Hypno Toad

A couple years ago, I retired an old AL frame bike. It wasn't spontaneous, it was very subtle. I notice the BB was feeling very soft (extremely soft). I thought the crank or BB was loose. But took a closer look and found a massive crack in the frame around the BB and down tube. Here's more details in a thread I started:

https://www.bikeforums.net/general-cy...ld-friend.html

Edit/afterthought: I have 3 other AL frame bikes, including my Kona Dew Drop. This AL frame gets some very hard miles, including: gravel centuries (and longer); winter rides with salt, snow, ice, cold; CX 'races'; single-track MTB trails .... 7 years and over 16,000 miles and I'm not worried about the frame in the least.

IMG_20160611_055959309_HDR.jpg

corrado33

You can find pictures of failures for every material of bike. However, what you don't see is the hundreds of thousands/millions of bikes that HAVEN'T failed.

I'm with some of the guys on the previous page. Frames will last forever unless they weren't built properly or have some sort of manufacturing defect or are crashed regularly. I have seen PLENTY of old 80s/90s aluminum frames that have been beat to crap but are still going strong.

If you are a material's scientist that works with stress/strain, you'll probably be familiar with the graphs that plot cycles of strain vs. how long the material can handle it. Steel is TECHNICALLY the only material we use today that has an "infinite" life, provided the strain stays below some point (I'm not a material's scientist... obviously.) Everything else works hardens and should eventually fail (whether that happens in practice is still up in the air.)

Ah, here we go, it's called the fatigue limit and endurance limit.



See, steel flattens eventually and ceases to get worse, whereas aluminum continues to get worse and eventually fails.


Unfortunately, unlike steel, welding aluminum has consequences. It often makes the aluminum around the weld and weld itself a lot weaker than the parent metal. (Hence the common failures in Al bikes are often at the welds.) However, I believe that can be mitigated by heat treating the frame after welding, and I'd imagine that's done on bike frames, but I can't be certain (I don't build frames.) I can definitely see cheaper Al bikes not being heat treated, making them more prone to failure.

fietsbob

note this date and count the days, till it fails.

A rigid frame in aluminum will last longer, than a flexing one.

70sSanO

It will last 32,612 miles based on the specific number of crank revolutions per mile.

John

himespau 

I wouldn't worry about it unless you need to convince your spouse that you "need" a new bike.

deaninkl

Dr Isotobe, I stand corrected, my knowledge of these alloys is somewhat dated.. to the late 70's.. thanks for the updated knowledge.

coominya

Yes, so if your bike isn't dented, don't worry.

cyccommute 

A couple of points. The difference in density between 6061 and 7005 is so small that to say there is a difference is splitting hairs. 6061 alloy has a density of 2.72 g/mL vs 2.78 g/mL for 7005. The difference between the two is 60mg. Assuming 500 mL of aluminum to make a "normal" bike, the difference would be 300 mg or, for the metrically challenged, 0.01 oz. Hardly a difference.

7005 is somewhat cheaper but that's because it doesn't need to be annealed after welding. It's not as easy to work as 6061 which means that it's harder to form odd shaped tubes like many modern bikes have.

bradtx

The great majority of failed frames (not crash damaged) that I've seen, or have seen pictures of, or heard about have failed where they were joined. This includes steel and Ti frames as well as AL. I've seen some pretty good dents in AL frames that didn't effect the frame's longevity. Dents with a teaspoon-like profile where there are no sharp edges that can lead to propagation of a crack (like from a handle bar strike) seem non lethal.

A couple of years ago I bought an aluminum framed bike with a sharp crease in the TT just to satisify my curiosity. When I gutted it for an overhaul, I found that the chain stays were disfigured from the add-on side stand. The bike was a very high mileage four pannier touring bike, doing so with chain stay damage. I use it as my beater bike with beater a focal point. I've been waiting for it to fail...

Brad

XCSKIBUM

There are B-52 bombers still flying that are older than most posters on here.

Most cracks/failures are due to damage and or being in a bind from being tightened down with some "spring" in the metal.

DiabloScott

You and a lot of other people are misreading this chart.

All fatigue happens because of an initial crack, which propagates with cyclic stresses. Once you have that crack in steel or aluminum, fatigue will begin and that chart does not apply. Almost always, that crack is a point where there has been damage from trauma or possibly a welding mistake; that chart is for undamaged test samples of aluminum and steel.

