Minneapolis Bridge Failure Report Due out Soon 2

[MJCronin]

To all..

I understand that preliminary details have been released about this long awaited report.

As I understand it, structural gusset plates were undersized (or marginal) and the strength of the corroded plate was not adequate for the load.

Also, (as I understand it) there will again be an argument as to who is responsible for the final review and approval of structural details.

Anyone remember the cause of the Hyatt Regency walkway collapse in 1981 (Kansas City, MO)? ...114 people DEAD..!!!

The investigators found a structural detail that was done by someone who was incompetent.

http://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse

Deja Vu all over again ???

My opinion only...


-MJC

 

[HgTX]

My understanding is that they don't have the original design calcs, so they have no way of knowing if it was an error on the design drawings or on the plans.

I'm not sure I see the point in figuring out who screwed up 50 years ago; they're probably dead and almost certainly no longer practicing.

Hg

Eng-Tips policies: faq731-376

 

[MiketheEngineer]

Hyatt case - the EOR did review the final details - apparently very quickly and signed off - even though they were NOT what he had originally designed. He was well known, respected and a good engineer. Just missed one little bolt detail.

Our Missouri PE dues about doubled to pay for the lawsuit aginst him!! Never understood that??

 

[Qshake]

I think the Hyatt Regency case is much more complex than what is noted above. It involved a change during construction and not during the design. What was supposed to be a single threaded rod was divided into two threaded rods at a support, such that instead of a single shear matter there was a double shear problem.

Once again, the Minneapolis bridge matter is much more complicated than a plates not meeting an allowable stress. Heck the bridge stood for 40 years and the last twenty with more load and heavier traffic than it was designed to carry.

I just hate to see engineering diasters trivialized in a few sentences. That hardly makes us look good to the public, if we can't agree amoungst ourselves that these matters are typically more complicated.

Let's not throw the profession down the toilet because we feel compelled to make some comments on a website.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[hokie66]

The reasons for the Hyatt failure are well known. The design was borderline at best, and the change to the hanger detail made it half as strong. Human error.

The report on the bridge should also give clarity, but enough has been released that we know that gusset plates were undersized in key areas. Whether calculation error or drafting error, unchecked, we may never know. Human error.

 

[MJCronin]

To all..

Again, I only asked....is it "Deja Vu all over again" ???

Qshake, can you better quantify "more load and heavier traffic than it was designed to carry"

I am not a bridge designer, so please excuse my questions, "How is it possible to overload a highway bridge with traffic only..?? Only so many trucks/cars will fit on the bridge ?

Was it the temporary loads of construction materials that were being stored by the contractor that pushed the bridge over the edge ?

-MJC

 

[JAE]

MJCronin,

I'm not exactly sure what you are asking. Deja Vu seems more like a statement than a question.

 

[MotorVib]

Qshake is right a 40 year old bridge and it fail right when repairs are being done on it?

And as for the plates being too small, the max load of the bridge was increased. (How much I don't know but it has been increased.) As engineers how can we see what people and (the goverment) are going to use what we desgine in 20, 30, 40 or 50 years from now? We make our best assumes at the time and hope for the best and then tell the world the max load, current, preassure, or aanything else that our system can do.

Chris

"In this house, we obey the laws of thermodynamics." Homer Simpson

 

[Qshake]

MJC -

From the reports, this bridge is about 1500' long and about 90' wide. The concrete bridge deck was increased from 6.5" to 8.5~9". This alone is 2100 tons of additional dead load. In addition, the reports show that two additional lanes were added in or about 1998.

Now think back to the era of the sixities and seventies where labor was cheap and material was at a premium. Bridges (like airplanes) are designed to carry the most payload (traffic in this case) with the least self weight, they are optimized in other words. You cannot simply throw additional load on a bridge and not check members and more importantly connections.

Today it is different as material is cheaper than the labor producing it. So it may be hard to imagine the scenario I suggest above.

Moreover, our traffic has gradually increased in weight from the 1960s. To see this, check out the design codes for bridges, AASHTO, and see if we didn't start out with H15 then HS20 and now we're designing interstate bridges for HS25. Railroads are moving to heavier loads too. So essentially you have two additional lanes than what was originally designed for and more dead load and heavier traffic composition.

And as I noted, the biggest problem we face is engineers casting this matter off with the wave of a hand with the idea that thin gusset plates and only thin gusset plates are responsible. This is irresponsible and confuses the public who rely on engineers for life safety.



Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[hokie66]

To the contrary, this bridge has been studied and studied and studied, with the new dead and live loads. But the studies were faulty, as they did not identify the deficient connections.

 

[Qshake]

Yes, Hokie66 and the one area which I noted above that needs to change in our industry is not to take for granted that connections are stronger than members. The connections were not studied.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[hokie66]

Amazing as it is, that appears to be what happened. But surely that is not industry practice.

 

[Qshake]

I wholeheartedly agree with your surprise and am sorry to say that not only does this practice exist, some will go so far as to say they don't know how to evaluate joints and only do members since there is no software that I am aware of that analyzes bridge truss joints. There is software that will analyze members. It is a tragedy.

Moreover, I've seen many DOT personnel rant against prices for analysis and complain that such and such doesn't need to be done....I hope this practice is coming to an end. And I hope that whatever or whoever created such suspicion in the owners minds will not do so again.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[MJCronin]

Qshake, hokie....

Qshake, you stated "the biggest problem we face is engineers casting this matter off with the wave of a hand"

Fundamentally, isn't this what happened at the Hyatt Regency ?

WRT the Hyatt disaster, I have read up extensively on the "structural detail" that was marginal, at best, on the steel detail drawings, then changed and approved by the engineer of record.

For those of you who did not understand my quote in the first post, "Deja Vu all over again" refers to a well known and very humorous statement by Yogi Berra, one of the greatest (but unintentional)US humorists.....

JAE,...The definition of "deja vu is here:

http://www.thefreedictionary.com/deja+vu

-MJC

 

[JAE]

I know what deja vu means. I just didn't see a specific question developed from the use of it.

 

[JAE]

I can see that the comparison between the Hyatt R. and the Minneapolis bridge may come to mind with people...but to some extent they only relate in that they reflect human errors or imperfections.

In the walkway collapse, the underlying flaw was essentially in the culture of buildings project development within the structural community. At the time, the culture there in Missouri was that steel fabricators designed connections and the EOR simply specified the intent of the connection. Jack Gilliam's firm sent the drawings off to the contractor who sent them to a fabricator who started the shops and the design but never finished them. The first fabricator farmed out that job to another fabricator due to workload and the new fabricator incorrectly assumed that the connections had been designed. Upon shop drawing review, Gilliam's firm didn't review them intently, again assuming that the design had been accomplished by the fabricator. The change in rod from one to two didn't help but that is simply one of many details that got by everyone due to the culture of farming out the connection design.

In the Minneapolis bridge collapse, I don't see the failure as a similar cultural thing but rather one of a simple design error in calculation, or perhaps a drafting error. The "cultural" aspect of this could perhaps be that of an insufficient QC process originally, and also what Qshake suggests above that the long term review of bridges made some assumptions that allowed the gusset situation to go unnoticed for years.

I sat at a conference some years ago where Ed Phrang, the NBS engineer who reported on the Hyatt collapse, stated that "we engineers need to remember that our structures are not beams and columns connected together. Rather, they are a multitude of connections linked by beams and columns."

Qshakes statement that so much software is "beam and column" oriented vs. "connection" oriented is right on the money.

 

[SylvestreW]

JAE, your explanation of the Hyatt collapse is how I understand it as well. It wasn't a design error, it was a case of "who does the checking" (which was thankfully clarified).

Re: the bridge, let's wait for the full report.

-
Syl.

 

[hokie66]

Syl,

As well as the undetected error in the shop drawing stage, there was design error.

The National Bureau of Standards tested the connection of the hanger rods going through the boxed channels, and determined that the ultimate capacity of the originally specified connection was only about 60% of the required capacity.

So the failure probably would not have occurred if either the original design had been adequate or if the change had not been made. Both contributed, and there was no redundancy, so the tragedy occurred.

 

[JAE]

hokie66 - yep - I remember that Phrang stated that the original design (using one long rod through the boxed channels) would have had serious problems as well.

However, Gilliam, in a later talk I attended, stated that the "intent" of what everyone called the "original design" was that they were just indicating a concept and not designing the connection in any way shape or form.

He said that, upon coming to the site that night his first words were, "where are the stiffners?". He saw that the rod nuts had just pulled through the channel flange-tip welds and understood right away that the final "design" was flawed.

 

[hokie66]

In which case, although I respect his honesty and openness, I would have a hard time understanding how the assembly at that point could ever be entrusted to a "connection designer". It doesn't, or at least didn't, fall under any category of connection that I know of. If the member size was decided by the EOR of record, he needed to make sure that it could carry the loads, both imposed loads and reactions. It was the member which failed, not the connection.