Delaminating Glulam Beam Repair 1

[Sam165]

Working on a rehab project of a four story multi-family apartment building. The front balcony beam (running parallel to railing) was water damaged to the point of needing to be replaced. Contractor replaced the beam per my recommendation and I was out there doing a QA this morning. While doing my QA I noticed the perpendicular glulam beam that face hangers onto the replacement beam is delaminating at the bottom plies and has a diagonal crack propagating from the over-fastened replacement hanger. The beam is a 3.5”x9.5” Anthony Power Beam that spans 7’-0”. It is supporting perpendicular top chord bearing floor trusses that span 9’ on either end. Live load is 100psf and dead load is 40psf.

I’m thinking of specifying lag screws that are driven from the bottom of the beam to fasten the plies back together. I’m not too worried about the diagonal crack. To figure out the required spacing of the screws, I calculated the max horizontal shear by determining the max bending stress and resolved that into a force couple. I feel like I need to take only a portion of the bending stress that is being resisted by the bottom plies (delaminating plies) but I wanted to get some input from you on how best to proceed. Thanks in advance!

Left face of beam looking into floor cavity.

Left face of beam:

Right face of beam:

 

[3DDave]

That happens when there is deflection - the deflection is done. It's deflected. So gluing an additional item won't change that. I didn't see anyone saying they would first jack all the load off the beam to allow it to be undeflected before any of the repairs, so gluing a gap that currently carries no shear won't magically make it carry shear.

 

[STrctPono]

3D, your assumption is based on the premise that this member is already carrying its maximum load. Live load is transient and can come in at any time.

You are correct, the proper thing would be to jack it, fix it, and then release it. However, this is probably not practical. Easier way is for the OP to see if the current condition meets dead load only and then check repaired condition for full Strength Level loads.

 

[dik]

3DD... with old structures jacking is nearly pointless... the deformed wood acts as a flat arch and the in plane resistance is generally very high.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik

 

[Sam165]

STrctPono, yes of course, good ol’ Mohr’s circle! I definitely need to dust off my mechanics of materials book to refresh my memory!

 

[jayrod12]

Shear flow is a function of bending. But you aren't designing the connection for the maximum tension force from bending, but rather a portion of it. It is the horizontal shear stress at the top of that ply that you need to design the fasteners for.

Can I suggest using SDS screws or GRK screws instead of lag screws? There's no pre-drilling required and are much nicer to work with. Your contractor will thank you.

 

[r13]

Just curious. Can you place the metal truss connecting plate on both faces of the beam, then glue the crack?

 

[JoshPlumSE]

I think your solution is perfectly reasonable. A little conservative probably. If you wanted to reduce the spacing of your repair screws, then you might calculate the "shear floor" at that location and design the screw spacing to resist that force. That's the "theoretically correct" way to determine the required shear. But, your method works and is conservative.

I like the manufacturers suggesting of combining this with adhesive. Though I'm not sure how to spec that out. If there is a "deflection" issue already, then you would want to consider jacking up the beam first. In fact that's not a bad suggestion if you're going to be adding additional loading to the beam as well.

 

[dik]

Make sure you use a clutch that prevents overdriving... on one of the Mike Holmes show episodes... his worker was installing lag screws with a hammer...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik

 

[r13]

When jacking up the beam is considered, I think the beam should be lifted rather than jacking from below, which would close up the gap before applying adhesive, or squeeze the adhesive out after application, and create uneven bond along the beam.

 

[BAretired]

Speaking of a hammer, what about glulam rivets? Usually they are used with a steel plate, but in this case they would be connecting wood to wood.

Whatever method is used, it is worthwhile to jack first to relieve beam stresses, then slowly release the jacks after repairs are complete.

VQ/Ib is the theoretical shear stress at any point in the span. For uniform load, the spacing of fasteners should be variable throughout the span because shear varies linearly from a maximum at the end to zero at midspan..

The total shear force required between the bottom two laminations in each half of the beam is f_b*A where f_b is the average fibre stress of the bottom lamination at midspan and A is its area. For uniform load, it varies linearly as mentioned above.

BA

 

[BAretired]

When jacking up the beam is considered, I think the beam should be lifted rather than jacking from below, which would close up the gap before applying adhesive, or squeeze the adhesive out after application, and create uneven bond along the beam.

I don't believe that is practical. Better to inject the adhesive, then jack to close the gap, then add fasteners.

BA

 

[dik]

subject to the caveat that jacking up may not close the gap, where actual fasteners may...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik

 

[BAretired]

@dik,

Well, I thought about using an adhesive, but rejected it because of the time required to install the fasteners. It would be nice to have adhesive and screws, but I don't know if the contractor could install all of the fasteners before the adhesive set. Maybe it would be better to forget about adhesive.

BA

 

[r13]

BA,

The lift idea was a steal from external post tensioning. Yet it is easy said than done, as it requires the setup of good lifting points. Jacks should be designed to press the wood uniformly, but not push too hard. Just thinking.

 

[STrctPono]

If closing the gap is the question, then that is best achieved by placing intermittent pipe clamps from top to bottom of the beam just after applying the adhesive. Will close the gap and squeeze out the excess adhesive.

 

[dik]

BART... most fasteners do not develop sufficient shear transfer... that's why I like glue...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik

 

[BAretired]

I like glue too, dik, except when I don't. Glulam beams have generally performed well over the years without the need for screws or other devices to hold the laminations together. In some cases, finger joints were used whereby a lamination was cut on a shallow angle and spliced somewhere in the span in order to save material.

We don't know why the bottom lamination separated from the rest. The most likely cause was that it had insufficient pressure during the period the glue was curing, but it is also possible there was something wrong with the glue. If the adjoining surfaces are coated with the old glue, the new adhesive may not bond with the wood but rather with the old glue which has already shown itself to be unreliable.

If glue is to be used, it should not be relied upon to carry horizontal shear[highlight #F57900], which is largest in the outer laminations because they are the most highly stressed in bending.
[/highlight]
EDIT: After looking at r13's sketch, it occurs to me that the last comment (highlighted in orange) is wrong. Shear stress is maximum at the neutral axis.
BA

 

[r13]

Shear and bending stresses in a beam, from text of mechanics of materials.

 

[r13]

Can this works without nailing? BA has a valid concern regarding the condition of the existing adhesive though.

 

[dik]

Correct... with rectangular beam it is a parabolic distribution... with average shear * 1.5 to give you the max parabolic stress... reason for the 1.5 factor in wood shear code. About 30 years back, in Canada, there was a bunch of delaminations... used to know the reason... but have forgotten...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?
-Dik