Poor design led to I-35W bridge collapse?

The bridge was inspected and noted to have accumulated flaws: 1) corrosion in / near bearing surfaces (a maintenance item), and 2) fatigue cracks (indicating loading or aging in excess of design).

Maintenance on *these things* were not underway. They decided to inspect it more often, to see if maybe inspection would fix them. They can afford inspectors, but they cannot / would not afford the repairs.

And there are 17,000 more bridges with similar diagnoses.

The purpose of government is to do for the people what they cannot do for themselves. "Lack of planning" they can do for themselves. It takes government to "cut taxes at the expense of basic services".

Robert, I don't think it was the engineering. I'd bet the engineer that designed the bridge did not design it for an infinite, (effectively) unmaintained lifetime. And I'd also guess that there are more lanes of traffic painted across that stretch of concrete than the engineer was told to design for so many years ago.

I'd say (from my armchair) the root cause was routinely exceeding the design loading and not performing necessary maintenance.

David A. Smith

"The purpose of government is to do for the people what they cannot do for themselves. "Lack of planning" they can do for themselves. It takes government to "cut taxes at the expense of basic services".

This is not true, I have yet to see the government bring me a beer while I am watching the Science Channel. The purpose of the government is to provide the fundamentals and infrastructure for the general walfare of society. There are some highways and bridges built entirely from private funds and users are charged a fee to driven on them. But generally, roads and bridges are the domain of government. The MN State government obviously failed to provide adequate maintenance for their bridges, electing to provide "More of something else" to their electorate, to "buy" a few more votes for specific politicians, at the expense of the infrastructure.

"I'd say (from my armchair) the root cause was routinely exceeding the design loading and not performing necessary maintenance."

I absolutely agree with this conclusion, but it will take a few months investigation to determine exactly what failed, initially, but when the big pircure is modeled, the above will most likely sum it up.

buckling"

bridge.http://en.wikipedia.org/wiki/I-35W_Mississippi_River_bridge>

Pittsburgh."

I am in agreement. The flimsy support is quite vulnerable to mechanical damage and corrosion. The consequence of damage/loss of this support is a design disaster... down-grading the entire bridge from a locally fixed shallow arch to a stiffened plate with edge- truss stiffeners, a massive loss of bend/twist rigidity of the arch. It is like removing the keystone not exactly at the center, but like removing some stones at quarter points.An arch of increasing reinforcement cross-sections designed all the way to the base/ground was needed and that should not have been compromised when once began at the mid-span.It would have responded better to unanticipated off- design loads like corrosion weaknesses or flutter of a ribbon bridge assisted by dysfunctional joints.

The tell-tale picture shows fracture roughly at the zones of sudden loss of bend/twist stiffness across span.

Narasimham

Dear Narasimham:

Was it not designed for the design loads "40 years" ago, with the expectation of reasonable maintenance?

The expected mode of failure was triggered by "fatigue" and "corrosion". Both of which require maintenance.

In 2005, had they responded by reducing the number of lanes across this bridge, until funds could be had to do actual repairs, we would or would not be having this discussion?

Your area of expertise is... ?

David A. Smith

I'm interested in reading what your area of expertise is.

buckling"

bridge.http://en.wikipedia.org/wiki/I-35W_Mississippi_River_bridge>

Pittsburgh."

--------------------- i dont ahve much information about thespecific case yet we know generally that steel bridges has a severe problem of corrosion OTHOH concrete bridges ar emuch better from tha tpoint of vew ie much less maintanance problems that is th emain reason that in our counrty we refrain as much as possible from steel bridges of corrosable steel (there are kinds of much less corrosion valnearble steels

th e'fatigue of steel is as well a factor yet we ahve to remember that that bridge was holding on i think more than 40 years !! moreoever the sayndards of desighn take in acount fatigue of material and itis taken in acount in a bigger safely factor (salmmer stresses ) so my guess would be tha tit was more a problem of bad maintanance!!

yet of course i cant judge from tousands of miles away wihtout al lthe needed data and facts anyway if it ia systemtic problem it is obvious that there is an urgaent need to avoid further disasters .. just immagine what could happen if that bridge was not falling to water but on a running train..... ATB Y.Porat

----------------

Concrete is not stone, you have problem with carbonation and resistance against a lot of chemicals. A concrete bridge will not last for ever unless you replace some of the concrete now an then.

