Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Bridge Collapse in MN 29

Status
Not open for further replies.
"If you are a bridge engineer you should know," said Astaneh. "You should know that bridges built in the 1950s and '60s -- they use details that are not good for fatigue fractures. That means that this bridge is 'fracture critical.'"

Wrong. At least they redefine "fracture critical" a little better a couple of paragraphs down.

Hg

Eng-Tips policies: faq731-376
 
"Meanwhile, a source close to the investigations said engineers are focusing on the thickness of the steel gusset plates, which he said were half as thick as they should have been. The source didn't know whether the original bridge design called for half-inch gusset plates or whether the contractor at the time used plates thinner than those specified.
"You cannot overplay the gusset design issue," the source said. "

 
NTSB officials said "people have run maybe a little bit too far" with the statement on the gussets.
 
from the Twincities article: "Gussets...are critical joints bearing a lot of stress... they were riveted and apparently welded to beams, both outdated technologies no longer used by bridge designers".

I didn't know we stopped using welding; nevertheless, a good article overall.
 
The over 200 discussions presented thus far cover a great deal of points of view. One of the questions I have, how do we learn from this experience? And how do we apply this experience to other bridges having similar conditions?

It is my view that the two in-depth inspections should be used in conjunction with a calibrated analytical model of the bridge components to assess the "fracture critical state" of the bridge. Included in this model should be an orthotropic deck.

From a detailed inspection, a map of the discrepencies can be developed. Missing bolts, bearing conditions, cracks (location and length)and corrosion (location and extent) can be identified. The effect of these discrepencies can then be introduced into the analytical model. New fixity conditions can be determined for each affected structural member and included as adjustments of the model. Then new critical load paths can be established. Either a buckling or stress ( with K1, K2, and K3 fracture modes) criterion should be used as the final evaluation for integrity.

If there is no plan to incoporate the inspection results into an engineering assessment, then why inspect the bridge? A critical assessment of the extent of a discrepency should be known, a priori, especially those bridges that have been designated as "non-redundant." If the standards of bridge design do not have specific recommendations as to fixes and the ramifications of these fixes on structural integrity, then what good are they?

There have been studies performed by this author that shows that the life of a structure with small cracks can be exteded (almost doubled) by treating the tip of the crack. By establishing a compressive zone immediately infront of the crack, one can arrest its growth. I see no need to apply a 2 to 4 inch diameter holes as suggested by the standard.
 
mtnengr -

Until a collapse mechanism(s)is found it is far too early to pass on learned knowledge from this tragedy. All there is now are a few educated guesses from experienced bridge engineers and many wild ass guesses from laypeople.

Speaking from experience, wherein possible, cracks and other physical distress are modeled in a rating analysis of bridges. This is not always the case though and is an area of improvement.

As for the standard of arrestor holes, this is viewed as the most effective means of crack extension mitigation. I don't see how your proposal of introducing compression at the crack tip can be accomplished in many cases considering the geometry of the members and ensure that it is reliable once in place. Clearly if it were reliable it would be a standard in use today. It's not very practical.

Regards,
Qshake
[pipe]
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
 
I'm not a bridge engineer, but it seems that the rating analysis of bridges may be part of the problem. It doesn't seem that it really provides an analysis of the actual risk so that decision makers can factor that into the O&M decisions. A risk assessment procedure for rating bridges might be more useful. This is often done with dams and could easily be done for bridges also. By constructing a dam or a bridge, you are creating a potential hazard if it fails. O&M of dams often involves risk assessments which looks at the probability of failure and the consequences of that failure. This analysis of risk can guide dam owners and the public in the necessary maintenance, inspections and mitigation measures to take.

Using the risk assessment example, if risk had been evaluated for this bridge, more proactive O&M might have been considered necessary. Even using a qualitative assesment, the bridge would be considered high risk. Probability of failure due to fatigue cracking, age, lack of redundancy, ship collision, etc. would have certainly been higher than normal. And consequence of failure for this bridge are very high due to the high traffic volume, economic impacts, environmental impacts, loss of life and cost to replace. A risk assessment would have likely shown this bridge to be at the high end of the risk spectrum as compared to other bridges in the state. Perhaps a new model for evaluating bridges should be considered.
 
cvg -

I agree that the actual rating analysis should be more quantitative providing a probability of failure, rather than a number which means it is more likely to collapse or not. At the very least, all fracture critical structures should undergo this type of analysis.



Regards,
Qshake
[pipe]
Eng-Tips Forums:Real Solutions for Real Problems Really Quick.
 
"Perhaps a new model for evaluating bridges should be considered."

Theoretically seems plausible, realistically, I don't know that it would help much.

DOT has a certain amount of funds available and subjectively allocates funds to repairs, maintenance, bridge widening, new bridges, etc. Bridge widening and new bridges seem to be heavily favored by the public (spending money to repair a bridge isn't as noticeable to Joe Driver as a new bridge or bridges with more lanes).

This and other bridges are/were identified with problems. There probably needs to be some statistical analysis to delineate when a particular rating factor an existing deficient bridge supersedes a factor for new / widening bridge project.

