The cause could be something as simple as a poor weld. If you look closely, the crack starts at the weld and moves outwards.
B&M used to do alot of overbanked turn supports like this. Very similar in design to Fury 325's and other modern hypers, but with one key difference.
The direction of the support....they are the opposite way around.
It is interesting because on the first photo, the support will be under compression when a train passes, very difficult for cracks to appear and welds to fail in compression. Fury 325's support however, due to it being the opposite way around, quite literally puts the opposite force to compression on the top area of the support (where it cracked), in the form of tension. The crack has appeared at a point in the support where tension would be at it's highest, further agrivated by the fact the force from the track would be completely sideways due to the 90 degree bank. Which in turn will push more load into the diagonal support through the weld.
B&M must have a reason for doing supports like this as many modern hypers use tension supports rather than compression on their overbanked turns. Mako uses them, but the overbanked turn is more overbanked than this, so less load would be transferred to the diagonal support through the weld. So I would imagine they generate less fatigue damage as the train cyclic loads the support. That could be a potential cause, if the weld was crap.
Even our own Nemesis uses compression supports on the overbanked turn. Compression as in the train and track will push into the supports away from the track. Rather than back in on itself, like a tension support.
Subsidence could be a possibility, that too will put a static tensile load / stress into the area where the crack has appeared. In addition to the dynamic tensile load as the train passes. Both would be too much for the support I would imagine. Leading to what we see here.
Whatever it is, a crack like this probably appeared very quickly, because it originates on the weld on the highest tensile stressed part of the support. That support would have been in two after less than 10 cylcic loads (trains passing over). It may have been detectable using ultrasound before it failed, but that is beyond the scope of a daily check. Once that weld failed properly, the whole thing would be cracked very quickly.
It would not say the fact they are fabricating a support so quickly is any indication of them not doing any changes. You could design a support, from scratch in a couple of hours. They are fairly simple to design and fabricate relatively speaking, lead time would not be affected as they are usually designed onsite in the same factory as fabrication.
As it is a tensile crack. On a tension support, it is possible they may strengthen the area where the crack appeared as a fail safe, even though it probably won't be needed.
I very much doubt B&M got their mathematics wrong. Which leads me to speculate that it is highly likely, subsidance, a poor weld or something else that doesn't point the finger at B&Ms legendary and usually spot on mathematics. There really is no room for the slightest error in a weld on a support you are loading with tensile stress. You could get away with it on a compression loaded support, not a tensile loaded one though because this will happen.
I think the video of the train going over the broken support, highlights better than anything, the excellent maths and engineering of B&M to be fair. I am not ruling out that they got their maths wrong, but I see that as a very unlikely scenario..
