LogFAQs > #979569371

I'm a bridge engineer, so I might be able to shed... Hopefully a little bit of light on the bridge questions. No background in forensic-type investigations, but I at least know enough to be dangerous. Also don't know a ton about this specific bridge so I'll have to make some guesses/assumptions.

I couldn't tell from the one video I saw what gave way first, but it looked like the column. Once the column went down, the whole pier went down, and obviously the truss followed. These bridge spans are designed to be continuous across all of the piers, which is pretty typical. You pretty much have to do that in order to reduce the magnitude of your bending forces by distributing where in the truss they occur, otherwise your bending forces would be too large for the structure to resist. So in order to span over the navigable part of the river, they HAD to design it with multiple continuous spans connected to each other. So you can see when it starts to collapse, the two spans resting on that pier begin to fall, and the span resting on the pier that wasn't hit starts to get pulled up, causing a lot of additional stress in the members that they were never designed for. Once those break, the bending forces redistribute to the span that stayed up. But since you no longer have the rest of the bridge connected to it, it puts a lot more force in the structure than it's supposed to have towards midspan (plus the fact that it's horribly unbalanced with part of the structure cantilevering off the pier). So once those forces exceed what that portion of the bridge was designed for, cue the domino effect. There may have also been hinges put into the spans to force a point where the bending forces were released in the structure, but those are less common on trusses than other long continuous spans, so I'm not sure. I don't know enough about the bridge. But if one was there, that would also have contributed to the rest of the truss spans collapsing sure to becoming unstable. There should be a joint separating the truss spans from the rest of the bridge, hence why those approach spans were left entirely intact despite the entire truss going down.

As for whether or not the columns should have been able to take the hit, that's a complicated question. Current code requires you to consider vessel collision when spanning over navigable channels, but the actual load you have to design for isn't well defined. Some states institute their own requirements, and for some states it's up to the engineer's discretion, and I don't know MD's rules there. This is further complicated by the fact that:

a) Vessel collision wasn't a code requirement until around 15-20 years ago, so it wouldn't have been part of the code at the time the bridge was built.
b) If they did attempt to design for it, information on collision at that time would have been very lacking in terms of its magnitude and how to apply it, plus the analysis tools are nowhere near as sophisticated as what we have now.
c) Even if they did design for it and did everything correctly, the size of ships have increased dramatically since the bridge was built. So even if it was properly designed for vessels of the time, that load is simply no longer valid.

One thing I'm curious about, because I couldn't tell, is if there was any kind of fender system. Those are typically required for bridges over navigable channels and are relatively easy retrofits. They both serve as a way to guide the ships along a defined path and redirect the ship. There are a lot of arguments about their efficacy and if they actually redirect the ship enough (not a whole ton of data on that), and you have to offset it from the columns a huge distance because of how much they deflect. Plus since they don't go the full length of the channel it's still possible to get behind them and hit the pier. But if there isn't any kind of fender system, I imagine that's what the feds are going to hone in on.