These failures are typical block shear failures. Timber failed in combination of shear and tension in these connections. Might not be the main reason but certainly something to look at.
In the animation posted by Sym P. le on the 22nd, the two sections of the lower chord are rotating about the fixed point that is the doweled joint. That looks like an excellent way to cause the failure described by molibden.
When there's that rotation, there are sideways vectors on the dowels that tend to pry the wood apart, along the grain lines. Add the tension, and whooops......
I suspect if the bridge had been designed the "olde" way, where members are only in tension or compression, the bridge would still be standing. Then there would have been bridge shoes atop the masonry pier, and that bending would have been tolerated.
Very good article really which also has no smoking gun other than it shouldn't have fallen down so suddenly.
No one seems to be interested in the lopsided load pattern given the car / truck lanes are offset to one side to allow for the fairly large single foot path / cycle way. Or the rather small bearing surface on the end points.
But any settlement or long term creep in any of those joints or beams will lead to overstress at the hard points at the support locations.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
Just thought those of you interested in wooden bridges should see THIS one. 1282 feet length, built in 1901. Wide enough for two modern vehicles (SUV size, not full size trucks). Capable of carrying several vehicles - each way - at same time. Still in operation. Location is Hartland New Brunswick Canada. Longest covered bridge in the world. Those concrete support piers are newer than the bridge, by the way; original piers were wooden as well. Reason for going with concrete was the ice buildup during the spring thaw (river runs from left to right in the photo).
Converting energy to motion for more than half a century
A few bits from a non-structural guy (mechanical):
From various pics, the cross beam resting on the stone pier has a very light connection to the lower longitudinal beam. The beam must be wider than the pier to accommodate 2 1/2 lanes vs one on previous bridge.
One would expect some steel of the cross beam to be below the longitudinal beam to take downward force. Once a load gets off of the pier beam, moves fwd, and starts putting load on the cross members (terms might be wrong), any weakness in those joints would result in a downward force at the connection between pier crossmember and lower longitudinals.
It does look like all the longitudinals (upper and lower, both sides) separated from the pier cross member and dropped, taking the deck with it.
In fact, post collapse, seems the pier cross member is nowhere to be seen.
And 12mm studs/pins/whatever passing through those joints seem incredibly light to me. Not so much that they would fail in shear, but crush the wood adjacent to the pins. Or at least worry the wood over time with cyclic loading.
That steel truss bridge built in 1895 sure looked spindly. But it lasted over 100yrs!!!
Depends on the country how much structures is included in the course.
Baltic States it seems to be an art degree with little structural content apart from running building codes. Well the ones I have talked to are not very up on it.
Apart from one who had level 8 license and part of that qualification was a post grad course in structures. He did know what a second moment of area was. The licensing stuff is something to do with the type of buildings they can sign off. He can do hazardous, library's, hospitals and the like.
New from Josh at BuildingIntegrity #tretten #trettenbridge How Did the Engineers Miss This? Sep 12, 2022, a video that consolidates current knowledge of the collapse. Looks closely at joints of the type used here.
Wood bridges have been working reliably (mostly) for a VERY long time.
Explore a Howe truss, which is typical of a wood truss bridge design. Find lots of wood in tension (yeah, there's the bottom chord. What else)?
IF you're going to use an unproven design, it might be wise to build a test bridge first, instead of ASSuming. I believe they do that a lot with airplanes.
Watched the whole thing and I can't say I'm all that impressed.
The first half of the video is an indictment of wood connections. He shows a bunch of bolt group tension tests in what looks like plywood, so not exactly relevant, but I get what he's trying to say. But to go on and talk about block shear failure as though it's some new and surprising phenomenon? Not so much. Standard check in any connection. "They wanted to break here [draws line across gross section at last row in group] because it would be stronger." Obviously. If that's stronger, then it'll fail in block shear first. If the block shear pattern is stronger, it'll fail across the gross area first. That's connection design 101 for any material dealing with fastener groups in tension. You check both and figure out which one controls and make sure your connection design loads are less than the lowest failure mode capacity (taking into account appropriate load factors, resistance factors, safety factors, or whatever factors your design methodology of choice uses).