Maybe something like:

Project Priority Index (PPI) = Risk Factor x Demand Factor x Benefit Factor

[tt] Bridge that collapsed Some New Bridge
Risk Factor 9 1
Demand Factor 2 8
Benefit Factor 3 4
PPI 54 32
[/tt]

Certainly crude, the gurus would have to work out the details probably with some weighting factors on each component of the formula.

I guess I'm basically agreeing with you, but just pointing out the public/political motivations for new / widenings have to influence the formula somehow.
 
I'm going to get in trouble again, but here goes.

What in the hell ever happened to the find it and fix it philospohy when it come to a fabricated structure.

 
unclesyd,

Glad to join you in trouble. The inspectors found a host of things wrong with that bridge, but the managers, risk assessors, fuzzy thinkers, bean counters, politicians, etc. dithered. And now they would like nothing better than to blame it on the long dead designers.
 
risk assessment was not popular with dams just a few decades ago either. Than, due to several well publicized disasters such as the Teton Dam failure, federal legislation was enacted for dam safety. But as you can see below, it took 10 years to implement and even then, only after the death of hundreds of people and billions of dollars in losses. We are now going through the same process with levees after Katrina and perhaps (some) bridges should be treated in a similar manner...


Dam Safety History
1972 Failure of Buffalo Creek Dam, West Virginia; 125 deaths
1972 PL 92-367 the National Dam Inspection Act of 1972
1976 Teton Dam Failure; $1 Billion in losses and 14 deaths
1977 Kelly Barnes Dam Failure; 39 deaths
1978 Corps of Engineers begins the National Inspection Program
1979 Federal Guidelines for Dam Safety prepared
1979 Executive Order 12148 from President Carter created FEMA; FEMA Director to coordinate Federal Dam
Safety efforts
1979 Memorandum from President Carter requiring the head of each Federal dam safety agency to implement the
Federal Guidelines
1982 the National Program of Inspection of Non-Federal Dams completed and National Inventory updated by
USACE; Final Report to Congress
 
note that risk does not equal the probability of failure. Risk = failure x consequence. therefore you can have two bridges with the same probability of failure which is high. A large high hazard one and a small, two lane low hazard rural one. Risk for the small bridge will be much lower since there are few consequences if it fails. This makes it very simple for the lay person, even a politician to understand where the dollars should be spent.
 
hokie66 & unclesyd,

There's not enough budget to fix everything that you find.

You might be interested to look here:

Especially around question: Q313-18

It still comes back to priorities. It would be interesting to see where the DOT had the repairs slated in their long range plan.
 
mjl23,
I agreed that the money needed to fix everything you find is not available as the monies have been spent in other places for other things. It just passed through the bridge fixing fund on the way to some bridge to nowhere.

Though I've never specifically be tasked with a bridge inspection I've assisted in the inspection of several large steel structures. One was a giant coal unloading gantry that was to be inspected as bolts were falling out and hitting the coal barges. As we were ascending the gantry for visual, MT, and PT testing it was evident that a major problem existed. Everywhere we looked there were major cracks and broken welds. I took the inspection crew down and conferred with the owner and much to his chagrin I told him that any inspection was useless and a waste of money was until he fixed the major problems. I offered to leave 2 technicians to mark all visible defects for repair. He opted out and and about three months later after another NDT company performed a thorough inspection and while waiting for an engineering evaluation a component of the gantry fell and severely injured two crewmen on a tug.

The moral of this tale is fix the obvious and mark the minor deficiencies for later. It appears to me that too much emphasis is put on finding the small fatigue cracks for study, evaluation, monitoring and ignoring the obvious like a warped gusset plate or a slightly bent member, etc.
Stop drilling a crack is just a band aid especially if you return and the crack and breached the stop hole.

Another thing that bothers me is that the people responsible for the inspecting and repairs are investigating themselves.






 
cvg,

I don't know anything about how risk assessment of dams is done, but it seems to me that all the steps necessary to determine the risk of this bridge had been taken except to proceed to address the risk. If the corrosion, fatigued members, and frozen bearings had been dealt with, we would not be having this conversation.

 
Just a point of reference, the Seattle Space Needle was built for the 1962 International Exposition, and the support legs use gussets similar to the photos shown of those on the bridge, i.e., loaded with bolts, so that appears to be the standard practice of that era.

TTFN

FAQ731-376
 
Risk Assessment is starting to find its way into transportation facilities, but it's the result of 911. Right now, some agencies are using it as a tool to prevent a collapse resulting from a terrorist attack.

A few years ago I was involved in a tunnel design project and we did a RA for that reason. It is useful and forces everyone to think more deeply about various types of failures and their consequences.

AASHTO has developed guidelines for its implementation
 
I agree that "Risk Assessment" has it's place in some designs but it should not be the controlling factor when possible fatalities are involved. Using RA there is major problem in that the first iteration is usually wrong and it takes several iterations to find some acceptable level of risks.
How would the number of permissible fatalities be determined say for a 10 car, 100 car, or a 1000 car bridge?

A prime example of RA failure is the tragic loss of the space shuttle Columbia with all it's crew. You might throw in the Challenger, though like bridges a little politics entered inot the picture. NASA was the daddy of RA.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor