Working Theory On Bridge Collapse Is Just That

The NTSB’s working theory on the St. Anthony Bridge collapse involves design flaws and overloading, according to comments by Transportation Secretary Mary Peters. That prompted questions by two state legislators about the role of maintenance and whether a lack of it didn’t also have some role to play in the collapse, but Peters said that the legislators have misinterpreted her remarks (via Mitch):

The top federal transportation official said that investigators have a “working theory” of why the 35W bridge collapsed in August: a poorly designed metal component called a gusset plate and excessive weight on the bridge that day.
U.S. Transportation Secretary Mary Peters’ comments Thursday mirrored statements she made in August, a week after the collapse, and like her previous comments immediately led to controversy. The National Transportation Safety Board, which is investigating the collapse, has said a formal finding will not be available for at least a year.
Sen. Steve Murphy, chairman of the Senate Transportation Committee, said Peters told a gathering Thursday in Washington, D.C., that he attended that “a finding of fault was not going to be lack of inspection or lack of maintenance” by state officials.
“I think it taints the findings,” he said.

The Senator jumped to a conclusion, although Peters may have inadvertently pushed him along. In any investigation of this kind, the team gathers evidence and starts testing hypotheses to see whether the evidence supports them or rules them out. A “working theory” means that the investigators have settled on one for the moment that fits the evidence seen thus far. It does not rule out that other theories will be tested, perhaps either replacing the working theory or added to it.
The NTSB says it still hasn’t ruled out any hypothesis at the moment, including maintenance or de-icing fluid issues. However, it sounds as if Senator Murphy has; any result that rules out maintenance and testing will “taint the findings”. He wants a finding of some form of incompetence with which he can beat the Pawlenty administration. A poor gusset-plate design won’t allow him to do that.
However, gusset plates seem a likely cause of the collapse. A similarly designed bridge in Cleveland had a partial collapse in 1990, and the gusset plates failed in that incident. As it turned out, the plate design was insufficient for the size and weight of the bridge. Given that history, using gusset-plate design as a starting point doesn’t seem unreasonable.
It will take several more months before the NTSB can make its decision on the cause of the collapse. Peters should probably keep her updates less specific until the investigators have their report ready. State politicians should keep their mouths shut as well, lest their desires to warp the process for their own political ends get too obvious to Minnesotans.

14 thoughts on “Working Theory On Bridge Collapse Is Just That”

  1. What? State Senator Steve Murphy shooting off his mouth prematurely? LOL! Sen. Murphy is one of the top five most devisive politicians in the State of Minnesota. This guy would criticize the Pope’s robes if He came to town.
    One can only hope that at the next election, the people of Red Wing will vote to take Steve Murphy out of the State Senate and put him in some serious therapy.

  2. Why the Minnesota Bridge Collapsed

    The NTSB is thinking design flaws and overloading as to why the bridge went down. But the agency hasn’t ruled other causes. Ed Morrissey is speculating that faulty gusset plates make sense at least part of the reason that the bridge collapsed. vadkinsQ…

  3. I liked the NTSB better in the old days, when they would take two years to do a thorough investigation, say squat to the press during that time, then issue a definitive report.

  4. Considering two bridges had issues with gusset plates , one verified and one suspect, I hope any bridge which incorporates them into designs would be considered for review under state of the art knowledge of new engineering data which may not have been known at the time of the original construction.
    Gusset plate replacement with more structurally robust plates would seem prudent especially in cases of aged or weathered conditions. It is much less expensive to do preemptive replacement or enhancement of existing plates than to suffer the loss of a bridge due to their failure.
    Even replacement can be avoided with a suitable doubler plate to add structural integrity to the bridge.
    The procedure is common in maintenance of ships of commercial ocean going size over their life cycle to counter areas of stressing noted during hull inspections in maintenance yard periods.

  5. IIRC, the bridge was down one lane for maintenance/construction work that day? How could excessive weight occur, unless they are considering unbalanced loading as excessive…or perhaps it’s just poor journalism?

  6. I understand the issues with the material staged on the bridge for work in progress, but one thing that struck me from the photos of the aftermath is there was a low number of SUV type vehicles in the photos.
    I would have expected to see more due to the adverse continuations for winter driving in the area for which SUV’s can provide better capabilities.
    Mostly I saw medium to small vehicles which would lower the total loading on the bridge compared to the weight of an average vehicle mix 15 or 20 years ago.

