Frame failure experiences

[novocaine]

6 months at the workshop in 2010 it was:

Titanium track bike being used as a commuter - spiral cracks along the seat and head tubes

Carbon road bike fell against a lamp post - top tube split right round like an egg

Recall for most of that year's Specialized full suspension - shock mounts shearing off

Most common was YouTube failures - people trying to emulate videos, pretty much destroying their Konas yet expecting warranty replacements

Historically, the early burly 7005 aluminum frames all suffered the same issues: the forging process didn't get rid of the granular structures in the aluminium tubing. This led to multiple failures at the head tubes, especially when headset cups were not always the exact size or the tubes prepared in the right way. This left many cheap to expensive frames failing, some even simply going pop in front of the owner at a retrobike meeting! Manitou being the most infamous but most early Fisher would have the familiar 'lighting strike' crack in the head tube as it followed the grain structure

Welding has been discussed so many times - the most important part being the weld penetration. A lot of arguments then ensued about the look of a weld, missing the point entirely that as as long as the weld has been done correctly, looks are not as important. With aluminium, the structure changes with the type of material. 7000 series has different properties to 6000 and needs to be treated accordingly. When you weld aluminum, you have the tubing of one type which is then joined buy a rod of a completely different type of material but once through the stress of the welding process, ends up as near to the original tubing characteristics as possible

Get that wrong and you'll have your joints failing like a 3am rollup

The same 'Olympic' frame builder that taught me all the above (which I've no doubt got wrong) also explained that titanium had similar issues with the heat and stresses of the welding process changing the physical structure at the joins. So the welding had to be done to strict proven methods and standards

If I also remember correctly, decent welds for all materials can often be stronger than the tubeset

Back to frame failures - exotic American frames imported into the damp UK often rotted from the inside out, many shown on this forum

And don't rely on those expensive brand names being micron accurate for their headset cups - they are not!

Overlong forks ripping headtubes off - horrible!

Front end collisions, have seen many a decent frame ended like that..

They are quite correct, the weld is often considerably stronger than the surrounding material. a weld rarely fails, what fails the substrate around it, normally right at the edge of the bead. this is partially due to a massive change in the grain structure of the material due to heat stress and also due to stress risers along the bead edge. aluminium and Ti are both known for this, it's normalised after welding by heating and cooling cycles. no such requirement in steel, but often what happens then is the bond isn't full at the edge of the bead and that creates a failure point.

We used to do pen tests on welds (as well as tensile tests). cut through the weld, polish and then pickle. the number of welds you'd get in with the "stack of dimes" look where the weld only penetrated 1/2 the substrate was enlightening. I'm well out of practice these days, but when I do weld (recently built a new office chair), I go with functional welds over pretty welds, and I know they must be full penetration because of the holes I blow (not really, you can normally see the penetration from the backside where you can also see the heat soak).