Sometimes, the crack can happen in undamaged metal from many many cycles of high stress, and the metal structure just starts to fail; this happens with both aluminum and steel. What that chart is showing for steel, is that if the stresses never exceed 30,000 psi, then those cyclical stresses alone will never cause the initial crack to get started, and with aluminum there is no minimum stress where you can guarantee that won't happen.

So if you put your aluminum bike through one million cycles of stress at 25,000 psi, you would probably see this cyclic-induced crack start to happen. And if you had a steel bike with the same number of cycles and the same stress you probably wouldn't. If you upped your loading to 30,000 psi, you'd see the same effect at 200,000 cycles for aluminum that you would at a million cycles for steel.

And if you can restrain your power to a more reasonable 13,000 psi per cycle, you could go a billion cycles without the subject crack initiation in either material.

corrado33

I'm certainly not misreading the chart (I just didn't want to get into it too heavily, post was already long), I was simply commenting that steel is the only material we have that can POSSIBLY have an "infinite" life, provided the cyclic stress stays below ~28 ksi (and no cracks form from other sources). Aluminum does not have this property. (CF gets really weird, but it definitely has a failure point.)

Now, is that sort of stress normal in bikes? Probably not, but the fact still stands. Undamaged steel will last longer than undamaged aluminum (or CF) provided the stresses stay below the endurance limit of that particular steel.

You commented much on welds, and that's another point of contention. Aluminum welds are often weaker than the parent material, where steel welds are often stronger. Another point in steel's favor for "lasting longer."

But in the end, I'm not here to convince people that steel is the be-all-end-all best material. It isn't really. Not for all bikes certainly. I think the point I was trying to make is that if a bike is well taken care of, not dropped, not wrecked, it's frame will likely last the life of it's owner, and probably more, provided the frame was built competently in the first place, regardless of what metal it's built out of. (Jury is still out on CF longevity.)


And I know those numbers on the x axis look high, but if you think about simply the stresses in your 700c wheel. Roughly a circumference of 2 meters, that wheel turns around 25,000 times in one 50km ride. Do that 50k ride 4 times a week, that's 100,000 cycles. Do that for a year, and that's 5,000,000 cycles. Basically what I'm saying is this we can easily ride a bike long enough to reach the point where steel has reached its endurance limit and aluminum is approaching its failure limit, especially if you're someone who rides daily.

But then again, many people don't ride like this, and other forces come into play when determining the longevity of a frame. Forces such as "how many times has the bike been struck by a shovel handle as you walk past it in your shed." or "How many heavy things have been leaned against it for long periods of time", or "How many times has the bike been knocked over in the garage?" etc.

Basically, unless damaged by a human, a frame (regardless of metal) can last an extremely long time. Longer than most people would want to keep a bike.

DiabloScott

This phrase implies that the curve is showing some kind of failure rate with respect to time, that is not the case. It is the cycles to failure at a given stress.
I always thought the x-axis should be the stress, makes for a different perspective.



Yup, it's not the number of cycles that's the unknown here, it's the stress. The fact that you can go 5 million cycles without fatiguing your aluminum rims is evidence that the stress isn't anywhere near the curve. Although the wheels are a better point for discussing this kind of failure because it does happen, especially near the spoke holes.

corrado33

Yes sorry, when I had written that I couldn't think of a better way to say what I was trying to say... But anyway, thanks for the clarification!

Phloom

https://www.michaelvester.ca/wp-conte...3/IMG_0111.jpg
It is over 15 years old and it get's ridden everyday through the winter, eh, An aluminum bike made by Canadians, eh. For the Canadian winter, eh.

Ironfish653

So, I suppose I shouldn't be riding centuries on a 40-year old frame with aluminum tubes and steel lugs, since it's prone to fail at any second now?

himespau 

Do you know if it's screwed and glued or just glued? I'm probably more paranoid about really old epoxy than most folks.

Ironfish653

No screws or glue. It's a '70's Bridgestone; The tubes are swedged on the ends, and the 'lugs' cast around them.
My 'SuperLight' has alloy tubes, the more well-known 'Submariner' had Stainless Steel.

Typical belt-and-suspenders Japanese engineering of the era, there's probably enough alloy in this thing to build two Synapses.

himespau 

Hmm, I'm not familiar with that form of connection. I'd thought they all used some form of glue.