Nanfang'ao bridge collapse Yilan Tiawan 2

[charliealphabravo]

I haven't found any mention of fatalities. Reported 20 injured.

Location

Collapse video

Bridge details

Photo of bridge system.

Edit: Updated dead video link. Thanks to JAE.

 

[Agent666]

While bridges as I understand it are often designed for the removal of a cable or missing a cable to ensure some level of robustness exists against say traffic taking out one of the hangers. I wonder in these assessments if the dynamic effects of instant removal of the hanger is accounted for in these types of analysis, time history or similar? I'd imagine for example you'd see significantly higher forces as the structure finds the new state of equilibrium. Any bridge guys have any experience in what might be considered best practice?

It would be interesting to see details of how the hangers/strands/tendons were anchored, it almost looks like they 'pulled through' whatever detail was provided setting off a domino effect of overloading all the other end anchorages. If you look at the photos, the one that initially went only has a few loose strands hanging from the top anchorage suggesting it possibly failed inside the arch box girder member

 

[IRstuff]

One would hope that the design process accounts for things like cables breaking; I might have expected the design to accommodate half of the cables breaking because of, say, small plane collision, or some-such.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[IRstuff]

According to Wikipedia, the bridge is only 21 years old.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[IRstuff]

Looks to me that the bridge deck falling pulled out the moorings of the arch; otherwise, the arch might have even stayed put. You can see the arch flex as the deck pulls the moorings, and arch, inward, and then relax when the load is removed. That arch itself appears to have had more than sufficient design margin.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[wiktor]

The bridge failed at approximately 10% or less of the design LL - just single typical fuel truck, which is rather difficult to overload. The design should account for a minimum one hanger failure, so the culprit will be likely corrosion of the hangers, or more likely fatigue cracking of the cable due to fluttering and over stiffened connectors at the terminals. Recent hurricane could be the last straw. Fatigue damage to the gusset plates connecting the hangers to the structure is fairly common and was observed on several bridges before. Typical bridge inspection should pick it up, or just notice visible fluttering of the hangers, and investigate further.
With the advance of the computers and design software, anybody could design a bridge - the biggest challenge unfortunately is in the details.

 

[Lomarandil]

That's a good point about the fuel truck wiktor -- unlikely to be over design weight.

As in many parts of the world, Asia is not immune to the push for faster, flashier designs relying on software output and without a lot of engineering oversight. So I wouldn't be surprised by a root cause that combined poor detailing for access/inspection/maintenance and poor execution of that inspection/maintenance program.

----
just call me Lo.

 

[izzotsai]

Sad to see such a disaster happen in my homeland Taiwan..., there are 4 found died and still 2 remained unfounded until now. R.I.P. for the victims.

The authorities said the formal report of this collapse will be published in 6 months the soonest. Some local news reveal that the bridge had been detected with some tendon rusting in 2016, but seems not been repaired properly after that. The root issue still in investigating.

 

[Denial]

Dynamic effects of instant removal

A simple, reasonably accurate and usually slightly conservative, way to get a value for this is to assume 100% overshoot. This approach results in
[ ][ ][ ][ ]P_peak = 2*P_static2 - P_static1

You raise a good point, and the breaking of the first cable would seem to be a classic case of a sudden (but impact-free) additional load on the neighbouring cables.

 

[Tomfh]

Yeah, could be simple progressive collapse. All of them struggling, and then one lets go....

 

[verymadmac]

Interesting how the cables on the left seem to fail at the top and cables on the right fail at the bottom with one exception. I wonder if there was already a failed cable or two, only one cable seems to show tension release, wouldn't one expect at least the 2nd failed cable to show some sign of spring back.

 

[IRstuff]

Not clear there were many failures at the bottom; the second video shows most of the cables still attached when the deck hits the water

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[msquared48]

Almost like the failure of the center cable allowed a pin to develop at the center of the deck, or lost vertical support at the pin location.

This implies to me that the cables were not properly tensioned, or some had relaxed over time, putting more stress on the cable that failed.

Mike McCann, PE, SE (WA, HI)

 

[charliealphabravo]

Hokie's point about inspecting the wishbone connections could also be relevant to the uniformity of tension in the cables. Any change in the shape of the arch might cause tension to concentrate at the ends or the middle strands.

From reports so far, it is looking like the main culprit will be corrosion that was flagged but not addressed.

Again, for inertial reasons, it would not be too surprising if the initial cable failure occurred at the top under near-static conditions and subsequent cable failures occurred at the bottom under dynamic conditions.

 

[LittleInch]

There are a few reports going around saying that inspections had found corrosion, but nothing much was being done.

Interestingly it also seems to be used for a light attraction. Weight is probably negligible, but the action of attaching these and then taking them down might cause damage and flexing. There is a bunch of junction boxes at the base of all the cables which indicates use on more than one occasion.

There is also no protection against vehicle impact and again a few reports saying vehicles had run into the cable supports.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bwing08]

Corrosion seems likely. The curb looks like a good spot for water to pool - I don't see any drainage systems, only some caulking around the cable sleeves.

 

[msquared48]

Interesting view here in looking at the right most exposed cable connection of four small bolts with plate stiffeners.

Seems pretty light considering the possible tension forces involved...

The whole design with a centrally supported superstructure seems highly susceptive to torsional forces from wind and traffic loads, as with the Tacoma Narrows failure, although that failure was solely due to harmonically induced constant speed wind.

Mike McCann, PE, SE (WA, HI)

 

[azcats]

That's got to just be a light pole, no? I think the red ones are the cables.

 

[msquared48]

You are probably right...

Mike McCann, PE, SE (WA, HI)

 

[IRstuff]

This is a pretty good shot of the bridge just before the deck hits the water. The cables on the right look the individual strands snapped from the extra load during the drop, but did not snap like the ones on the left. Interestingly, it looks like the cables are not wound, but simply wrapped with the red stuff.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Ingenuity]

Interestingly, it looks like the cables are not wound, but simply wrapped with the red stuff.

The stays are possibly multiple 'single-strand' stays: multiple, 7-wire individual strands (6 wires around a straight king wire), inside a external (RED colored) HDPE sheath. Today this would be probably done with multiple single-strands however, each one individually corrosion protected (i.e. greased and sheathed monostrands, individually stressed and replaceable), making up a stay group, inside a external sheath.

EDIT: added 'multiple' to define that the stay is NOT a singular strand, but a multiple of single strands.