Here's the remaining photos from the incident:

http://members.cox.net/younggg/Wheel4.jpg

http://members.cox.net/younggg/Wheel5.jpg

http://members.cox.net/younggg/Post1.jpg

The last shot is of the seat bolt that broke when I landed...

If you look at the break right above the spoke in the 1st picture, you will see a spot that looks discolored, this could be an indication of a pre-existing crack that expanded.

Look at my previous post about metal crack analysis.

You mean it looks like the fatigue striations of stage 1 in fatigue failure?

You could equally argue that the actual of angle of shear through that section is sufficiently off boresight that its partially obscured by shadow.

Without seeing it right in front of you, we're limited to partial analyses from low res pictures. I'm just thinking massive oxide bifilm defect that finally went.

I'm betting that this is a delayed failure of a cracked rim after it got hit by something like a pothole or a curb (my father destroyed a wheel in similar, albeit immediate, fashion with similar damage pattern after running into an unseen curb).

From what I can see, the impact started a crack on the outer edge of the rim on the break with the jagged edge. The crack propogated under fatigue stress to the spoke hole, then turned and propogated along the tip of the "V" in the V-section rim to the next spoke hole. The crack then turned at the hole (because the hole is reinforced) and propogated toward the opposite braking surface. Failure was when the crack finally reached the other edge of the rim. (See attached picture)

This is definitely not due to air pressure or tire size (because the initial failure is a radial crack on the rim, rather than circumferential), and for similar reasons, not caused by a worn braking surface (besides, the rims only have 500 miles on them); these two failures often go hand in hand besides. It would have failed either due to a hit on a pothole, curb, or other object known to taco wheels; or it was due to a manufacturing defect or mishandling to cause the initial crack.

Uh, translation? I'm not trying to give the impression I'm an expert at analysing metal failure, or that I know what happened here, all I'm saying is analysing the fracture site could possibly tell us what happened.

So in layman's terms, what did you say?

I have the wheel. Everyone has looked at it. There are two other places the rim is starting to tear apart from a spoke hole on the tire side of the rim. It looks like it is being pulled apart. We think it is a combo of things but primarily the wide tire. Hopefully we will get up with Mavic and get an RA so they can examine it.
"pause"
I just checked the front wheel he sent back. Using the park tensionmeter, I built the wheels with, it averages 21(99kgf) with no spokes higher than 22(111kgf). Mavic in their manual lists 90-110 and the guy I spoke to when I had a Reflex with cracking at the eyelets said a max of 110. So I feel that overtensioning is not the issue.
I have seen many damaged wheels but I have never seen one break on the inside, the failures are always radiating from the eyes. Or in the case of a worn rim around the brake surface.

Sorry.

Fatigue is a two (really three) stage process. Ignoring crack initiation, stage 1 would be the part where a crack grows slowly with virtually zero overload for a while, and is characterised by deformed and worn surfaces to the crack - where the crack has repeatedly opened and closed/moved, bringing those crack surfaces together again and a again. The second stage is the catastrophic failure part, where the remaining material is too small to hold the load anymore. Normally, the first stage takes hundreds or thousands of cycles, so that worn area will be polished and smooth and shiny. But in cases where it has only taken a few cycles before failure, it won't have ad tiem to polish or smooth and will be dull and rough.

Second; all metal components start off life as castings. Then they're forged and shaped, etc. Aluminium is prone to oxidation, and the vast majority of manufacturers of the bulk alloy either don't know or don't care that there are set rules as to how fast and how far you pour aluminium, and any other alloy that readily oxidises that readily. You throw the metal into the mould turbulently and the surface film of oxide which is dry and hard wil be wrapped up and dragged in into the casting. Long story short, all Al alloys, all stainless steels to a lesser extent, Ti alloys, Mg alloys suffer this problem. These films are cast-in cracks. Sometimes they weld up during forging, sometimes not. Especially wil Al and Mg alloys. I've never seen it in a rim before, but I've seen elsewhere Al components that have failed from a film-crack that survived forging and went on to remain in a component.

Unlikely, but possible.

The other wrenches and I at the LBS I work at looked at the pictures. Here, based on the pictures alone, is our theory:

First, a fact - this is not a high end wheel. It may say Mavic, but is one of their lower offerings.

Now, the theory: judging from the cracks around the spoke farrells that survived, it looks the spokes were over tensioned. You can see where some of the other spokes are pulling out. It is possible - I have seen it done - to tension spokes to the point where they will break or they will put the rim badly out of true. It is possible that the last time the wheel was trued, somebody got a little too boistrous with the spoke key. If a number of spokes were too tight - combined with the normal stresses caused by a rolling wheel and the fact this wasn't the best wheel in the world - it is possible for the wheel to explode.

At least, that's our theory.

The wheelbuilder says on Page 2 that he used a tensionmeter.

I still say a rim defect.

Hmm, from the hundreds of wheels that I've built while employed at a shop for 10-years, too-tight of spoke-tension will turn a wheel into a taco shape well before it's tight enough to pull spokes through. I may do it through fatigue with sufficient time at just below that critital tension (takes years). And they typically start cracking at the spoke-holes themselves and propagate outwards in a spider-web pattern. From looking at that 1st picture, that crack definitely did not start at the ferrule.

All this talk... It can't really be anything but 1) defective rim 2) Something hit the rim, starting the crack. eg. a crash in which the rim landed on a rock or some such. Rider weight certainly contributed to the final failure.

Thanks for posting this. Learned a lot about making wheel decisions. I know my price range theshhold just went up by $100.00, I think that seatpost picture did it.

Mavic should take notice of the number of reads of this post, approaching 4000 in one day, and realize the P.R. impact this could have on them. They are already getting bashed as to there responsiveness and what there expected answers will be . CYA position. Right or wrong thats there Reputation to 4000 veiwers.
If they are smart they should get a spokesman on here or at least send a +message through the Builder.

Ozrider, this was not a low end rim. The CXP33 is one of their more expensive rims. It has a welded seam, machined brake surface, and sprung inserts that are beveled to fit the interior of the rim and spread the load of the nipple/spoke.
The rim has two other places where it is cracking, that are not shown in these pictures. It is cracking from the tire side of the rim, not the nipple side of the rim.

Chuck, if you can make some measurements on the tyre I can caculate the forces on the rim-flanges. The width of the flattened tyre is all I need, from bead to bead. Basically the lateral circumference of the casing. :)