I emailed Dave from Mammut again. He apologized for not following up andsaid he planned to be in Switzerland next week. He said he would look to see if any of the testing mentioned in my post from last July had been done. I'll try to follow up with him again in a couple of weeks.
I think the thin hard sharp steel of bolt hangers in general is a big part of the problem. The carabiner is soft aluminum. Alot of force, a little twist, and the bolt hanger will cut right into the aluminum.
Put permanent steel rings or chain repair links on all bolt hangers:
Do you have any sharper photos of the biner? I'd love to see closely what the metal around the fracture looks like.
Doesn't sound look like the gate was open, I've seen a few broken 'biners where that had been the case, and the biner is always significantly bent, not broken clean like it appears yours was.
I think that our gut instincts about a burlier carabiner being better than a light one are likely to be meaningless. There is so much design put into modern hardware you can't just eyeball it. It's better to trust the engineers and the testing labs and focus on the things you can manage best yourself, which is biner orientation, sling type and length and other factors.
In the OP, the biner which failed was the first one from the belay when the FF was the highest. One thing to consider in such situations is to use a locking carabiner on the bolt. I've climbed with guys who carry one or two draws with lockers and use them on their first one or two bolts, usually on slab climbs.
I doubt most people would consider doing this, but it is one solution.
Here is the latest I received from Dave at Mammut.
If you look at the pictures you can see some patterns—the dynamic breaks with the gate open virtually always break in the same place, right where yours did. This is the most common type of break, as it takes far less of a load to break an open-gate biner, and if the load is somehow applied out onto the nose (i.e. not aligned directly along the spine) it can take a truly small load to break—it’s not common by any means but it is not unheard of. Closed-gate breaks are FAR less common and pretty much always show significant deformation of the overall carabiner shape, and on a pinned gate or a wiregate like this they pretty much always break on the nose of the carabiner, as you see in the photo.
Generally in a real climbing situation it isn’t realistic to generate that much load in a static manner until you get into wacky hauling situations, etc.
The scenario that your carabiner broke under is actually the most common one for a biner to fail—a swinging fall causes the gate to rub against the rock, as the biner swings the rough rock pushes open the gate, and the load ends up applied out nearer to the nose of the biner than the spine (either due to the way the biner hangs in a bolt hanger or because a sling is pushed out to the side, etc), and near a belay/tied in tight so there is less rope out and less dynamic element to the belay to absorb the force of a fall. Under these circumstances it’s unfortunately not too difficult to break a biner if all the stars are aligned just the wrong way—for this reason I would recommend always doubling up pro on critical pieces or using a lightweight locking biner or something like this, especially if there are terrain features that could come into play against a biner.
The batch testing we conducted showed that there was nothing abnormal in the metal itself, and the breaks we conducted again on our stock from that batch confirmed this. All the breaks were well within the parameters we expect and well above the stated loads printed on the biners.
So in that case, is a wire gate is more likely to be sprung open when dragged across the rock as it would only take a little nubbin to hook the wire and pull it out of place ?
A non-wire gate biner may require a higher nubbin size threshold to open it as it dragged across the rock. Would the rounded surface area of a non-wire gate deflect more nubbins than the wire gate and therefore be less prone to opening.
I thought that the premise of wire gates was they are less prone to opening ? Perhaps the wire gates were only tested in free space and not in rock surface contact type situation ?
Not that I know much about it .... just speculation