Full Fixity Connections in Wood

[bigmig]

I have been in discussions with an architect and builder from another location of the country (one with less of a snow load) who swear up and
down that they do a flitch plate connection between two sloping beams all day long for their entire life. This is a center steel plate, sandwiched by two sloping
members that meet at a ridge, which are then bolted up the wazoo.

There years of experience and where they are from makes them the experts and I am an idiot I guess. But I am the one stamping it, so we will go with my
neck hairs for the moment.

I have looked through every guide I could find, have diligently browsed as many eng tip posts on the subject and am not convinced.
My take so far is that:

1) it is hard to install a fixed connection in wood because absolute quality control is the only way it as any chance of even remotely working. I cannot be there while they are doing this, so I am just not confident it will be done right.
2) if ANYTHING is off or loose, the connection will sag out.
3) the wood basically changes dimension with humidity, which is an issue if the weather ever changes
4) their is the cross grain split out where the system wants to flatten out, and in doing so, pulls the nearest fastener to the end grain right out.
5) there are no full fixity connections in wood....period. Which translates into "how much sag can you live with". This seems like a very philosophical question.
For me and my insurance, the answer is "none".

Any input on this would be of help.

Thank you in Advance.

 

[KootK]

1) This is something that is done but usually, I see it done in one of these ways:

a) The flitch plate extends the full length of the beam or;

b) The flitch plate extends many multiples of the beam depth on either side of the joint.

2) Where #1 is the situation, these moment connections are made possible by spreading out the lever arm of the moment connections by a great distance. When we say that we don't believe in wood moment connections, what we really mean is we don't believe in localized wood moment connections.

3) An article for you on this: Link.

4) Member XR250 is quite adept at this. If he materializes here, take heed.

 

[jayrod12]

I've seen "full moment connections" in wood, but they often are of steel flange plates with lots of small diameter fasteners to minimize connection slip. I've seen some glulam ones where they epoxy in rods to provide tension continuity.

with a bolted flitch, I feel it's essentially your number 5. I guess with an appropriate end distance and enough bolts, it likely is to work. However you'd need to ensure the wood was fully kiln dried, not that soggy shit I see all the time that at one point was kiln dried but has since been left outside. Also tight fitting holes would be necessary, which in my experience is also something the contractor tends to screw up so they just ream the hole in either the wood, or the steel bigger to make it work, introducing more slop in the connection.

 

[Celt83]

5) there are no full fixity connections in wood

This all day long. However, that does not necessarily mean you cannot provide a connection with some rotational rigidity or that your structure actual requires fully fixed joints to approach that performance level.

Which translates into "how much sag can you live with"

Nope, it translates to how much joint rotation you can live with which will impact overall member/structure deflections.

The biggest challenge with providing rigid connections in wood is determining all the sources of "slip" in the connection and then actually quantifying what the slip is to determine a rotational spring stiffness to model with.

You could model the joint fixed and design for that moment then run the model with some rationally determined rotational spring stiffness at the joint and compare the results.

I'm making a thing:

http://www.thestructuraltoolbox.com

(It's no Kootware and it will probably break but it's alive!)

 

[phamENG]

You mean something like this: Link?

I don't have a problem with it as long as you're not designing it as a real flitch beam, but rather a steel plate beam that is stabilized by connections to the diaphragm through the wood that's fastened to it. If possible, I'd want the steel to be full length (no splices).

If you can't, I'd be okay with an analysis that splits the fixity into a pair of fastener groups in shear. Bolt slip is still an issue, so you may try to use shear plates or have the holes drilled in the wood exactly the same diameter as the shank of the bolt. Or, if you can get away with it, put the steel on the outside and fasten through with screws - like Simpson's Combo-Head SDHR screw.

 

[phamENG]

Man. That took me too long to type up...everyone else beat me to it.

 

[XR250]

I agree with KootK's insight. The farther apart your connections, the less bolt slop will affect "sag"
I would drill all the holes the same diameter as the bolt. For assembly, they can drill thru the holes in the steel into one side member and then drill back thru this hole for the opposite side member. That way there is no "reaming out".

 

[bigmig]

phamENG, it is sag and there is no "nope" in front of that. I understand that sag is caused by the slippage and partial fixity is the question. The bigger question is
how far can it deflect (sag) before the sheet rock splits out and I get a call on the 24th of December. In other words, the real world, what matters to the client is the sag in their ceiling under the design load. From what I am gleaning from the responses, the more spread out tiny fasteners you have, the higher the chance of things working because it means that 80% success during install means you have some factor of safety just in the nature of mulitiple fasteners.

I like the full length flitch plate. The issue is that the loads are high enough that the flitch plate it would take to do this is substantial (like possibly a pair of 1 inch plates)
which puts us right in the realm of "just do a steel frame" to begin with.

 

[KootK]

From what I am gleaning from the responses, the more spread out tiny fasteners you have, the higher the chance of things working because it means that 80% success during install means you have some factor of safety just in the nature of mulitiple fasteners.

That's not how I see it. XR250's excellent installation advice aside, imagine that your fasteners will have the same amount of slip no matter what your fastener configuration. Apply that amount of slip over the lever arm of a 48" connection, and you'll get one estimate of sag inducing joint rotation. Apply that same amount of slip over the lever arm of an 8" connection and you'll get a great deal more sag inducing joint rotation.

phamENG, it is sag and there is no "nope" in front of that.

I feel that the most likely serviceability issue is the cracking of ceiling gypsum joints at the knuckle. And I would expect that to be governed by joint opening at the knuckle rather than vertical sag. In this respect, I feel that this system excels because the steel plate ensures that the knuckle joint will not open up appreciably until that joint fails in steel plate, M/Zx bending.

The issue is that the loads are high enough that the flitch plate it would take to do this is substantial (like possibly a pair of 1 inch plates)
which puts us right in the realm of "just do a steel frame" to begin with.

I'd probably start with a ballpark design of the plate thickness at the knuckle. If the result is a stupid thick plate, move on to something else.

 

[HouseBoy]

FWIW... I've designed the type of steel "flitch frame" that is illustrated in KootK's first reply many many times over the years.
I do not like the idea of 2) 1" thick steel flitches. When I come up with that much section needed, I usually will utilize a channel or wide flange section (with a full pen load at the peak).

As pham suggested, I tend to make the steel components run the full length or nearly so. I prefer to use LVL for the "sandwiching" material when I need to stop the steel core pieces short of the bearing.

I don't think there is a huge dependance on the wood to steel connection to resist the major bending moment since there is not wood continuity at the peak. This is different from the behavior of a simple span flitch beam.

BTW - USUALLY I can convince the client that a wide flange shape is going to weigh a lot less than one or two 1" steel plates PLUS there will be many fewer bolts to contend with. That and the lower dependance on the bolted connections (since the wood blocking can be fit "tight" between the flanges) generally carries the day.

I'm curious if your architect and builder are referring to something with a full pen weld or if instead they are thinking of some sort of bolted connection (where tow plates are lapped at the peak) that resists to moment. Just wondering based on the terminology you used in your initial post.

 

[molibden]

There are really stiff connections in wood. But you need bigger sections of members and appropriate connectors like long screws with full thread or glued in rods.
You can make moderately stiff connections with bolts and steel plates but you also need bigger sections as normally and you need to control splitting (best to use screws with full thread).

 

[dik]

or revise it and use a ridge beam. I forgot to add, I really like using adhesives, too.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik