A couple of people on the bulletin board asked me what I thought about the Minneapolis bridge collapse and what caused it. I wrote the following, which provides very little insight, but I haven’t posted anything on the blog in a while. We were kidding on vacation that I would need to get back to help fix the bridge, but I was at least able to scan the documents I mentioned below which were requested by the investigators. Every bridge collapse causes engineers to change how we do our work, and this won’t be any different even though it isn’t known what caused the problem yet.
Here’s what I wrote about why the bridge fell:
I don’t think anyone knows right now, which seems odd because usually some likely theories surface pretty quickly. One of the problems with truss bridges is that each side of the bridge is a beam. Most bridges have a bunch of beams underneath the deck spaced out evenly. With this bridge, if one beam failed then that whole side fails and the rest of the span can’t handle twice as much load (plus the rotation). Worse, since the beams are continuous over the piers, you lose multiple spans, so essentially the whole bridge fell.
I saw a report done several years ago where an investigating consultant was saying that the bridge was subject to fatigue problems (repeated bending, like when you bend a paper clip over and over until it breaks). So there could have been tiny cracks that finally gave way. Or it could be a corrosion problem that went undetected. But it also seems highly suspicious that repair work was going on at the time, so it seems like that will have something to do with it, whether they accidentally cut something they shouldn’t have, or overloaded the bridge with equipment. GDOT was asked by NTSB to provide our rules for what kind of equipment can be placed on a bridge, but I’m sure they’re looking at every angle.
Or it could be none of the above and the foundation failed somehow. Or something I haven’t thought of. It is certainly being watched closely by all of the bridge engineers across the country.
One thing that is very bothersome is that it went so suddenly without any kind of warning. Bridges should be able to take a lot of punishment that would render them unusable or force them to be replaced, but without giving away completely. Even the bridge in Montreal that failed last year (from internal corrosion probably) had hunks of concrete falling off several hours before the bridge gave away.
One theory I heard was that while the surface is not supposed to be part of the supporting structure (load bearing?) that it can become part of it when another part of the bridge fails. Since the repair was resurfacing, it is possible the crew unknowingly broke the very thing holding the bridge together.
An amazing thing is how the bridge fell flat such that relatively few of the hundreds of cars involved ended up in the water. It seems like a left or right side failure would have created more rotation… and more cars in the water.
Possibly, but they weren’t even taking off all of the concrete, just the top two inches (though if they hit a bad spot they would go all the way through). I imagine it will be a number of factors. Heavier and more frequent truck traffic will be significant if it was a fatigue problem. There may have been bad maintenance, inspection, some design flaws, etc. None of them by themselves would have caused the bridge to fall (it stood for 40 years after all). Something probably happened that day that was the straw that broke the camel’s back, but that may not be the whole story.