputation for fatigue, there are still plenty of aluminum airplanes flying around that are 50, 60, even 70 years old and older still in regular service. The last B-52 (The U.S.'s main strategic bomber) was manufactured in 1963 (almost 50 years ago), and is expected to be in service until 2040 (about 70 years). I personally would be hard pressed to claim one bicycle frame will outlast another, soley based on the material it is made from.

As far as steel goes, I've been riding TK's own Prestige silver brazed fork for the last 19 years, still going strong. It was gifted to me off his former road bike, that hung on the wall in the shop as an inspiration to what perfection could look like. :)

Similarly, I’ve been riding a steel (Columbus SLX) Nobilette also for 19 years. While the frame is still intact, one of the fork legs cracked a few years back, just at the bottom of the internal lug portion on the semi-sloping crown (a known stress riser in the internal sloping crown design). And now the Serotta content – the cracked fork was replaced by a steel Serotta external crown fork, and all has been well since.

[Mark McM]

As I said earlier, I have always seen ti failures near the joints only, and while I have seen the same for steel - I've also seen steel fatigue mid tube. That is where I get my incredibly unscientific (and probably unnecessary) conclusion that a good ti bike may never fatigue, IF every joint was perfect executed and the its built for the rider's weight.

I think if you look closer, you’ll see that even if a tube appears to fatigue mid-tube, that there is actually a “joint” or or other discontinuity in the tube that caused a weakening of the material or a stress concentration. Fatigue, like other failure mechanisms, starts at the weakest link. For example, it is not uncommon for cracks to start at welded or brazed-on fittings like water bottle bosses or shifter bosses (like the mid-downtube cracks I earlier mentioned on Merlin frames). I've seen a few steel tubes crack at butt transitions if the transitions are poorly designed or located. I’ve even seen some seamed steel tubes crack and split at the seams. Possibly the cracked steel Possibly the cracked steel chainstays you’ve seen were due to some features like poorly formed crimping, butting or swaging, or near the joint at the chainstay bridge.

[Mark McM]

Fatigue, like other failure mechanisms, starts at the weakest link. For example, it is not uncommon for cracks to start at welded or brazed-on fittings like water bottle bosses or shifter bosses (like the mid-downtube cracks I earlier mentioned on Merlin frames).

Just a short follow on -

brazed or welded on fittings don't even have to have poor fabrication to result in fatigue cracks. Even if the base metal retains it original strength, the extreme stiffness differences between a thick fitting (like a shifter boss) and a very thin tubing wall can cause a large enough stress concentration at the corner to accelerate the initiation of a fatigue crack. So I'm not sure I agree that a titanium frame will never crack even "IF every joint was perfect executed", if the the joints are not properly designed to start with.

[AndrewS]

Mark,

I agree with you about aluminum: several tests have shown that modern aluminum frames are the least likely to fatigue because they flex too little to get fatigued. But aluminum airplanes fly for 40 years because just about every structural part has been replaced at one time or another, and all of it has been constantly inspected and repaired. All the helicopters I flew were massively rebuilt overtime, and the aluminum rotors replaced very regularly. If bicycle frames were full of access panels and reinforment girders it would be more comparable.

The breakage study brings me to a simple conclusion: The stiffer the frame the more likely it is to take a beating. That's good news if unbreakable is the only characteristic you want, but not as helpful if you are trying to produce a certain ride. A test that factors in stiffness in the number of cycles might be more interesting.

The steel story I told were three different models cracking behind the chainring, 2 to 3 inches from the bridge. I'll leave it at that.

With ti, I was trying to articulate that if built with the right thicknesses to counter annealing and with clean welds, the material is unlikely to fail. I understand that the joints are the most likely failure point on any bicycle, but even your own examples seem to almost always involve a weld.

I think this might be illustrative: Can anyone think of a ti frame failure of a bike that was ridden regularly for a decade? If my bad weld theory is wrong, then fatigue should show up at all different points in the lifecycle, and mainly in well used bicycles. If the welds are the reason, then the failures should be earlier, rather than from long use. PJBike mention TK's broken Prestige frame that lasted for many, many years, then failed. Anyone think of a ti frame acting like that?

