I've been watching the media reports about this, about the big dig collapse in Boston, the aftermath of Hurricane Katrina, and all the way back to the Hyatt Regency walkway collapse in '81.
Based on all of these, my opinion is, if its the least bit technical, and you heard in the mainstream media, they probably have it wrong.
I think I'm going to reserve judgement until the final report from some competent engineers is released.
I don't know the details of this bridge but here would typically only be a gusset on the other side of the chord, with nothing behind the diagonals' flanges or chord's webs, so that the gusset connections are in single shear.
Gussets are very susceptible to corrosion as moisture often collects at the joints in the debris (bird nests, etc). If they're undersized so that stress levels are high, particularly accounting for the fatigue stress range portion, then you have a reduced allowance for corrosion pitting/loss.
chichuck, just go to the link provided and within that pdf document there is another link that includes some of the calculations. This isn't a media report but an NTSB report by engineers.
When ever I go to a jobsite I can usually find a dozen or so structural problems. Often, I go within a day after a field person or an achitect has been there but yet they have not caught these issues! What happens if I don't go, is these problems get ignored.
Owners often don't want to pay for the service to have a structural engineer on site even if it is only periodic. They often claim that their field personal or the architect or a testing agency can do that job, which is often not true. There is nobody like the guy that did the design that is more familiar with all the aspects and what can and cannot be done.
Knowing how fatigue life goes down as stress level goes up and knowing that in some cases the stress was twice the allowable, it is hard to understand how the bridge remained standing for 40 years.
eureka - The matter of it standing for so long, IMHO, is attributed to the fact that truss joints are designed for an envelope of member forces. So unlike the truss joints in the old statics class, these joints may never even have equilibrium due to the envelope and based on the envelope concept you can see that it would be a magical day, indeed, if the same combination of loading actually occured on the bridge.
Another matter that I attribute the longevity to is ultimate strength. The NTSB did an allowable stress design to show how they reconstructed the analysis to the practice in 1967 and then show the results compared to other joints in the same bridge. however, the ultimate strength of the connection is not presented that I'm aware of.
Regards,
Qshake
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
It is sad to say, but anyone who even remotely had anything to do with this bridge will probably be dragged into this legal mess including the State DOT, inspection companies, construction companies, fabricators, engineering firms, etc. whether it was the original design, studies or subsequent renovation work.
As to why it collapsed 40 years later, the gussets may have been significantly undersized originally but as others have said corrosion and fatigue reduced their capacity over the years to the point that a "straw broke the camel's back". What the straw was is still up for speculation.
Unfortunately, it is much easier to prove design flaws than it is to prove any other cause.
Once again, the bottomless legal coffers will be used to prosecute the underfunded engineers.
It would be one thing if this was used as a lesson on why not to cut costs on engineering. But sadly, once the lawyers are involved, it becomes a blame game, and no lessons are learnt, only compensations paid and scapegoats found.
As a bridge inspector, I can say this. For a fracture critical bridge, such as this, the inspections must inspect each connection element (bolt, plate, weld) etc visually for signs of distress.
Either this was missed during the inspections, or it showed no signs of distress. If the plates were underdesigned from the beginning, I cant imagine that no signs of distress couldnt have been noted, since it stood for 40 years, and most likely saw more load than was on it that day it collapsed. Something doesnt make sense.
Slightly off-topic, but I just received an e'mail solicitation (by way of eng-tips.com) for Norwich University which uses the I-35W bridge collapse as the basis for it's pitch of a Masters of Engineering. Does anyone else find this to be in bad taste? I mean, to say that if people take their online courses, collapses would be prevented strikes me as crass. Below is the opening paragraph:
"Dear Structural Engineer,
Although catastrophic and heartbreaking, the August 2007 bridge collapse in Minnesota may not have been a surprise to many civil engineers. It was only a matter of time before a disastrous event such as this put our country’s aging infrastructure under intense scrutiny. Since this devastating event, the need for more experienced civil engineers, like you, has risen. Norwich University’s Master of Civil Engineering program has been designed to upgrade your skills and help you take control of similar challenges."
The bridge had been modified several times in the past 10 or 20 years.
As I recall, there were some wider walkways and new crashrails that could have increased the dead loads.
It was never a "bouncy" bridge in the times when I was stopped on it and seemed to have a good reasonably good alignemnt through the years.
I suspect (without looking at the deatails, but looking up at it from a boat), that it will end up being a combination of thin gussets, age and unbalanced load conditions (during traffic changes and construction progress) when it all comes down to it.
1.) It will be the deck repair contractor who gets stuck with this - it is always the contractor who takes the fall. Thats my bet
2.) What really amazed me was that the bridge was designed in '67, and the comps can't be found (not reall shocking) The bridge was significantly modified in '98 and acording to the Feds the designers only checked a few critcal membersand if they passed, the modifcations were deemed safe. This appearently is viewed as standard practice and they are okay with this. A.) If we couldn't find the design comps and no one anayzed the strucure, how did anyone know what the critical members were? B.) Since '67 codes and allowable loads have changed significantly which may have made a non critcal member critcal. C. The last inspection report noted significant deteroration of the strucure. I find it hard to believe that there was no deterioration of the structure from 67 to 98 then rapid deterioration from '98 to 06. Certianaly the deteriorationmust have reduced capacity of some elements.
Appearently the feds recognize more should have been done and are urging DOT's to check gussets in future designs, but don't see this design practice as abnormal.
3.) Speaking of deterioration, the last inspection report should there were significant problems that needed to be addressed. The state opted to extend the life of the bridge 15 years with more frequent inspections. I,m sorry, but if you are ignoring the current reports, why bother with more reports?
The failure was not a tragedy, it was a governmental policy.
As I read the report, the Sverdrup calculations have been located and assessed, but they did not include calculations for the gusset plates in question. Therefore, it is not known whether or not the design intent was correctly shown on the drawings. To me, production of the drawings and checking thereof is most likely when the errors were made.
The truss member forces were shown in the Sverdrup calculations and on the drawings, and those forces agree with the analysis of NTSB consultants, so the original forces were used to check the gussets.
It is still unknown, or at least unreported, if these gussets were the elements that failed first. I think the early release of this section of the investigation is just to give us all something to chew on.
hokie66 - just to add to your last comment, I think the news conference was helpful in that common practice should really be to check all structural elements when modifying a structure and hopefully that is what the industry will take away right now.
On a couple of points by DRC1 - I think it amazing to find calculations from 67....in fact, I think the business standard is 7 or 10 years and out the door they go. For someone to offer the calcs up shows a willingness to help. Hell, anyone could have started throwing work into the shredder when the news it the stand.
I defintely agree with DRC1 comments on code changes, there was a significant code change for steel in 1989 AASHTO. At present, I'm not aware of any mandate that when upgrading bridges for new loads, all member stresses and member proportions must now meet the new code. I do think there are building codes that spell this out. I could be wrong though, would like to hear others comments on that.
I also agree with the deterioration comment and Strguy11 about inspection.
Regards,
Qshake
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
![[pipe]](/data/assets/smilies/pipe.gif "[pipe] [pipe]")
