Sunday, June 14, 2009

Sin-gineering (an FMEA exercise)

By now, every self-respecting cyclist on earth has heard this account of a Mavic R-Sys wheel exploding spectacularly under a VeloNews reporter. Needless to say it's not very good press for our snooty friends across the pond.

If you compulsively scour online bike news to the the first one with every tidbit of new gossip, you may have also already read Mavic's strained response where they attempt, with a questionable level of convincingness, to make the point that the wheel's failure was a result of the crash, as opposed to cause thereof. Even if you did fleetingly entertain the belief that this were the case, it still doesn't absolve Mavic engineers of being cognitively disabled (notice how PC I am today?)

Those of you who are engineers probably recognize the Acronym FMEA--Failure Mode and Effects Analysis. Simply put, FMEA is an exercise that one performs with a product or system to identify the many ways that the product can fail and assess the potential risks posed by each mode of failure, the point being that one must expect that some products will fail and that an acceptable failure mode is equally important a design criteria as the performance characteristics of the product when intact.

Generally, when performing FMEA one identifies the failure mode, (catastrophic atomization of wheel resulting in high-velocity faceplant), the severity (high-velocity faceplant seems pretty severe...) and the frequency of occurrence (I can't imagine they made more than 10,000 of these things, so even one failure is a pretty big deal).

If you're an engineer, your job is to make sure the stuff that will get you sued basically never happens. For instance, you should be pretty darn sure that if you're a wheel maker, your wheels don't blow up if you breathe on them wrong. It may be true that being breathed on is not in the description of the wheel's normal use, but it's darn likely, given that they cost $1,200 and are designed like the wheels on an 18th century pushcart, that someone sooner or later is going to lean in real close to get a good look, and since they likely haven't ridden the wheels yet, they'll still be breathing.

It is for the reason above that the engineering world invented "Factors of safety". If you know that a certain component failure--say the lift cable on an elevator snapping--is going to hurt a whole bunch of people really badly (hopefully you're making the analogy here all by yourself), then you build that component a whole lot stronger--say 11x (which is actually the factor of safety for elevator cables)--than you ever expect it to be strained. When was the last time you heard of an elevator cable snapping, despite there being thousands of elevators in every major city, many of which are shoddily maintained and decades old? Yep, never. It simply doesn't happen.

Anyone who's been riding bikes for a long time knows that awful things happen to wheels. You hit potholes. You graze other cyclists or bushes on the side of the road. Occasionally you slide out or have to jump over something. Rims get bashed. Spokes break. Normal wheels, even the ridiculously expensive deep-dish carbon tubulars, weather these sorts of things every day without exploding and tossing anyone, but if any of the above happened to an R-Sys wheel Mavic would throw up it's hands and say, "not the intended use?" Sorry sir, you're not allowed to ride this wheel in the real world, it's simply too dangerous...

The fact that the road-cycling world would even accept a product so ridiculous in the first place, nevermind allow any attempt to defend it, is another example that there's something seriously amiss in roadie culture. A MTB product that performed like the R-Sys wouldn't last a day. Clearly playing in the dirt keeps some of us grounded.

Ride stuff that works. It could save your life.


Anonymous said...

Hmm, that is a difficult one. I agree with you that the overall idea and engineering of the wheel seem somewhat ridiculous, but not more so than, for instance, Lightweight wheels. Let us face it, these wheels are for sponsored pros or rich wanna-be athletes, totally impractical and not intended for daily use and abuse. I agree with Mavic, however, that the conclusions of the VeloNews guy are premature and, frankly, bad form. I do not want to defend Mavic, personally I would never buy said wheel, but given the many unknowns of the crash and despite the horrific pictures of the destroyed wheel it does not necessarily follow that it was the wheel that caused the crash (remember how the front wheel after Markus Burghardt's crash looked: We can all only speculate what the true cause was, maybe it was the wheel, maybe it was the frame, maybe it was the tube or tire, who knows. Anyhow, one more reason to avoid all this dodgy-costs-be-dammed-lightweight-centred-carbon-fibre-in-everything-philosophy.

Unknown said...

Certainly no argument on the philosophy being a large part of the problem, but there's a big difference between hitting a 70lb, hub-high brick at speed with your wheel at a 60 deg angle and the incident under discussion.

There are plenty of situations that one is never going to ride out where it effectively doesn't matter what happens to your bike. My point is that everyday things like hitting potholes, rolling tires, etc. shouldn't result in catastrophic failure of the kind seen with the R-Sys. Such things may permanently damage ultralight race wheels to the point where you might not want to ride them more than the 20 feet it takes you to stop, but they should still resemble wheels for that 20 feet, as opposed to literally blowing to bits.

I'm on those pics from a comments back, btw. We're still working on the bike...

Anonymous said...

Completely agree, my point only was that just because a wheel looks awfully destroyed does not necessarily indicate that it was also the cause; although I admit that considerable evidence points towards Mavic in this case. Ignoring the overall philosophy of such wheels, where we both can only shake our heads, the case nevertheless remains a bit of a mystery. Personally, I doubt very much that the wheel failed in such a dramatic fashion without any prior damage or impact; a bold statement I know. My speculation is that the wheel was either accidentally damaged before (eg during transportation or shipping), or that the tire/tube blew and caused the rider to crash, which at 25mph also destroyed the wheel and bike (but in that order). I also remember the episode with Richard Virenque who crashed badly some years ago and blamed it on a Time's fork. The result of the investigation was, however, that his QR were not secured properly and that this caused the crash when he tried to jump over a small obstacle. Anyhow, I guess we should wait for the outcome of Mavic's investigation. Looking forward to the Yeti pics!