Fatigue problems are design problems. If you don't call it maintenance to replace details that have shown to have a shorter lifespan than that of the rest of bridge.

But things like this have happened before, and probably will in the future also.=20

--=20 Mvh Anders Lager=E5s

-------------- i forgot to tell you aht i am a bridge engineer with more than 40 years experiance...

concrete is saied to last more than 15 years

youare right that in some cases it has problems mainly if not properly desighned

ie not enough cover of reinforvement stell or in sea while it demads special concret that is chlorine resistance and there are such . another problem is salt that is spread fo rdifrosting ice

if can be as wel ahndles by soecial coating of the comcrete . yes the satndard of bridges deoand

Dear Irving Layton:

Licensed professional mechanical engineer in the State of Arridzona. Got my degree in machine design (graduated with honors from Arizona State University).

And I have seen "politicians" find more room on the top of bridges for more than 50 years. And "inspect until the problem goes away" for almost as many years.

Are you satisfied, or do you care to snipe some more? The person I was responding to seemed to have some failure analysis capability. I enjoy talking to someone who knows what they are talking about.

David A. Smith

Dear Anders Lagerås:

These charaters in your name are going to be butchered upon posting. My apologies in advance.

Absolutely true. There are only two materials (as far as I know) that have an infinite fatigue life: iron loaded to less than 80% of hot rolled steel yield strength, and carbon epoxy composites (I don't recall the load limit, 40% of yield perhaps). Concrete will fail eventually.

That is long for a food or software designer. But that is usually less-than-minimum life for a "permanent" bridge.

A fatigue problem *may* be a design problem. Catastrophic failure due to failure of a "single joint" is a design problem. Any structure that does not very conservatively load steel, and have no other materials in its construction, will have fatigue. And hence requires periodic maintenance, not just periodic inspection.

Like the politicians-in-training that schedule, and actual politicians that fund, maintenance? It would not take much to cut the number of lanes that can be used during any hours, once a finding of "structurally deficient" has been made.

Do you have the design specifications for this particular bridge?

• Was this a permanent routing for I-35W,

• was this a temporary bridge that simply had its term-of-service extended,

• was it designed for the traffic that it was getting 40 years after being put in service,

• is it a design standard in Minnesota to make all bridges good for 100 years without maintenance?

Politicians also fund least-cost design, and send work to those that meet their cost limits. Which would be shame on the engineering firm also, should that be the case here.

Or this one, due to contractor shortcuts:

Really, they could have cast iron in a monolithic structure down to bedrock. Why not argue that it is a design flaw that they did not do this?

Civil engineering design, just like anything else, has cost limitations. And it is standard practice to *perform* maintenance on bridges... every municipality on the East Coast of the US with a bridge has a pretty significant annual maintenance budget.

David A. Smith

...

Which reminds me: the first iron bridge there ever was, closed to traffic years ago. This bridge at Ironbridge, Bridgenorth developed multiple cracks in the cast ironwork, when traffic multiplied in weight and volume far beyond the Coalbrookdale ironmaster's expectations. Imagine that: After only two hundred years!

Brian Whatcott Altus OK

I just saw this NY Times article that had this alarming fact:

Engineers See Dangers in Aging Infrastructure. By JOHN HOLUSHA and KENNETH CHANG Published: August 2, 2007 "Still, a study by the Federal Highway Administration found that visual inspections, the primary method used by bridge inspectors, only rarely detect cracks from metal fatigue. "In the study, completed in 2001, 49 bridge inspectors from across the country examined test bridges in Virginia and Pennsylvania. Only 4 percent correctly identified a fatigue crack. Worse, many inspectors identified nonexistent problems, suggesting that bridges sometimes undergo unnecessary repairs while some serious conditions are not detected. Inspectors now sometimes employ tools like ultrasound, but those add time and cost to their work."

