HOW TO: tie a steel frame system to an unreinforced masonry wall

[FreeRangeMonkey]

I am currently designing a frame to retrofit an existing unreinforced masonry building. The existing building has been in existence for a long time (50+years) and has experienced the expected settling, shifting, etc, which resulted in some diagonal cracks at the corners of opening and at the pilasters. There are existing trusses which are to remain for architectural reasons, but they must be worked around. The building is a 59'w x 83'l x 20'h box. The contractor has already removed approximately 26 tons of ceiling and there is still about 30 tons of roof. The trusses have thus been relieved af about half their load. The wall span is about 19', with a 1-2' parapet.The frames I am designing will work as a hinged frame in the short direction, with 3 spans; the long direction will be cross-braced. I am leaving a 4-6" gap between the back of the framing and the inside face of the URM walls. I am thinking that I want to attach a ledger to the URM walls, and my thought is that an epoxied ledger would provide a LOT of shear capacity without piercing the wall envelope, which dictates a mandatory full seismic upgrade of the entire buidling, which is out of the budget of the new owner. I figured that by attaching the ledger with a lot of epoxy and them creating a bolted clip connection from the frame to the ledger that I could tie the walls together with the epoxied ledger, brace them with the frame and not have to get into a full seismic upgrade.

My questions are:

1) Are there any resources for this type of work...web articles, code referneces, etc.?

2) anyone have personal experience with this type of stuff?

3) are there any issues I am not considering?

4) can you recommend an epoxy system that would work in this application...or maybe a better solution altogether?

Thanks in advance!

Greg Cashen

 

[LUGuy]

FRM,

This sounds like a very interesting and challenging project. I am curious what you mean by epoxying an angle to an existing wall. It sounds like you want to use epoxy only (?) because if you bolted or otherwise penetrated the wall, it would need to be upgraded with consideration of a complete seismic review.

First, what kind of epoxy is good enough that it can replace a conventional hard connection.

Second, if this epoxy is acting as a structural connection to the URM wall, how can you not comply with seismic upgrade requirements?

It sounds like you are trying to find a loophole in the code. Am I missing something?

Jim

 

[FreeRangeMonkey]

YES. I am trying to find a loophole in the code. This is a building that has been in service for many years and the worst it has to show for it is a few cracks in the URM. That is pretty standard. The idea is to basically "smear" epoxy onto the ledger and epoxy it into place. You ask what kind of epoxy is strong enough to resist this loading, and I tell you, almost any kind, including the kind you can find at your hardware store. But I would probably spec a structural epoxy like Simpson ET or SET or something. While in college, some classmates of mine actually did some testing on epoxied ledgers for Simpson for theis senior thesis. It turns out that the shear strength of the epoxy is much greater than the strength of the wood being epoxied. For a vertically loaded ledger, the failure mode is a peeling motion from the top of the ledger down, occurring in the extreme wood fibers.

For the application that I am considering, we have to transfer the existing roof loads to the ground other than through the failing wood trusses which bear on the URM walls. To that end, we are going to install a steel frame system which will pick up the roof load and take it off the trusses. This does not, however, deal with lateral loads, unless we make the roof act as a flexible diaphragm. That can only be accomplished either in the frame roof (flat) or in the wood diaphragm, but the wood diaphragm is existing and is otherwise serviceable, so I would prefer not to mess with it. So that leaves the frame. once that is accomplished, all that is left is to brace the walls so that they are not overspanned. Right now, I think that 20' for URM walls is a bit excessive. My idea was to tie the walls with a slip connection parallel to the wall so that for wind/seismic loads perpindicular to the face of the wall (out of plane) the walls would be braced at every 5 feet or something. In that case, the failure mode of the epoxied ledger would never occur becuase there would never be a vertical load on it, just the bearing of the wall against it...essentially, it gets pushed/pulled between the frame and the wall and is attached via bolted connections to a steel angle on the frame. The failure mode would be bearing failure on the bolted connection.

The reason for trying to avoid the full seismic upgrade is that the review process will be much longer and the cost to the owner will exceed the value of the building by a lot. As it is, the owner thought he was buying a serviceable building and now has to retrofit to some degree, but if he can avoid having to lay out the additional 50k that it will take to do the full seismic upgrade, he may be able to stay in the project. Otherwise, he will likely just sell it to someone willing to take the loss. Unfortunately, even with the building completely upgraded, it does not comp out at what it would cost to do the full upgrade at this point. And since the building officials have basically told him that he can avoid the full seismic upgrade by not penetrating the walls, this seems like the best way to go.

Clearer? Thanks for checking it out!

 

[ERV]

FreeRangeMonkey,

Is it possible to set and epoxy a channel on top of the masonry wall and connect to the steel frame? This should help reduce failure at the connection.

 

[FreeRangeMonkey]

Alas, no. The top of the wall is actually the parapet, which is 1' above the bottom of the joists. The joists are set into pockets in the wall, as are the trusses on which the joists sit. I think that some sort of epoxy solution is the answer. Maybe I will just use Dur-o-wall URM wall repair epoxy to dowel into the wall without penetrating to the outside, but I think that is actually weaker than the epoxy pasted ledger solution, because the pasted ledger will act to tie the brick together, whereas epoxied dowels will concentrate stresses at the bolt locations and then it is a matter of adding dowels to distribute (read: disperse) the stresses.