Frame failure: Idle speculation on root cause

[roubaixtuesday]

Induced by boredom suffering a cold, paint removal and frame failure investigation time today.

The frame on our 1990s built, columbus steel tubed and fillet brazed tandem failed this summer.

The bike was used regularly but not with huge mileage, I'd guess maybe 15,000 miles total, no more than 30,000 tops. It was completely refurbished including a new paint job in about 2018, which would have revealed any significant cracks at that point, I think.

It had been creaking for a few months previous to the failure.

The failure is at the pilot seat tube, also propagating along the top tube. The seat tube has quite a substantial upstand above the top tube, and because of my height, a lot of seat post was showing above that. And remember the stoker's bars are also attached to the seat post. So the bending moment for this joint is *way* in excess of a normal bike seat tube.

There are several cracks and some of the fillet brazing has come away completely.

There are cracks both ways from the clamp. There's a long horizontal crank one side of the top tube and another smaller one the other side. There's also a crank at the front around the upstand of the top tube, that doesn't seem to connect to any other of the cracks.

So, what say you as to the root cause?

Pictures from front and both sides:

 

[Paulus]

At a guess I would say that maybe you had too much seat post showing, even if it was not above the limit mark. Over time the seat post has rocked back and forth and has caused the failure.
Of course I could be totally wrong.

 

[roubaixtuesday]

I have two pet theories:

Firstly corrosion under the brazing weakening the joint to the point of failure. I can't see any other reason why such a large chunk of brazing would come away.

But that doesn't really explain the cracks from the clamp which surely wouldn't start afterwards? So pet theory two is fatigue at the clamp joint, perhaps exacerbated by overtightening (you need that post really tight as it's holding the stoker bars) and bending from the stoker bars. But I'm not sure why you'd then get the other cracks.

 

[roubaixtuesday]

Oh, and in the unlikely event anyone would like a completely borked tandem frame, you'd be very welcome, drop me a dm.

 

[biggs682]

A couple of new tubes and a lick of paint it would be good to go

 

[roubaixtuesday]

A couple of new tubes and a lick of paint it would be good to go

I thought duct tape or maybe a proper jubilee clip job is needed in this case?

(If you're being serious, apparently the ovalised top tube is no longer available)

 

[Profpointy]

A couple of new tubes and a lick of paint it would be good to go

Black tar and newspaper laminate for a truly professional repair

Joshing aside, it does seem a very short lifespan, and I guess as a tandem it should have been built rather stronger.

 

[fossyant]

Silver lining and all, new tandem time !

 

[Jameshow]

It's a tandem frame so has massive loadings in compression on the top tube coupled with the loadings on the seat tube and it's high probably of failure, be it steel or alu...

 

[wafter]

That looks in a sorry state

The load / stress situation at the clamp is probably pretty complex although the failure seems pretty intuitive. The crack is starting near the root of the cutout which will act as a stress raiser; that area being subject to quite a lot of forces. There's the relatively static bending moment of the rider's mass exerted rearward via the seatpost, the static bending / compressive stress from the clamp itself and the cyclic loading to and fro across the bike during pedalling.

The failure further down is perhaps more interesting as there's no obvious stress raiser - unlike a welded frame I'd expect there to be little to no heat affected zone from a brazed joint (rightly or wrongly). One assumes that the top tubes are butted against the continuous length of the seatpost tube, so perhaps we have an issue with fatigue in the brazing from differential strain / different elastic modulii between that and the steel.. I could see how this could propagate through the brazing but it's interesting that it's gone through the steel too.

It does look pretty grotty inside so perhaps corrosion was a factor in initiating / propagating the crack..?

As for loading I'd expect the cause to be the same as for the clamp area without the loads associated with the clamp itself. Seems like a pretty common area for failure - this is where my old ally Giant frame went, with broadly similar cracks propagating horizontally from the weld forward into the top tube and rearward into the seatpost tube..

 

[raleighnut]

I'd say the corrosion played a significant part.

 

[roubaixtuesday]

Silver lining and all, new tandem time !

This wasn't a very recent failure, the new (to us) tandem was acquired the same day it went.

https://www.cyclechat.net/threads/the-big-tour-of-24.295943/page-2#post-7178885

And very fine she is too, though I say so myself

 

[roubaixtuesday]

I'd say the corrosion played a significant part.

I'm not sure if the crack came first or corrosion (from the inside?) I guess it could be crank first then exacerbated by corrosion??

 

[Ian H]

I'd guess the main culprit was too short a seatpost, the inserted part not extending below the top tubes.

 

[si_c]

I'm not sure if the crack came first or corrosion (from the inside?) I guess it could be crank first then exacerbated by corrosion??

Looks as though it could be a mix of strain and corrosion, the metal around the brazing looks to have delaminated somewhat, so I'd guess that corrosion had been present for a while, I'm not a materials person though so it's as good a guess as I'm going to have.