1 Yikes.  It just folded up.

Some reports of fatalities now, but too early to know the real number.  Hoping it stays low.  There was a bus full of kids on the bridge when it collapsed, but it sounds like they all escaped okay.

2

I lived in Minnesota for a couple of years... I've driven across this bridge dozens of times.  I'm quite shaken by this, oddly enough.  I've put a post about it at The Pond, but it doesn't really say anything you can't find anywhere else. 

Except that Ph.Duck's family, who lives in and around the Twin Cities, are all okay.

3

Wonderduck, I'm glad your family is OK.

We probably won't get any real coherent reports of what happened until daylight tomorrow. I've seen reports of 5 dead, but I'm sure it's more than that. What in hell failed? That sucker stood for 40 years.

4 Oh, and by morning someone will have figured out a way to blame this on Bush and the war, if they haven't already.

5 One expert on Fox News mentioned scouring damage (soil erosion under the footing?) as a possible cause.

6 Well, I've heard they already are blaming it on Jesse Ventura.

As for the cause, the only thing that seems excluded is terrorism, as really, something this big?  People would have noticed the explosions.  I hope.

7

This guy lives in a warehouse apartment that's a mere handful of yards from the bridge... he SAW the darn thing collapse from his rooftop, and says there weren't any bombs or explosions.

They're saying seven dead now.  Ph.Duck gave a reason for the collapse that sounds quite possible: the construction work closed the left-hand lanes going in both directions, thereby forcing all the traffic to the outside lanes.  From family member reports, it's been bumper-to-bumper and really slow (as in 'walking would be faster' slow) for months now.

Perhaps that constant level of traffic, grinding away at the outside of the bridge, mixed with the construction work AND the train going underneath just made something go *sproing*.

I'm not an engineer of any type, so I really can't judge, but it SOUNDS plausible, at least.  Of course, considering what we know at the moment, metal-eating beavers making a dam sounds plausible, too.

8 Read a report over at Captain's Quarters that the cell phone system in the Twin Cities is getting swamped. If you're trying to reach someone there by cell, I strongly suggest using text messaging. Much less bandwidth & the message can wait for an open patch of network to transmit. (Handy trick I picked up after Katrina...)

9

Wonderduck, that better not be the case. Civil engineers try to incorporate a 3- to 5-fold overdesign for safety reasons. By which I mean, they figure out what the anticipated maximum load will be, and then design for 3-5 times that much. If the bridge failed because it was overloaded, then someone screwed up bigtime.

Do they salt the roads in Minnesota in the winter? That's what I'm wondering about. There was a bridge near where I used to live in MA where saltwater had really corroded the concrete badly. If you walked under that bridge and saw what had happened to it, you'd never feel comfortable driving over it again.

The bridge that went down relied on four specific concrete pillars, and the recent "before the collapse" pictures showed rust stains. It wasn't a closeup shot, so I couldn't see it in particular, but I would think that one of the leading candidates for a cause would be concrete failure.

But it's really pretty pointless to speculate at this time. People who know what they're doing are going to be all over the wreckage during the next week, and they'll figure it out.

10

Do they salt the roads in Minnesota in the winter?

They do, though some places (like where I lived) use it and sand both.  Really, the trick is to actually find the road first.  My first snowfall in MN was the Halloween Storm of 1991... 27.5" of snow in 12 hours, then 50mph winds the next day.  We actually had blizzard conditions without it snowing at all then, as the wind just picked up the snow and blew it around.

The whole town shut down; the only places open were the grocery stores, the video stores, and the liquor stores, and they were out of everything (yes, all three categories).

I knew it was bad when I helped push a snowplow out of a snowdrift.  Salt or sand, neither would have made much of a difference.

I'm almost positive they salt the bridges in the Twin Cities, though.

11 It says on Fox News that it was the first bridge of its size to get a computerized anti-icing system in 2001.

12

For a disaster of this magnitude, there have to have been several failures, not just one. When we get to the bottom of it, we will likely find out that there were problems with corrosion, and botched investigations, and ignored warning signs, and perhaps the anti-icing system.