[RPS]

Is there a way - any way at all - to measure life expectancy on certain materials? Say, for example, one has a 25-year-old steel frame: as has been said, every material fatigues. So with that in mind, is there a way to measure a material's remaining strength, which in this concern is steel?
If the gist of your question pertains to how many “additional” miles a used steel frame may have left on it, I think there is no way of knowing unless of course it has already started to fail, which means it shouldn’t be ridden any more. Even with a brand new frame fabricated of a known and tested material, it’s doubtful anyone can predict with any accuracy how many real-world miles it can be ridden before it fails. There are simply too many variables involved.

I don’t see how knowing something like the material having 50% left would mean much anyway unless you can ride the next 100,000 miles under the same exact conditions.

[dd74]

If the gist of your question pertains to how many “additional” miles a used steel frame may have left on it, I think there is no way of knowing unless of course it has already started to fail, which means it shouldn’t be ridden any more. Even with a brand new frame fabricated of a known and tested material, it’s doubtful anyone can predict with any accuracy how many real-world miles it can be ridden before it fails. There are simply too many variables involved.
This is exactly what I mean. Because the frame in question is a 1987 Appel. Now, to be fair to the frame, I've ridden every route with it, including the nastiness of Cahuenga Boulevard west into SFV and Mulholland above Hollywood - two of the most oppressive roads in L.A., so bad in fact, that motorists have tried to sue the city of L.A. for wheel and undercarriage damage to their cars.

With that said, the Appel, which "looks" as if it has been raced on, has held up fairly well. My issue is I need to stop myself before becoming emotionally attached to this frame as I can't be sure how hard it was ridden before I took purchase of it and now long it might last.

I can tell you, though, that the frame has impressed me so far. I have a 2010 Ridley Orion as well, and it's beginning to chatter quite a bit in the head and seat tube region only after about five months of ownership. The Appel, not in the least.

Do I trust the bike? No. Would I trust a new Serotta CDA or Classique Ti over the Appel? You bet. I'd also trust a Gunnar or Motorbecane Le Champion above the Appel.

But the thing is the emotional attachment. This is (was) a great frame. It's set up old school with DT shifters and gives nothing away to newer bikes except for one needs to sit to shift. Where the dilemma comes in about fatigue is also the fact that I'm not sure I want to drop $800 (roughly the price of a new Gunnar, or a nice down payment on a Serotta custom frame) on having the Appel repainted. The pearl white on the bike is trashed. British Racing Green would be my choice. I already have a Brooks saddle waiting for it. :(

Maybe that explanation clears things up. Maybe it doesn't. :crap:

[AndrewS]

Maybe stop riding it on the worst roads?

[RPS]

Maybe that explanation clears things up. Maybe it doesn't. :crap:
Unfortunately on this subject there is little that can be cleared up to make a difference.

I don’t believe that you’ll find what you are after for various reasons; most already stated above by others. Most importantly the economics of testing are cost prohibitive – it would make buying a new frame cheaper (assuming testing was even possible without destroying the frame in the process). And even then I’m not sure it would predict failure all that accurate anyway.

In large and/or very expensive systems what you are after can be done. As an example, a small highly-stressed steel member of a bridge or other structure can be replaced with a new one and the old member tested in a lab under different conditions for fatigue or whatever is deemed pertinent. The information can then be used to predict the life expectancy of other similar members without having to replace them all sooner than needed. On a steel bike frame this approach isn’t practical because even if you had one main-triangle tube replaced it wouldn’t tell you anything about the others which may fail at different times. It also wouldn’t tell you anything about possible failures at dropouts, lugs, chainstays, etc…

I can’t recommend what you should do since I don’t know you or the bike, but will say that if it were my bike and it had sentimental value I’d inspect it carefully for visible cracks on a regular basis and then ride it as I would any other. I normally avoid bad roads regardless of what bike I ride so that wouldn’t be an issue for me. For what it’s worth, I have an old beater I still ride around town that I bought around 1987. The frame is built like a tank and because it has relatively low miles for its age and has very little corrosion I don’t worry about possible failure. I sometimes worry about the stem or cranks but never about the frame or fork.

[dd74]

Maybe stop riding it on the worst roads?
LOL! That'd be too easy, wouldn't it?