  7. Frank, I’ve been wondering that too. Add to that the fact that some of the decking had been removed in the lanes under repair, and it seems like the load on the bridge should have been considerably less than a normal traffic load, even with some heavy construction vehicles present. (After all, presumably similar construction vehicles were driven back and forth over the bridge in normal traffic all the time.) The only possibility I see there is a severely abnormal weight distribution depending on which traffic lane was open and where the construction vehicles were placed, but that seems like a rather gimmicky theory.
    daytrader, there’s a bridge of that design in Alabama. The state is in the process of installing additional temporary supports until the NTSB determines the cause of the Minnesota collapse. Depending on the outcome, the state may decide to modify the bridge or replace it. Everyone involved acknowledes that something will have to be done, and that it won’t be cheap.

  8. Recall that a train was also passing under one end of the bridge at the time it fell. It may have been just the combination of factors–a perfect storm of stress, imbalance, vibration (from train and construction,) with a few corrosion-weakened poorly-designed gussets. The “rogue wave.”

  9. Since we are all engaging in pure speculation, it is possible that the asymmetrical loadings resulting from bridge construction, materials storaged an the like may have been a factor.
    I expect that the NTSB will come up with the right answer. Given the state of the bridge engineering art, this shouldn’t be a particularly difficult structure to evaluate.

  10. Removal of the decking may have contributed to the collapse. The design combines elements of the arch and the truss. Instead of unbroken arch members, it has a series of segments which can become unstable if the mistmatch between load and geometry is greater than what the truss members can absorb. (It’s more complicated than that, and the mechanisms depend critically on the arrangement of the diagonals, among other things.) If the bridge is stable with the deck load in place, it may lose stability with the deck load removed. (Remember that stone arches depend on the load above for stability.)

  11. nj, I thought about that after I posted previously, although my thought was a bit different: Does the bridge depend on the arch of the top surface for some of its overall stiffness? If so, did removal of the decking make the structure more prone to sagging in that area?
    RBMN, I’d forgotten about the train. Not sure how it plays in, though… I wonder what kind of soil conditions exist there. Soil structure is one thing I don’t know much about.

  12. If loading on the bridge caused the collapse (which I doubt) then one would have to point the finger of blame on MnDot. Would MnDot let a contractor work on the repair of the bridge without their supervision? If so, on a bridge with the problems documented on the I-35, no supervision would be negligent.
    Also, I don’t think it is a good idea to bias thinking with a “working theory” of why the bridge collaped. It is good to go in with an open mind.
    I used to do work for a company that used micron microscopes to analysis metal fatigue. I suspect the reasons for any gusset plate or beam fatigue will be determined by this method.
    Then the reasons for the fatigue will be investigated. And on this bridge, there no doubt could be many reasons, such as oxidation of the metal, severe temperature changes, expansion joints not functioning causing stress; amoung some of the reasons.
    But it is said that a picture is worth a thousand words. Did the bridge look like it was suffering from a lack of maintenance? What did the inspection reports reveal?
    I am sure when the investigation is completed and the truth comes out, it will be a tough pill to swallow.
    People want to know why a bridge fell down, when it was supposed to be inspected and maintained for the public safety. That is the bottom line.
    Coming out with news that the investigators have a “working theory” that there is a design flaw with the gusset plates and overloading of the bridge during construction, is putting a spin on the subject, that is not responsible. People should keep their mouths shut until the investigation is completed.

  13. you guys are right about a closed mouth gathering no foot.
    in earthquake country we see buildings with large and heavy internal bridge crane structures that have cross bracing between the columns twist the gussett plates to some degree.
    for certain types of construction these things are absolutely necessary but i have never seen them with cross ribs to stiffen them.

  14. “I wonder what kind of soil conditions exist there. Soil structure is one thing I don’t know much about.
    The soil is mostly sandstone, limestone, gravel, and clay, which is what you find over most of central Minnesota. It is not very stable and is prone to severe erosion. This is why, for example, Saint Anthony Falls, was covered with a concrete overflow spillway over a hundred years ago to stop the falls from retreating upstream due to erosion. Also, there are several old water tunnels dug into the riverbed and surrounding soil just upstream of the site. These were built in the 1800’s and early 1900’s to provide as a source of power to run the sawmills, textile mills, and flour mills that were scattered throughout the area.
    A lot of these tunnels, unused for several decades and with little or no water flow, were reopened and the water flow restored just a few years ago to help clean up the stagnant water problems (pollution and a hell of a lot of mosquitoes!) associated with them. I’ve been wondering just how this affected the underground water flows in the area and if that contributed to undermining the soil downstream of the tunnels.
    Limestone and sandstone don’t react well to increased water flow (even underground) as it erodes that type of soil very easily and the clay here tends to expand when it absorbs water. I wouldn’t be surprised if excess erosion of the underlying soil was a contributing factor in the collapse. Could it be possible that our good intentions of cleaning up that portion Mississippi river and surrounding areas actually lead to the collapse? Stranger things have happened.

Comments are closed.