35W Failure Investigation 1

[unclesyd]

I may (already have) get run off but I've got to ask.

This is from an grizzly old codger that has seen a lot of money spent for the wrong things.

Failure analysis was my stock in trade for 42 years and realize the need for it but there are times where it becomes a waste of money that could be spent elsewhere. In this case inspecting similarly constructed bridges. We are going to drop sevral millions of dollars into this failure analysis that will conclude that it failed. We know this is a very poor design and is subject to catastrophic failure at any time. The bridge had been categorized as a bridge in
a failing mode, they were in a wait ans see mode. This bridge is now a non-entity and is no longer a threat to anyone. Give an experienced engineer a couple of hundred grand and couple of good technicians and let him evaluate the failed structure.
The NSTB has come up with all types of teams and groups to do this and that, none of which has any bearing on any other bridge. They are hiring a consulting firm for a ton of money. Their walk-in fee will probably be in excess of 2 million dollars.

The cause of failure of this particular bridge will not help anyone in future failures of other bridges. There will be no failure if you detect the cracking and fix it instead of waiting to examine the fracture surfaces

What really set me off is that they have a survivability team. This is quite evident all you have to do is count the number of people that got off and the number that died. If the bridge doesn't fail everyone survives.

If you build a good bridge and keep in good repair it want fall down.

 

[RossABQ]

While I agree with many of your points, I don't think it is a given that nothing can be learned from a failure analysis. (I am speaking in an engineering sense, obviously the lawyers want to know who to sue) What if the bridge was really fairly sound, but the contractor doing repairs did something stupid? We could spend billions replacing other bridges prematurely. $2 million is chump change compared to the cost of one new replacement bridge.

I saw on TV that the state of MN has already ordered re-inspection of 22 other similar bridges.

 

[unclesyd]

A failure analysis of failed bridge has to contribute something but it's the approach of throwing money at a dead bridge. There are now 3 independent investigating groups involved now so the tab is now $6 million. This money could pay a lot of inspectors to a lot of inspections.
All of this is probably mute as when defects are found there is no action.

What worries me about your concern about a stupid contractor is where was the DOT inspector.

If you want to get scared try to catch the NTSB update on the news. This is by the Chairman of the NTSB.

 

[SteelMover]

I would maintain that we learn....maybe codify...more from 1 bridge that fails than 1000 that stand for their whole life.

 

[bridgebuster]

unclesyd - I agree that mandate to re-inspect trusses is an overkill but it's the politics of the situation.

I disagree with your statement "We know this is a very poor design". Trusses are reliable structural system. Granted a truss is fracture critical (so is the arch and the suspension bridge) but consider: A human body cannot survive without its head, does that make the body a poor design?

We have to deal with things for what they are not what we want them to be.

 

[unclesyd]

bridgebuster,
Appreciate the feed back.

As I've mentioned sevral times that I wasn't a bridge person so in making that statement I was quoting people, given credit as head of civil engineering, on the major news networks that said that this was an inherently bad design due to the lack of any redundancy in the structure. Their quote was there was no safety in the two main planes. It was also mentioned that this same type of structure wouldn't be built today due to these safety problems and the labor involved in fabrication. Both said this was built with cheap steel and cheap labor in it's day.

I totally agreed that politics are the biggest impediment to good engineering and design and this case failure analysis.

They need to pick up remove the rubble and get
the bodies out and forget about all those scientific toys they want to try out and all the training for the different divers.

I've seen a lot of very similar structures built a quarter mile from my boyhood home by the Nashville Bridge Company. Our neighbor was the head of layout for this shop and I've heard him tell my father many times that they had a new "Architectural Wonder" to fabricate.

 

[bridgebuster]

unclesyd

I guess there's a lot of truth in the saying "those who can do; those who can't teach."

Actually, trusses are still being built, although not as frequently. Cable-stayed bridges are in some cases more ecomonical for long-span crossings than trusses and they have lower maintenance costs with respect to painting. And, people would rather have a cable-stayed bridge rather than a bulky looking truss.

Current design practice is to design the truss bridge as three-dimensional structure.

go to

gobridges.com

they have a good article on the collapse.

 

[DarthSoilsGuy]

3 independent studies, i hadn't heard that a 3rd stepped in.

they should all be released at the same time (at least on a prelim level).

i'll be proposing for my role as the 4th independent reviewer for another $2MIL. you can expect my report some time in the 2-week period immediately after the NTSB publishes their investigation results.

 

[IRstuff]

Talking heads are NOT always reliable sources.

