Juggy done a bridgedozy

[buzby]

Sorry for the bump, but I want to respond to posts in this thread.

The timeline that I understood from my half watch was that some of trusses were put under tension for the move. These then had the tension released and it's then that the cracks were seen. The decision was made to put the trusses back under tension, to put it back in the state it was in before the cracks had appeared. Since they were "resetting" back to a previous state it was thought this was safe and didn't need a review process (or closing the road, presumably). This was wrong and re-tensioning the trusses caused the cracked areas to break completely.

I read the full NTSB report this morning, coincidentally.
Full report (in exhaustive detail)
Abridged (ahem) version

The cracks at the 11/12 node first became visible as soon as the formwork was removed after the upper deck had been cast and the bridge's dead weight was being borne by the ends of the deck resting on the construction abutments at each end for the first time (the construction crew heard a loud cracking sound and all work as stopped while this was investigated). The tensioning of the rods in the two end diagonals then took place ready for the lifting and transport operation, and no significant growth in the cracks was visible (this is most likely because the construction abutment at the 11/12 end of the deck was restraining any movement of the 11/12 node joint).

After the bridge was lifted and placed into position, there was no restraint for the 11/12 end of the deck on the intermediate pier (when the bridge was complete, the shorter deck section over the canal would have acted as some form of restraint, transmitting the load through to the pier at the other side of the canal) and the deck rested on the pier via 4 plastic bearing shims placed under the outer deck sections, with no support under the central truss members. The transport tension rods were then detensioned, and the cracks started to grow. The initial assessment by the designers was to place a fifth bearing shim under the centreline of the deck so that the area that was cracking was no longer suspended in midair. This was done but the cracks continued to grow. It was then that the design engineer decided to retension the rods in member 11 to 'put it back to the last known-good conditon', but this was totally flawed as the 11/12 joint had started to fail as soon as the shoring had been removed after construction and before member 11 had been tensioned for the transport operation.

The retensioning operation added some vertical clamping force at the joint between member 11 and the deck, but the horizontal component from the tensioning added to to the existing force that was pushing the 11/12 node out of the end of the deck, and caused the final failure.

The NTSB's findings were:

1) Figg, the designers, had a) designed a non-redundant structure (a failure of any one of the truss members would cause the entire deck to fail) but had used the safety margin figures for a redundant structure when sizing the truss members, and b) used the wrong figure in their calculations for the amount of restraining force steel reinforcement added across a cold joint (a joint between set concrete and fresh concrete), and as such had vastly underestimated the amount of reinforcement and dimensions of the joints required between the truss members and the deck (the 11/12 node was only the second-highest stressed joint - the other end of the deck had even higher stresses, but was being restrained by it's pier).

2) Louis Berger Group, the peer review engineers contracted by Figg to check their calculations, were not qualified to perform a review on a novel design of this type (nobody had built a concrete truss span of this type before - as it turns out, for good reason). Figg had not asked to see their state-mandated accreditation before giving them the work (unsurprisingly, the peer review contract was decided by the lowest bid). Also, their contracted review was only for the 100% completed structure (where the shorter span over the canal would be in place), not for the bridge's intermediate stages of construction.

3) The cold joint areas on the deck that would interface with the truss members when they were poured were not roughened as per the US concrete construction specifications, so there was no additional frictional force in the joints to help the steel reinforcement restrain any movement between the diagonal members and the deck. The NTSB noted, however, that even if the member 11/deck joint face had been roughened, their analysis showed it would still have not provided sufficient additional restraint for the undersized and under-reinforced joint.

4) When the cracks became too big to ignore, Figg took the decision to retension member 11 unilaterally without any peer review, under the assumption that it was being returned to it's previous 'transport' condition. This was a wrong assumption, as by that time the 11/12 node had already failed,  the steel reinforcement in the joint had already undergone post-elastic deformation and so it could never be returned to it's previous state.

5) Despite Figg's reassurances that the cracks were not safety critical, there were other contractors, the project managment company and even the customer who had seen the pictures of the cracks or had been onsite to inspect them who could have raised the alarm and had the site evacuated and road closed, but nobody did (if you could fit a ruler into the cracks, then they were obviously not due to post-cure shrinkage or thermal expansion - the only acceptable reasons for a prestressed reinforced concrete structure to crack).

[Alberon]

Great post as always, Buzby.

Now when are the fuckheads responsible for this going to prison?

[buzby]

Great post as always, Buzby.

Now when are the fuckheads responsible for this going to prison?

Unlikely, unfortunately. About the only consequence so far is that the U.S. Highways Agency have disbarred Figg and W. Denney Pate, their engineer for this project, from working on any federally-funded bridge projects for 10 years, which Figg have been fighting against in the courts. They are still able to bid for state-funded projects, if anybody feels like taking a chance with them after this. The main construction contractor, Munilla Construction Management went bankrupt as a result of settling the lawsuits with the victims families. The Louis Berger Group were the only defendant firm named in the lawsuits who have refused to contribute to the settlement fund, instead paying off 3 victims and taking the other cases to court (MCM's afministrators were also suing Louis Berger for negligence).

One interesting outcome of the NTSB investigation is that they no longer consider OSHA (the US equivalent our HSE) a 'friend;y' agency and will no longer allow them to access material and evidence the NTSB has gathered for it's own investigations. This is because OSHA's own report into the collapse included material from the then-unpublished draft of the NTSB's report (which had been sent to OSHA for comments) that they had reproduced and published without consulting the NTSB.