This is of more than academic interest to us in Houston, as we're having issues with the constant rains we've had for several months now. Several major bridges were already being monitored due to problems with scour erosion.  Ironically, this story was posted to the Chronicle's website only hours before the disaster.

13 Slideshow: "Before" pictures

Adding a little data to the speculation, the main span was an arched truss, supported on four piers (you can see the two on the north end in the first photo above). Trusses are a good method of spanning a long distance (so no mid-river pier, unlike the adjacent-- downstream-- bridge). Problem is that this design has no redundancy: It's two trusses (tied together), each supported at two points, like the four legs of a table. If you lose a support, it WILL come down. And because it's a truss, a failure at a joint will shift loads through the remaining structure. If the result exceeds physical limits, hello cascading failure.

Another possibility is shifting of one pier, which would impart a twisting stress on the whole structure. This would be hard to detect until something deformed/failed.

Re: Ice removal, a while back the bridge was retrofitted with a chemical-spray ice control system. (The bridge was subject to icing problems because of water vapor from the adjacent University of Minnesota steam plant.) Can't find the reference now, but read the chemical was ?something?-phenol, supposedly less corrosive than *Cl, but still corrosive.

My "eyeball assessment" is that they designed the bridge using "lots" of "little" steel, rather than fewer "big" members (probably for cost and time-to-obtain reasons). That works, but it means a lot more joints, and some of those connections are complicated to assemble and difficult to inspect afterward. Certainly there's nothing "elegant" in the design, it even lacks the brutal beauty of classic railroad construction.

14

The fundamental problem with that approach is that if it fails, it doesn't fail gracefully. Of course, the design engineers will respond, "We will design it so it doesn't fail, so that doesn't matter."

And that's not really wrong. We have to keep in mind that the bridge stood for 40 years before it went down. This isn't like Galloping Gertie, which fell 4 months after it was built.

I don't think that the post mortem will decide that the design was faulty. I think it's going to end up being a problem of maintenance, most likely due to corrosion.

There's a video of the center span going down, taken with a 1-frame-per-second security camera. The whole center span fell simultaneously; it's not the case that one end went down first. I'm having a bit of a hard time understanding that. It will be very interesting to see the analysis of the failure cascade.

(DAMN I hate cascading failures. "For the want of a nail...")

15

The design of the bridge next to it (which I gather was built later) is not only much nicer looking, it's also far more robust. Each of the spans is supported by two independent arches, so even if one arch fails, the other can at least partially support the span. And I suspect that those arches wouldn't cascade linearly, so that even if one span did go down, the other would would remain standing.

The arches are supported near the bottom, so the design doesn't rely on the foundations being stiff and resistant to lateral force to anything like the same extent. All in all a much more robust design.

The truss design of the bridge that did go down relied pretty heavily on tensile strength of the truss. That's why, when the span over the river failed, the other parts went down too.

16 "...which I gather was built later..."

Actually earlier. It's the 10th Street bridge, and this photo taken during the 35W construction shows it proudly standing.

17 Lileks has a link to a video segment from just after the accident, in which the anchor reads from the most recent inspection report (some *damn* fast work laying hands on that!) to a safety engineer.  I can't recall the exact phrase, but it involved a "medium-low" risk due to some cracking in girders and the "diaphragm" plus a mention of "out of plane" stresses.  The engineer seemed to think that was significant.  Linked here. From their explanation, I gathered a diaphragm is an assembly of angled girders, perhaps those that make up the "floor" of the arch.

And I agree, the security film looks weird, now that you mention it.  The central span just dropped straight down, I'm sure it will make sense once the cause is determined.

18 Sorry for double post, but I wanted to correct myself: and add an observation:  it doesn't drop straight down; the near end drops down first.  My observations were getting too long to abuse the comments with here so they're at Houblog  (or will be in a minute).