Wednesday, November 19, 2008

A wobbly bridge - part 6

What a difference a week makes! Lots of information has come out in the NTSB board meeting last Thursday and Friday. Also I attended a presentation given by Prof. Roberto Ballarini tonight at the University of Minnesota.

Four thoughts now -

  1. It's clear the U-10 gusset plates failed due to the load. The gusset plates should have been thicker. As it was, the safety factor was only 1. (That's the ratio of the load that can be supported to the load that is actually being supported.) It should have been 2. The bridge had been sitting on the verge of failure for years.

  2. I asked Prof. Ballarini a question at the end of his talk. I mentioned the reports that the bridge had been wobbling in the days leading up to the collapse. I asked if such wobbling could have been an indication of imminent collapse. He thought that was a good question. From his response, it sounds like there could indeed have been wobbling in the days before the collapse, related to the loss of elasticity in the gusset plates.

    As I wrote August 11, 2007, "When a bridge starts wobbling, best not to ignore it." Get people off wobblers.

  3. Someone has lied about whether or not there were sparks flying between the work crew and MNDOT about mixing concrete on the bridge, either the people involved or the short grapevine. The stories are at odds.

  4. Concrete was being mixed on the bridge and had been for days. Heavy ingredients, sand and gravel for making concrete, were in piles on the road above the U-10 gusset plates. This concentrated load was the straw that broke the camel's back, or perhaps the lead pipe.
Version 1.2

Sunday, November 16, 2008

A wobbly bridge - part 5

The NTSB hearings took place this Thursday and Friday. I watched online on Thursday. In the final hour that first day, Acting Chairman Mark Rosenker asked twice what was different that day, the day the 35W bridge collapsed. Three people mentioned the load. No one mentioned the motion.

I wrote an email to one of the board members with the question,

Was the mobile concrete mixer mixing concrete ON the bridge instead of OFF the bridge, in defiance of MNDOT?
The board member emailed me back that he had asked the question right after lunch on Friday, to which they had replied that
there was no concrete mixing at all going on at the time or immediately before the collapse. [emphasis added]
I am grateful that the question was asked.

Questions though remain. In response to an editorial today in the Minneapolis Star Tribune Lessons from a doomed bridge, I asked,
In the days leading up to the collapse, did the bridge wobble unusually and increasingly?

In the days leading up to the collapse, did the mobile concrete mixer mix concrete on the bridge instead of off the bridge?

Was mixing concrete on the bridge forbidden by MnDOT?

Did a MnDOT employee tell the crew not to mix concrete on the bridge in the days before the collapse?

Did the crew use a low slump concrete which must be laid and screeded within 15 minutes of being mixed? If so, did this prompt them to mix it on the bridge to save time?

Was the U10 gusset plate underneath the mobile concrete mixer at the time of the collapse?

Was the U10 gusset plate weakened by the repeated motion of the mixing, if such mixing took place?

Did the truck next to the mobile concrete mixer hold unmixed ingredients for concrete?
Why do I keep asking? It must be the influence of my godfather, Jim MacInnis, who was a lawyer in San Francisco.

Version 1.0.1

Update (Nov 16, 2008, 5:09 pm Central): More questions, simply about the load -
What was the total load on the bridge that day?

How much of it was piled or parked above the U-10 gusset plate?

What percentage of that load would have been safely off the bridge and not concentrated at that point, balanced on or off center, had the crew mixed the concrete off the bridge instead of on the bridge, if indeed they did mix it on the bridge?
According to the New York Times, NTSB reported in March that the total load on the bridge was 630 tons that day.

MPR reports,
On the day of the collapse, 270 tons of construction equipment and material sat piled just above the bridge's weakest spot. [emphasis added]
The Minneapolis Star Tribune reported that the NTSB released documents in March noting that more than 192 tons of sand and gravel had been loaded onto the bridge. Is that how much there was at the time of the collapse?

Was the sand and gravel used for making concrete?

Update (Nov 16, 2008, 8:10 pm Central):
Here's a presentation to NTSB which has some answers -

Construction Activity & Traffic Conditions on I-35W Bridge
by Robert Accetta.

Total load (over the entire bridge) - 630 tons
Construction materials and equipment load (near U-10) - 289 tons
Construction materials load (near U-10) - 192 tons

The piles of sand and gravel were directly above the U-10 gusset plate. It's still unclear exactly where the mobile concrete mixer was. Did it pull up next to the piles in the second lane?

Update (Nov 16, 2008, 8:40 pm Central):
Here is another of the presentations -

Bridge Description and Collapse
by Jim Wildey

It includes an image depicting the initial failure of the U-10 gusset plate. It appears both the west and east plates failed simultaneously.