Do you trust your handlebar stem?

Fatigue needs two things to do damage to a structure:

1/ repeated cyclical stress
2/ a stress concentration

So if an item is unused it won't fatigue, it might corrode, but it won't fatigue.

So time limits set by manufacturers are probably based on moderately heavy usage to suit their insurers. Stress cycles close to the failure load of the structure cause a lot more fatigue damage than smaller loads so they probably make adverse assumptions regarding rider weight and road smoothness.

Stress concentrations can come from 3 sources - Poor design, manufacturing defects and corrosion. Poor design is pretty rare these days as the risks are well known. Many fatigue failures result from manufacturing defects, and these mostly come from welds, getting a weld right is not a trivial task and all kinds of defects can lurk unseen in the body of a weld and these defects will often be in the form of cracks which generate stress concentrations at their tips and you get fatigue. Castings reduce the risk of defects but a poorly executed cast can also contain defects, if you are looking with suspicion at your stem following this conversation the manufacturing process with probably the smallest risk of fatigue is forging - but you normally pay extra for that.
Corrosion can generate changes in profile and cause cracks so even if the corrosion has not resulted much loss of material it will increase the risk of fatigue damage.
By the way I ride a bike with a welded 2 bolt stem, but it's from a reputable manufacturer who I'm sure will have engineered the weld and have appropriate QA procedures that will significantly reduce the risk of a defect that's likely to result in fatigue damage

I started building custom wheels recently so have learned a thing or two about liability, as well as real shelf life of bicycle components.

All the bits on your bike fall somewhere on the weight/ durability spectrum. Older stuff, like quill stems, were very durable, but the industry ceded durability for weight in the 90s. These days you can buy a bike for not a lot that's half the weight of Bernard Hinault's bike, but I can guarantee it won't last as long when subjected to the same use/ abuse.

Stems are much the same as spokes (and anything else); they have a certain number of cycles they can tolerate before failure. 10 years for an alloy stem may be a long time, but it's more dependent on the use you've put it through in that time.

2 years for a stem is legally right; almost all people will not put a stem through enough abuse for it to fail in that time. Beyond that and it becomes a matter of rider weight/ road quality/ weather exposure. Chances are a stem will last a decade if ridden on smooth roads by a light rider on sunny days, but from a liability perspective you can't guarantee that and so have to mandate a shorter shelf life, so you don't get sued if your product eventually fails and causes an accident.

- coincidentally was checking out stems on Rose's website earlier today and came across the following advice:
"For your safety, please change aluminium stems and handlebars after two years or 5000 km at the latest. After a fall you should change these parts immediately."
Does ANYONE do this? 5000 kms? !!!

The only time I've ever sheared a bolt head was when using a torque wrench, that was also on a stem faceplate.

Glad you survived in one piece, a 30mph tumble in the dark must be horrific.
Outside of cycling, I enjoy a spot of rock climbing and it's interesting to compare the industry response when equipment fails like this. In the event of safety equipment failure, the standard practice is to send it straight back to the manufacturers, who by and large are incredibly responsive - precautionary recalls are not at all uncommon. SRAM hydraulic brakes aside, cycling manufacturers seem less bothered?

I remember these ITM 2 bolt stems failing about 10 years ago, the bolts were gash and there was the torque vs thread friction issues.

Not quite happened to me, but my clubmate's stem faceplate cracked. Luckily I happened to notice it when we were stopped, so a careful ride thereafter, no swinging out of the saddle. Stem binned.

I've had a nosebolt go on a Cinelli 1A stem, but that was due to overtightening. Very hard to get them tight enough to stop the bars moving if you hit a hole. After such an incident I wanged it up. Next day, speeding throguh some village in the Ras - BANG - bars shoot round and flop sideways. Lucky to stay upright there.

MKultra wrote:

I await the storm of pedantic rage but be honest chaps, buying a torque wrench is not a cure for being cack handed.

That's exactly what it is a cure for, do bolt up to specified torque, then stop.
Not, tighten, tighten, tighten, and then a bit more till you think you got 11.8Nm

The thing is most people don't realise just how little 15Nm and less is, especially when they're swinging off the end of an XL length Allen key.

You can even get the ones that stop going beyond the set torque if you're really stupid/hamfisted.

I have never had a stem fail on me or a bolt snap (for now touch wood), I also have never used a torque wrench on ally or steel stems.

With the exception of carbon where it is critical to get it right with a torque wrench I would be of the opinion that if you don't have the feel for getting it right yourself with just an alan key then perhaps you should be taking it to the LBS rather than doing it yourself.

I await the storm of pedantic rage but be honest chaps, buying a torque wrench is not a cure for being cack handed.

When you design a bolted connection like stem / bar joint you calculate the tension in the bolt that gives you a strong enough joint. The problem is that measuring the tension in the bolt is pretty tricky without special equipment so the most practical thing to do is measure the torque you impose on the nut or the bolt head as a "proxy" for the actual bolt tension. The problem is that it's not very accurate! It's a classic first year engineering degree lab to torque up a series of bolts in different conditions and then measure the actual tension in the bolt. If I remember correctly it's about +/- 25% accurate. The problem is that friction in the bolt/nut system eats up the torque and very little is used to actually tension the bolt but the friction in the system is also very dependent on condition. Some factors include, bolt & nut material, type of grease used (if any) location of the grease on the bolt, foreign bodies & corrosion. If you don't replicate the conditions the manufacturer assumed when they worked out the recommended torque you could be a long way out from the bolt tension intended. And of course we are all using calibrated torque wrenches aren't we? So Dr Fabulous0 when you torqued up to the manufacturers recommendation you may well have prepared the bolt better the manufacturer expected and so didn't reach the required tension.

Quote:

I am writing this just to warn others to look very critically at their stem and ask yourself is it fit for purpose?

Maybe *that* is what Chris Froome is doing?

http://chrisfroomelookingatstems.tumblr.com/

ITM Millennium!!! Isn't that over 10 years old? I think that, right there is your problem.

Truth is the stem/bar interface is never going to be that brilliant from an engineering point of view... but things have moved on significantly from those Millennium days.

Muchos kudos for simply taping your bars back on and riding home... that is truly hardcore!

Get well soon.

I had a stem faceplate fail once due to corrosion, it'd done a couple of harsh winters.
A big tug on the bars pulling onto a roundabout at rush hour and the faceplate disintegrated leaving the bolts in situ. Luckily I rode no handed across the roundabout like a boss carrying the bars, more luck than judgment mind

Would you be able to name the brand ?

Two bolts is a poor design as you now realise, if they are not tightened evenly then one can be taking excessive force, four bolts or one are better. I've seen a few failures like this though, threads stripped and cracks and even once a stem shear off at the weld, luckily none of these resulted in injury but poor manufacturing quality was the issue in all these cases rather than incorrect installation.

I've been a bike mechanic for many years and I have to warn you guys not to put to much faith in reccommended torque settings, it is not uncommon to do up stem bolts to the manufacturer's specification and find that the bars will slip under a slight load. An experienced mechanic can not only make a fair estimate of torque without the torque wrench but can also feel when the bolts are under sufficient but not excess tension.