Take a look at the image accompanying the NY Times story. You can see the supports for both the I-35W bridge and 10th Avenue bridge in the image. (Note: this image shows the 10th Avenue bridge actually has 4 lanes not 2.) The I-35W supports look puny in comparison, and this is for a bridge that had twice as many lanes and did not have the extra supports in the middle of the river.

In this article, an inspector for the bridge also says he believes it was a design issue:

Inspector suspects design flaw. BY KEVIN HARTER, JENNIFER BJORHUS and TOM WEBB Pioneer Press Article Last Updated: 08/04/2007 11:27:08 PM CDT "The two members were fracture critical. If one fails, there's nothing to support the other," he said, adding that the 40-year-old bridge was built using old national guidelines. "We don't design bridges like that anymore," he said."

Bob Clark

You could have thrown in a picture :-)

Notably, the bridge has a high, curved arch, which extends all the way to the ground.

Bob Clark

Good thought!

Brian W

It is far too early to be pointing fingers and throwing rocks at anyone or anything for being at fault for the catastrophe. I will add that there are other bridges in the Twin Cities area that cross the Mississippi and Minnesota Rivers that are steel truss designs and much older than the I35W bridge. The Robert Street bridge in St. Paul is much higher and older and is not in imminent danger of collapse. There's much more to the story than the type of bridge design.

BTW: Unlike what the History Channel said, the I35W bridge does /not/ connect Minneapolis and St. Paul.

: > : : > : : > : >Do you have the design specifications for this particular bridge? : > : : > : >* was it designed for the traffic that it was getting 40 years : > : >after being put in service? : > : ... : > : >David A. Smith : > : : > : : > : Which reminds me: the first iron bridge there ever was, closed to : > : traffic years ago. This bridge at Ironbridge, Bridgenorth developed : > : multiple cracks in the cast ironwork, when traffic multiplied in : > : weight and volume far beyond the Coalbrookdale ironmaster's : > : expectations. Imagine that: After only two hundred years! : > : : > : Brian Whatcott Altus OK : >

: > You could have thrown in a picture :-) : >

Imagine that... all the way down to the ground, huh? Amazing. I had a girlfriend with legs so long they went all the way up to her arse.

So that's where McDonald's got the idea for their sign from...

Whenever there is a disaster there is always early speculation about what the cause will be. The video of the collapse and the computer modeling done on it on possible failure modes before its actual collapse will probably allow the cause to be determined quickly. I like the idea of the extensive computer modeling. It should be done on such outmoded designs and bridges that are structurally deficient. It will give us an idea of where the most detailed examinations should be done. That report in the NY Times I cited suggests visual inspection is not enough, and the collapse itself proves it. Extensive modeling might tell us where we can focus expensive, time consuming methods such as ultrasound to get a better idea of a bridges condition. BTW, the Robert Street bridge is not of the steel truss type. It is of the concrete, multiple arch type. Note also there are supports within the river:

Robert Street bridge.

Bob Clark

Looks alright here.

But to make sure it is conservatively loaded enough to avoid fatigue problems during its expected life is a part of the design. Maintenance like repainting and maybe replace corroded parts are required. And inspections also to identify problems with the design, a well designed bridge will have no problem with fatigue during its ordinary lifetime. And after a bearing capacity classification the new lifetime.

no

If not should a bearing capacity classification have been performed and it should have a new design lifetime, if now fatigue is he problem. The number of lanes could be reduced, some parts replaced or what ever.

No, you always need to repaint it, replace bearings and things like that. But to replace every single part of the bridge at least two times since they do not last for more than 40 years, sounds a bit odd. And it is probably not supposed to be that way either.

Guess that that can cause the same kinds of problems as it always does when it comes to quality. But you are supposed to get your self another job instead of designing bridges that kills people. And the design lifetime should be written on some drawing. =20

Not as long as the ashtoo standard specification did not tell them to and you can do it some other way that works. =20

But if the design lifetime according to the standard specification is

--=20 Mvh Anders Lager=E5s