Bridge Collapse 5

[bridgebuster]

"Sheikh added that there have been an estimated 34 bridge collapses in the United States, Canada, Europe, and Asia over the past 160 years."

34 collapses in 160 years seems QUITE low

 

[JEmH]

I would have to think that the 34 collapses in 160 years is a misquote or typo. Perhaps 34 collapses resulting in fatalities? Maybe 160 collapses in the last 34 years?
If we just count all bridges on public roads with spans over 20' in the US, India, the UK, and Canada (i.e. where most of the people on this form are) we could name at least one collapse per year.

 

[cooperDBM]

unclesyd: I think the crack you're referring to is simply wet due to a transverse construction joint above it. The white crack is filled with efflouresence (salt).

I think the shear failure started as a vertical separation of the top long. steel lap at the back of the bearing seat (along a horizontal plane). Long. steel on the bearing side of the failure went with the falling concrete while the long. steel on the abutment side stayed embedded in the renmants of the topping. Seems that there was no shear steel tied over the top of the top steel to prevent this separation.

If this is true it would have been quite sudden and prior inspections may not have seen much of a crack unless they could have seen the back face of the bearing seat (probably a hor. crack a couple of inches below the top of the deck). The inspector who looked at the bridge immediately before the failure wouldn't have had a change to make that observation. The span seems to be box beams placed side-by-side which wouldn't have allowed a view of that face.

Annual or biannual inspections in Ontario, and I believe Quebec, are carried out by trained technologists or maybe young engineers. Experienced engineers would inspect the bridge if a structural evaluation was required, say for a rehab. I don't think it mattered in this case since I think it was a sudden brittle failure due to bad design detailing. If I'm right I'm amazed it lasted as long as it did.

 

[cooperDBM]

Looking at the 1st photo on

http://www.keystonebridge.ca

at the twin bridge still standing (right side of the photo) there does appear to be a diagonal to horizontal crack in the side face of the cantilever near the bearing seat. There are a couple of streams of efflouresence at that crack. The surviving cantilever of the failed bridge (photo 2 & 4) also shows a similar horizontal crack (with eff.) near the top at the back of the bearing seat where the separation seems to have started.

 

[prost]

I dunno; if an inspection 9 months ago doesn't tell you what you need to know to avoid a collapse (and fatalities) today, then I still think you have an inspection problem. Almost any design, no matter how poor, can be made safe with adequate inspections. The aircraft industry assumes that there will be problems with design, so that redundancy is designed into the system at the beginning, but proper inspections are still needed to avoid aircraft crashes caused by unanticipated problems.

 

[HgTX]

Inspection won't help you with sudden brittle failure.

Hg

Eng-Tips policies: faq731-376

 

[dik]

There was a news article this AM that they had ruled out corrosion problems and were checking reinforcing to see if it had been installed as spec'd. Can't verify this.

Dik

 

[prost]

Sudden brittle failure? Caused by what? Again, if this were an airplane, and a failure mode could not be inspected out, so to speak (that is, inspections catch indicators of impending failure before catastrophic failure occurs), then the structure or system would have to be retired before the failure mode could kick in. If this isn't how the road systems are maintained and inspected, then I just started feeling a whole lot unsafer than I have felt in the past.

 

[HgTX]

All I'm saying is that in-service inspection in itself is no substitute for proper design and construction (and inspection during construction). If ductile failure was not a consideration in design, or if construction was shoddy and led to a situation that is prone to some kind of sudden brittle failure that doesn't give much external warning before it happens (which is kinda what sudden brittle failure is), then periodic inspections, especially those based on the assumption that everything in construction went as it should, are not going to help you.

Hg

Eng-Tips policies: faq731-376

 

[unclesyd]

HgTX,
If you had inspected the bridge and it checked OK according to the inspection parameters and as you were walking back to your truck and turned your ankle on a chunk of concrete and at the same time a chunk fell on your head, thank goodness for hard hats, wouldn't one think something is amiss.

 

[HgTX]

I was speaking more in a general sense. I'm not saying that there WEREN'T signs in this case that something was wrong (and they did realize there was enough wrong to send someone out there). But the blanket statement that any competent routine inspection program will prevent any kind of failure (barring getting hit by an outside actor of some kind) is just not true. Bad materials and bad workmanship can lead to a situation that deteriorates very quickly over a timeframe shorter than the design parameters suggest the inspection interval should be.

