Under design? Is that an isolate case (by one design firm and state DOT), or it is a standard practice/detail adopt by many states/institutions? Any comments?
I have long list of questions for NTSB to single out the "undersign" as the sole cause after it (the bridge) has been there for a long while. Hello, bridge designers?
It would be interesting to see the report. Looking at the collapse video at the time, the failure seemed to be quite symetrical. That is one of the things I do not understand. If the gusset plate 'went' i would have expected a non symetrical failure
All we can do is look at the actual report before making any engineering opinions. A Yahoo report is really not proven to be accurate and the headline may not be accurate.
The collapse seemed to be somewhat symetrical, but the other elements may have tied things together in the short term until the similar conditions/weak pints were triggered.
The preliminary reports issued earlier this year by the NTSB identified these same gusset as being underdesigned. It was not just one gusset, but rather at the same location on each side of each of the trusses. The plates were shown 1/2" thick, when they should have been considerably thicker. No design calculations were found for these gussets. Likely a drafting error, not found in the checking process.
A successful design of the gusset plate could be made using 1/2" thick plates, if the other dimensions were correctly proportioned. (Thickness is only one part of steel area calculation.)
I can't believe that actual gusset thicknesses were not checked over the years especially with inspections and URS report using LSD. Again as noted elsewhere, the bearings may be the culprit.
In my reading of the URS report, I could find no mention of checking the gussets. I could have missed it, as the report is quite long.
In this bridge, to memory the gusset plate thicknesses varied from 1/2" to about 1 1/2". The ones which failed were some of the most heavily loaded, and were 1/2".
Maybe _under_designed, maybe _not_ designed, since nobody could produce original calculations for the failed gusset plates. It just sort of smells like, "Hey, Junior, just make it like the other one."
... Which is not nearly as scary as the fact that the gussets buckled enough to notice, and enough to cause people to photograph them in 1999, and then they were left in service.
Mike - According to the report, the plates were not photographed for the buckling but the general condition (further noting that the buckling wasn't noticeable) and for the attachment of strain gauges on some of the plates.
As for the conclusion that this bridge fell due to underdesigned plates, I can only note that this was certainly a problem early on but thanks to many things done right in the design, it functioned and as steel should do, it gave warnings. The NTSB says that based on their interpretation of the inspection and rating codes no one had any reason to suspect a problem. It's this last sentence that I have a very difficult time with.
No engineer bridge or structural is going to wave off plate distortion like that especially when it coincidentally occurs at four of the same panel points through out the bridge.
It's much easier to blame a dead engineer and a defunct firm that to own up to poor maintenance. And what's worse is that our entire infrastructure is poorly maintained due to lack of funding....it's no secret and since 8/1/07 it shouldn't be shoved under the rug either.
Off my soapbox.
Regards,
Qshake
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
I was going to respond to this post, but I procrastinated and Qshake beat me to it. You have a bridge that stood up for 40+ years and they say the design was incorrect!
The way I understand it, there were new surfaces applied to the bridge, materials stored on the cantilever span and heavy equipment on an active span. To save a few bucks, the bridge loading was changed and no re-analysis was done.
It's a chicken bleep excuse to blame a 40 year old design, when you have whole state department of transportations employed to insure the road and bridges integrity.
Shortcomings in maintenance will never be completely eliminated, and I imagine that budget restraints will make the problem worse over time.
But if the gussets in this bridge were thick enough, we would not be having this discussion. Read all the reports and you will agree. Human error will never be completely eliminated either, but that should be our goal.
I would like to point out a few things with regards to this thread:
1) We use load factors for a very good reason - ultimately, we can't know exactly how much load a bridge will see, or how it will distribute. So we bump things up considerably, and design for that load.
2) Temperatures in Minneapolis that day were in the mid-to-upper 90s, which could easily produce higher loads. I wouldn't be surprised if temperature load had something to do with it, whether due to fatigue or some other loading.
3) I have "checked" structures that were designed according to older codes using newer codes and found them deficient. Yet, they're still standing.
It is conceivable, then, that a gusset plate could be underdesigned yet still support a bridge for 40 years.
![[pipe]](/data/assets/smilies/pipe.gif "[pipe] [pipe]")