"This is a tension member all day long." (I'm paraphrasing a little - but he said the member is primarily a tension member.) No, it isn't. Even the free model way up there in this thread shows it's a compression member. Can it see tension? Sure. But it's primarily a compression member. I will admit that I agree - it looks like a tension failure. The most plausible explanation I can think of for that is that it failed not while the truck was right there, but rather when it was on the first span and the passenger car was on the third span. That would have put the center span - assuming the loading was sufficient to overcome dead load - into negative bending and that member would have gone into tension. That alone is unlikely to make the bridge collapse immediately - there is a fair amount of redundancy there - but once the truck moved over the bridge and it needed to go into compression it wasn't there - hence the dramatic bending failure in the top chord over the support.
Can't treat glulams/treatment messes with the glue? Sure, if you don't know what you're doing. But there's a whole section of that market that has been working for some time to ensure adequate treatment of exterior use glulams. AWC Tech Note S580D
Glulam in software as a homogenous member? Not exactly. Different programs do things differently, but the ones I have use equivalent properties that take into account the layup and anisotropic nature of wood to the extent it makes sens to do so. But those programs also don't design connections in wood. The ones that are used to design connections in wood absolutely take all of those things into account. That is, of course, assuming the engineer isn't doing their job and just trusting the computer. If the engineer is doing their job, they're checking that stuff and using the computer as a tool to automate it.
Talking about leaving all the steel exposed as a bad thing? Has he not heard of weathering (trade name 'Corten') steel? Popular in the bridge industry for its ability to develop a patina of corrosion products that seals the underlying material from further corrosion? I'm pretty sure he mentioned it at the beginning of the video. Also used in the manufacture of fasteners and hardware. I can't say for certain that the pins are weathering steel, but I'd be surprised if they aren't.
I'm not saying that this was a good design, or that the engineers focused enough of resiliency in the design. Or even that they had a clue what they were doing. There simply isn't enough publicly available evidence to make any of those conclusions yet. But I find the attack on the concept, the materials, and the engineers without providing any real proof of what decisions were or were not made, materials used, etc. and only relying on pieced together pictures from half a world away and a whole 'double lorry full' of assumptions to arrive at this opinion and broadcast it to thousands of lay viewers who trust you to give a good, honest, unbiased assessment is...irresponsible.
Edit: Also...who hits the breaks when the bridge is collapsing? I'm getting the @*&^#%$* off that thing! I wouldn't be surprised if it started to go while it was near mid-span and he gunned to it to try to get off. Haven't seen an interview, though, so it's impossible to say either way.
Problem with Corten, and I've seen it often is that it corrodes and leaves an ugly 'rust' stain which would make the connections look ugly. Because the corrosion layer is relatively 'tight' and bonded it can have problems if the oxide layer is removed, such as by flowing water. This can enhance the corrosion effect... seen that happen,too. The guiderails for the Manitoba floodway hydraulic ram were originally Corten. Parts of them had nearly disappeared.
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates
Great post pham... I was not as impressed, either, and left a note about the compression element on his blog. I did't read your post until after I posted... I generally post before I read the thread... I know I shouldn't, but I do. I've not experienced it, but I can see the splitting (almost like block shear, without the tension; wood without occulusions, is extremely strong in tension.) occuring and you should be able to design to reduced stresses... just like designing for flexure or shear. I've never used that... you have to be really good at drilling holes in wood.
I wasn't thrilled about cladding the top in copper, and I didn't realise that's what was done. It may not be as bad as implied. Trapping moisture is a problem... copper due to it's anti fungal properties is probably the best material to use. I've not looked into that type of cladding. The copper cladding appears to have 'drip edges' to prevent a backflow underneath the cladding, and is formed in the middle to reduce oilcanning, I suspect.
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates
Thanks, dik. Wood does have a brittle failure mode in tension...but then it also has a brittle failure mode in pretty much everything. It's a design issue that must be overcome through competent design.
Your comment about Corten and staining - I noticed on a few of the pictures above some staining around the connections. Could indicated Corten was used there. But, again, I'm pretty sure none of us know that one way or the other.
It looked like the cladding was only over the top, which can be beneficial. AWC recommends it for exterior glulams here in the States - particularly if they're installed on the horizontal - to keep water from sitting on them and penetrating the wood/connections. The key is, as you said, to keep it from getting trapped.
![[ponder]](/data/assets/smilies/ponder.gif "[ponder] [ponder]")