Similar talking heads lamented going up against Saddam's thousands of T-72's prior to the first Gulf War, even though anyone who was remotely in the know knew that the T-72 was a clanky rattletrap compared to the M1, and that the Apaches and Kiowa Warriors could probably kill the T-72s without much assistance from the M1s.

TTFN

FAQ731-376

 

[HgTX]

Their quote was there was no safety in the two main planes. It was also mentioned that this same type of structure wouldn't be built today due to these safety problems and the labor involved in fabrication. Both said this was built with cheap steel and cheap labor in it's day.

Not exactly. "No safety" is an unreasonable statement. Not even "no factor of safety". No redundancy, yes.

As someone else said, trusses are still built. I know of one big one being built right now. And plenty of other fracture-critical bridges, not necessarily trusses, are still built. They are designed to be safe for certain loads. You overload the bridge, you'll have problems. You overload a fracture-critical bridge, you'll have more dramatic problems.

What's changed since 1967 inludes (1) higher toughness requirements for both the steel components and the weld metal (2) a greater awareness of connection details that may be particularly prone to fatigue and fracture. So even a similar-looking truss bridge built today would be more robust. But it's inaccurate and unfair to say that it was just a crappy bridge built in the bad old days.

The cause of failure of this particular bridge will not help anyone in future failures of other bridges. There will be no failure if you detect the cracking and fix it instead of waiting to examine the fracture surfaces

The cause of failure of this particular bridge may point to a particular fracture-prone detail that hadn't really been paid much attention to in the past. Some bad fatigue details are worth keeping an eye on; others are worth a retrofit. Knowing exactly what the problem was will help state DOTs decide what they need to do about their own bridges.

I totally agreed that politics are the biggest impediment to good engineering and design and this case failure analysis.

They need to pick up remove the rubble and get
the bodies out and forget about all those scientific toys they want to try out and all the training for the different divers.

I fail to see how just clearing out the rubble and avoiding the use of available technology ("scientific toys") is in the interest of good failure analysis.

Hg

Eng-Tips policies: faq731-376

 

[unclesyd]

A little clarification on my part. The conversations concerned this particular design of bridge design and I should have stated that realizing that truss type bridges are still being built. I used the word safety probably because of my notes and as a mixed metaphor because of a statement made several times about the failure of one component of this design would cause the bridge to come down. The point was that this particular design wouldn't be considered today for this location.

I agree that fracture critical bridges are being built and hopefully with the required safeguards of inspections and load limits. Hopefully they don't fail before completion.

As far as the bad old days a lot of my school buddies were designing and building bridges in this era.

It has been put forward by several people interviewed that this particular design was chosen sorely to eliminate a center support in the river channel. Maybe it was the best they had at the time.

I see mainly rivets in the fabrication with very little welding.

I've been in involved in several incidents where loss of life occurred and the primary directive was recovery of the bodies, nothing else.

All the surveying in the world isn't going to give you any information on any critical detail. It is just to accommodate someone using a computer model to emulate this failure. This was a statement by the head of NTSB. Why wasn't his model run to emulate the failure prior to the actual failure to see how the bridge would fail. I was involved in failure investigations for many years with most information being resolved at the micro level. Macro investigation played part at different times though the acquisition of fresh fracture surface was the prime directive. Hopefully they will find a big red arrow pointing to the problem.

If you start removing the rubble you take the divers out of harms way and recover the bodies.

Addenda:
There was a detail on this the original construction of this bridge that that was modified on the the first retofit.
It is the design of the median/divider.

 

[unclesyd]

To finish my last post which was accomplished by the power company.

The median was no more than a concrete rectangle with beveled sides in the middle of the bridge. This divide was modified by installing posts with an Armco like rail.
2o years later this same type divide was used on a 3.5 mile long bridge over Pensacola Bay. It was evident from the start that this wasn't going to work as cars would bump into it an either jump the median and into head on traffic or over correct and hit the outside guard rail. It took numerous accidents and 2 deaths before this was properly corrected. The last death was a relative of state politician.

The point is that we have a hard time learning from previous problems and solutions.

 

[Qshake]

unclesyd - Toward your item on rivets and no welding....aside from areas like Houston and Louisiana and locations that are close to large oil works where field welding is performed everyday and is damn good, bridge engineers shudder to think of field welding for two reasons:

1. It is a big fatigue and fracture items, i.e., the origin of stress risers.
2. Most DOT's performing the construction inspection do not have the expertise to properly inspect field welding.

Further to point two is that field welding may require special testing due to point one and that cost money which is unnecessary thanks to rivets and bolts. Mainly bolts now.

Regards,
Qshake

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