Shrinkage in deep glu-lam to steel connection

[EmilC]

Hello all! Long time reader, first time poster.

I'm looking at a connection in which a deep (36") glulam timber tees into a W beam, with steel deck and slab over both. By my math, this timber will shrink by roughly 5/8" from its as-manufactured moisture content of ~15% to in-service MC of ~8% (conditioned interior space in a institutional building). Having T.O. timber drop 5/8" below T.O. steel seems to me like a serviceability problem with the slab in the corridor above. I'm trying to come up with some options for the architect here; do any of you know of a way to mitigate this?

I'm looking at some of the fancier timber hanger products out there and wondering if maybe there's a way to suspend the timber from a point other than the bottom, but so far that seems like a red herring.

Thanks!
-Emil

 

[EmilC]

jayrod, you make a good point- glulam mfg tells me that 12% MC is average out of their doors (and we should be able to keep it that way), which is better than I had in my head. We typically assume wetter stock, but this project won't be sitting around in anybody's yard for long. Unfortunately, I am the glulam sub and detailer of the steel-to-glulam connections on this project, so I am coordinating!

Re connecting higher on the member- some of the nice European hangers (Knapp, etc.) seem like they restrain the timber about its centroid rather than from the bottom like a regular saddle hanger, but I don't have anything to back that assumption up.

I think I've got a good quiver of options here, thank you all for adding to it! Time for a RFI to the architect.

 

[BridgeSmith]

It's been a long time since timber design class, but generally how deep you have to go on the beam to support it would be dictated by how much of the cross section is required to resist the applied shear, wouldn't it?

 

[jayrod12]

HR10, correct. Although glulam is a tricky animal since its lamination laid flat, I haven't looked to deeply into their shear failure modes from knife plate connections, but I assume the plies splitting apart would be a concern.

 

[EmilC]

My understanding is that supporting timbers from points other than the bottom is a tricky art... NDS has rules for notching at supports, but glulams are only permitted to be notched to a tenth of the depth at most in the NDS. There are tricks like using fully-threaded screws as reinforcing, and techniques for calculating the additional stresses on those fasteners from resisting shrinkage (My-Ti-Con has some whitepapers). With the lower shrinkage estimate, I could look at that more... it didn't seem plausible with my initial ~5/8" shrinkage estimate to be using a connection that would be restraining shrinkage with the same fasteners that are holding up the beam.

The Wake Tech pedestrian bridge collapse a couple years ago was caused by notching glulams at supports without regard to splitting, it's a brittle failure mode and definitely something to watch out for. My understanding is that the member can still fail even though the x-section at the notch is sufficient to resist the shear, if the member isn't able to develop the tensile stress across the grain required to concentrate the shear in the upper portion of the member.

 

[bones206]

My local airport has a gorgeous glulam-framed roof, with steel framing at the perimeter. I was standing in the security line last week (staring at the ceiling as engineers tend to do) and noticed that there were in fact ~1/4" gaps between the roof and timber beams extending about 3-4 ft from the connection to the steel beam. In this case, the serviceability of the roof was probably not adversely affected, but it does show the shrinkage gap phenomenon does happen.