I-35W Bridge Collapse (updates)

by Michael Miller

Original article is here.

Added December 12, 2007: Professor Hassan Astaneh-Asl gives a very accessible explanation of the problem with truss-arch bridges in this article from UC-Berkeley, published August 3, 2007.

Added November 7, 2007: We just found two recent articles from the Star Tribune about the bridge collapse and added them to this page.

The first article (17 October 2007) is about how certain gusset plates might have failed due to rust and high temperature. That article contains a nice graphic showing the locations of those gusset plates (one error in that graphic is the reference to "L-10" instead of to "U-10"). According to a note in that graphic, the rust on the L-11 gusset was detected in 1993 and MNDoT noted it in their records at that time. If the rust on L-11 caused the bridge to collapse, that means MNDoT had 14 years to do something about it, but they failed to protect us. If the construction of L-11 is "a design flaw," it's a design flaw that they could have taken care of 14 years ago. The Star Tribune's copy of the article can be found here.

The second article (1 November 2007) is about U.S. Transportation Secretary Mary Peters' statements about a "working theory" of the bridge collapse. Isn't it unwise for a person in her position to make such statements more than a year before the official report will be released? Apparently, she was making similar statements a week after the bridge collapsed. Isn't Peters' "working theory" that the bridge collapsed because (A) MNDoT failed for 40 years to fix a "design flaw" that had been known for at least 14 years and (B) MNDoT failed to keep the weight burden off of the bridge when they knew it was vulnerable? Why didn't MNDoT detect failing gussets in an inspection before undertaking the deck work? The Star Tribune's copy of the article can be found here.

Added September 16, 2007: Messages from an engineer that were listed here earlier (September 14, 16) have been removed because they are not credible.

Added August 29, 2007: An acquaintance posted a link to my bridge web page on reddit.com. You can see the reddit.com comments here. I appreciate all the ideas and I will now add a few remarks to clarify some issues:

  1. If I were to write the page today, after learning quite a lot about bridges during the past three weeks, I would write it somewhat differently. The main difference would be that it has become clear to me that many errors were made in management of the bridge and quite a few things likely contributed to the collapse. A focus on any one problem is a bit misleading.
  2. Some people suggest that I am presenting a "conspiracy theory," but they are mistaken. I'm only pointing out one fact about the construction project. It is true that in the days following the collapse, state officials and the construction company claimed that the construction work was mere "resurfacing" which could not have contributed to a structural failure. I noted that on my page and showed this image from the NY Times. The state officials and construction company were inaccurate, or possibly inaccurately quoted, because, as the "construction gash" photo shows, they were doing much more than mere resurfacing.
  3. I pointed out that I am an epidemiology faculty member because I wanted it to be clear that I am a scientist, but I am not an engineer. I did not claim to know why the bridge collapsed. I have no expertise in steel structures or bridges. Does this mean that you "should not listen" to me? I think you should listen to this: All I am saying is that major work was being done on the bridge deck and the expansion joints, this is clearly visible in the photograph, and I can place the location of the construction work in the photo very precisely on the bridge. As to what this tells us, I am not sure because I am not an expert. Any claims I may have made about the cause of collapse should be taken with a grain of salt (which is why I stated my qualifications, or lack thereof, at the outset).
  4. Someone presented this claim: Concrete has almost no tensile strength and its use as bridge decking material is solely as a platform to drive upon. It has nothing at all to do with the structural integrity of the bridge itself. Any entry-level statics course would tell you that much, but apparently a handful of armchair reddit'ers prefer to take the "expertise" of an epidemiologist on this subject as prima facie fact. Unfortunately, the individual who wrote that [buford13] does not know what he is talking about. I spent several days with a preeminent expert (structural engineer, full professor, specialist in steel structures) and he told me something entirely different. Namely, the concrete has two layers: the top layer is two inches thick and it plays no structural role, but the lower layer is nine inches thick and it plays an important structural role which must be taken into account when construction work on the deck is undertaken. He couldn't have been clearer on this point. Another engineering professor told me that my web page contains important information that should be sent to NTSB, and I would have sent it to NTSB if I could have figured out how to do it (feel free to tell them about the page if you know how to contact them).
  5. What I wrote in Point 4 bears repeating: Construction work on the bridge deck that goes deeper than the first two inches of the surface layer is not "resurfacing" and it can endanger the structural integrity of the bridge. The deck is part of the structure. Don't believe this if you don't want to, but you can read similar statements in the NY Times and I'm telling you that I was told this very clearly by an absolutely rock-solid expert structural engineering professor.
  6. Again, I cannot tell you what caused the bridge to fail, but I can tell you that work on the concrete of the bridge deck, as shown in the "construction gash" photo, and in the photo added today (see item below), can play a role in a collapse. I don't know what the final story will be, but it is likely that the construction work played a role in some way. I'm not telling you this as an expert, but as someone who has read the words of several experts and can report on what they said.

Added August 29, 2007: I've added another pre-collapse photo here with analysis.

Added August 19, 2007: This is the original MNDOT traffic video from the time of the collapse: collapse2mins.wmv.

Added August 14, 2007: Jim Pankow sent me this great email message to draw my attention to this MPR article about the bridge collapse [source].

Added August 13, 2007: This Star Tribune article shows that the possible impact of bridge resurfacing on the structural integrity of the bridge was not considered in planning the construction work [source].

Added August 13, 2007: My colleague Jim Pankow noted the mile marker at 18.5 miles in the construction photo. Combining that information with his observation of still-standing mile markers at 18.4 and 18.3 miles (both next to WBOB) he can show the location of the construction gash in a second way that corroborates what I had found using the vanishing-point method and the location of Elliott Hall. In addition, all three of those markers are visible in the Microsoft aerial view of the intact bridge which makes it obvious that the construction gash is just where I said it was.