Hg

Eng-Tips policies: faq731-376

 

[dimjim]

I would wonder if any unusual heavy
equipment passed over this bridge in
the last month. Did the overhanging
walkways put too much tensile stress
on the reinforcing bars?

 

[prost]

This is certainly a very interesting discussion!

The bridge was 30 years old, right? I would think normally a bad design would be evident much sooner than 30 years. The fact that the bridge was erect for 30 years (assuming no problems encountered during that time) would suggest that the bridge was properly designed and constructed.

The statement was made by 'dik' that corrosion was ruled out (of course given the recent history of early news reports on various sensational cases in the States, I would tend to 'trust but verify' all such reports). After 30 years, some kind of age degradation is inevitable of course, the question is did the age degradation cause the failure?

 

[apsix]

I did a quick google on fatigue failure of reinforced concrete. There are a number of papers/studies on this failure mode therefore it cannot necessarily be assumed that the bridge was properly designed and/or constructed because it stood for 35 years.
Failure due to inadequate design or construction doesn't mean that it didn't meet the required standards back in 1970.

 

[MikeHalloran]

One of the reports mentioned that they've started using huge quantities of salt on the roads in that area in recent years.

I'm guessing that chloride corrosion was not given a lot of consideration when that bridge was designed, because nobody bothered to salt the roads then. When I lived near that latitude decades ago, salt was used only to keep the sandpiles from freezing solid in their shelters. Salt is just an expensive abrasive when the temperature is below 0F all the time.

Sand was applied after each snowfall to provide traction, with maybe a little extra salt in the spring, to help break up the accumulated foot of ice on which we had been driving all winter.

Maybe global warming made it less futile to salt the roads, and hence catalyzed increased salt usage, and eventual failure of bridges not designed to withstand salting. Speaking of which, are _any_ steel- reinforced concrete bridges actually designed to deal with heavily salted decks?



Mike Halloran
Pembroke Pines, FL, USA

 

[Tomfh]

It wasn't lack of maintenance that killed this bridge it was lack of ties.

The product was defective from day due to bad design/construction. The little bit of corrosion was simply the straw that broke the camels back.

 

[Dinosaur]

I am amazed at the conclusions some get from these photos. While I only looked at the photos for a few minutes on the web, I could not draw conclusions except one; that I thought the drop in span detail and design philosophy was por for this very reason, lack of redundancy. That does not mean the design didn't meet the code or that the consrtuction didn't meet the specs. I also suspect some degree of corrosion contributed to the problem but the bars I could see looked very clean.

As I understand it, shear reinforcement provisions in the codes have been increased since the fifties but I don't know a better date on that change. This is one of the problems I see in our industry (transportation infrastructure) that rehabilitation of existing structures is rarely undertaken to address code modifications. That is a major discussion in itself. Also, I believe bridges are rated based on their moment capacity alone. Shear is not part of the rating. This should be appropriate for 95% or more of the structures in the inventory but I am sure some, as possibly this one, are controled by their shear capacity.

I hope we can find a specific cause that can be addressed in the future.

 

[JEmH]

This goes a little off topic, but Mike H., in answer to your question about designing for being pounded by salt, "sort of." In my experience, in Pennsylvania, most mild steel above the beam seat, and anything below that we really like or won't ever see again (like the back steel in the stem of a cantilevered abutment) is epoxy coated. Obviously, there are other corrosion protection systems, including using non-corroding bar in the first place, but I've only seen that a couple of times. I've never really seen a good answer on just how protective the usual epoxy coating is. I have, however, seen bars in a parapet that had been hit but never repaired, thus exposing the bar. The green shell of epoxy looked great, but contained nothing but rust. Don't know if that was a fluke or what.

 

[JEmH]

Sorry to double post, but Dinosaur:
Right, this is going to take a lot more analysis than can be had from a few pictures. Hopefully this all comes out in a form that we can all use on our own designs. It's fun to discuss now, though.
Shear is conisdered, and I find that shear often controls the ratings of concrete bridges. (PS and RC)
I will admit that we can be negligent on keeping a current rating on the substructures, which the failed portion of the subject bridge could be called, depending on your point of view.
I would also say that you have plenty of company (including mine, for what that's worth) on your dim view of drop in spans.

 

[civilperson]

The area of concrete usable for shear strength is only that portion below the location of loading. Some configuration of straps and hoops that transfer the load to the top would have to be present to use the concrete above the beam seat for shear calculations.