Does anyone have any thoughts on this failure? I'm curious what the typical inspection schedule and process is for a roller coaster like this.
I'm also curious about the root cause of the failure. I have a guess, but it would be interesting to see an actual failure report. I don't expect to see it in the news, though.
Here is a video of the crack moving as roller coaster cars go by.
Someone also got a photo of the crack a week prior before it had fully propagated through the support.
A wise senior engineer once told me, "There's never only one problem. There are at least two and they work together against you."
My armchair quarterback take is that there was a stress concentration at the weld, and also that inside corner wasn't well painted so corrosion was a factor in initiating the crack.
I don't see corrosion as a likely issue; there is no staining of the paint in the location of the crack and the crack likely propagated from an area protected from water by the joint to the track.
A stress concentration at the weld would be due to poor craftsmanship, which I mentioned earlier.
Good point about the lack of paint staining.
The Wikipedia article for the coaster says the crack started along the weld line, so my current hypothesis is a faulty weld.
I know I'm just an out-of-work aircraft repairman so the next thing I'm going to say is probably insane, but if it were my roller coaster I wouldn't re-open it to guests until I'd replaced that support column and done some NDT on that section of track as well as all the other support columns. I'd likely have nightmares for years about some weld somewhere failing. See why I don't work on planes anymore?
You aren't wrong, but they are also going to need to verify the fatigue design of that detail. That connection detail is rather rigid, so the model they used for the design of that weld may not have accurately modeled the stress at that weld.
This is the kind of issue someone can get a PhD trying to solve.
Not insane at all. I hope an NDT contractor is going to get some business out of this incident.