Freeway Bridge Collapse in Twin Cities

[davidbeach]

http://www.npr.org/templates/story/story.php?storyId=12436334

MINNEAPOLIS August 1, 2007, 8:20 p.m. ET · The entire span of a four-lane interstate bridge over the Mississippi River collapsed during evening rush hour Wednesday...

 

[wiktor]

It is a very sad accident, unfortunately looking like some kind of maintenance/design issue. Relying on the flanking spans to bear the thrust of the side arches lead to the progressive collapse and may even trigger the entire event.

 

[mshimko]

Can anyone provide some info, or point me to a decent source, to better understand the terminology and sound-bites we're all hearing about the bridge collapse?

I've heard "structurally deficient", and "only rated 55 on a scale of ___", etc. All I know for sure is "structurally deficient" means less than perfect, but beyond that I know little, and want to avoid myself (and others) from jumping to conclusions.

 

[SlideRuleEra]

mshimko - "Engineering News Record" is the most likely source of sound, current technical information on this subject. Here is a link

http://enr.construction.com/Default.asp

http://www.SlideRuleEra.net

 

[strguy11]

Check the FHWA website under the bridge section. They have posted the Coding guidelines. I think in there there is a description of the sufficiency rating.

 

[LCruiser]

I wonder if it was an eccentric loading problem. The surface as milled on 4 of 8 lanes on the same side - outside in one direction, inside on the other. Apparently the first collapse left the superstructure 50 feet offcenter. That would explain that...

 

[LCruiser]

Also, the eccentric loading would be exacerbated by rush hour traffic...

 

[bridgebuster]

Here we go speculating again .

Two lanes in each direction were closed and on some newscasts there was talk about the unbalanced traffic loading. I doubt that this caused the collapse. For example, if just the two outer lanes in one direction were fully loaded (assuming AASHTO lane load)there would be a LIVE LOAD uplift on the opposite truss. However, the dead load would counter the uplift.

From the aerial pictures it wasn't bumper-to-bumper trucks.

 

[strguy11]

I also doubt the unbalanced loading had anything to do with it. I am sure this was not the first time work was done to this bridge, or that lanes were closed in its 40-year service life.

 

[LCruiser]

I doubt it would have had an effect on a competent structure. I'm just saying that the fact the rush hour traffic loading was off to the same side as the construction unbalancing, and the initial collapse was apparently rotational and at the time of the maximum unbalancing seems like too much of a coincidence - the odds of it being random at that point in time were astronomical.

 

[northarm]

MN DOT has put their inspection reports online accessible from
their bridge collapse website:

http://www.dot.state.mn.us/i35wbridge/index.html

These reports give you a fairly good overview of the design of the bridge, and contain some pictures to illustrate the inspection findings.

Here's what I know.

1. The only video of the collapse I've seen is a surveillance video from the Mpls police. It was shot at a low frame rate, approx 1 per sec, and shows the west side of the center span of the truss (span 7) looking to the north. Pier 7 and span 8 are visible, but not pier 6 where the collapse may have started. Having viewed this video numerous times, it is apparent that the pier 6 end of the span started down first. This is the span that crosses the river. The truss separated right before pier 7, and the remaining truss remaind seated atop pier 7 for a second before span 8 broke in the middle and pulled the intact truss off the pier.

2. A woman interviewed on Channel 5 here, KSTP-TV, was proceeding south on span 7 at the time of the collapse, and stated that the left side of the span sagged in the pier 6 area before the entire span went down. Her account seemed to me clear and to the point.

3. While all other spans of the truss came down more or less in place, the roadway and debris in the pier 6 area
looks as if it fell sideways to the east. That tells me that something failed first on the east plane of the truss near pier 6. It appears that span 7 fell straight down into the river, even at the pier 6 end.

4. This is what I got from the various MNDOT inspection reports: The truss is anchored at pier 7, the north end of span 7 that crosses the river. Going south, there are roller bearings at piers 6 and 5, and a hinge bearing at the south end of the truss where it connects to the approaches.
The reports state that the pier 6 bearings are badly rust fouled, and jammed at the contraction limits, even in summer. Concurrently, the hinge bearing is jammed at its expansion limit. The pier 5 roller bearing is in good operating condition and well inside its operating range.

There is a deck seam over pier 6, and none over pier 5, and there is apparently a lot of runoff (we love salt in the winter) getting past the seam seal.

5. Pictures of pier 6 show the pier tops clearly. Both roller bearings seem to be gone, and both pier tops show fresh concrete gouges in the direction of the river (north), as if the truss had been shoved off the pier tops. The east pier shows a bigger gouge than the west pier.

Unfortunately, the steel from the east face of the truss at pier 6 is all under the collapsed roadway on the ground. The steel from the west truss face, including the bearing foot that was on top of pier 6, is in plain view and substantially intact.

6. My question for you bridge experts, after you have read the MNDOT reports, is: What's wrong with this picture? Why are the bearings of the truss from span 6 to its southern end locked in opposing directions? Is it possible that the truss has sagged, forcing its bearing points apart lengthwise? Is it possible for the essentially flat approach road to creep northward, shoving on the truss? What about span 7, with one end fixed and the other end on roller bearings locked in full contraction in the middle of summer? Does any of this matter on this sort of structure?

7. The resurfacing contractor was working on the pier 6 end of span 7 at the time of the collapse, with piles of concrete debris of unknown size and heavy materials trucks parked in the same area. There is no evidence of any debris pile now, but the pier 6 end of span 7 is under water. Unfortunately the public has seen no pictures of the top of the span shortly before the collapse; I hope these will be forthcoming - they might allow an estimate of any possible overloading.

 

[bridgebuster]

northarm - thanks for the detailed report.

The original drawings are available on the MN DOT website. My gut feeling is that the south end was the problem; the framing is a bit unusual.

Spans 3 through 5 are made up of continuous curved girders; the roadway is superelevated west to east (east being the high side.)

At the north end of Span 5 the girders frame into an end floorbeam at the beginning of the truss (Span 6). (At first glance the connection of the FB to truss (Sheet 19 of 94) is unusual. I have to look for more details) The roadway on Span 6 is also superelevated, transitioning to a normal crown at Pier 6/beginning of Span 7 (River Span).

All curved girder superstructure experiences uplift; the inner girders are the ones most affected.

enough said for now.