Glulam Beam - Checking Problem

[BigGG]

Engineer also mentioned that the glue in the glulam does NOT transfer shear, per their calculations. The wood fibers transfer shear while glue adheres wood together but for calculation purposes, they do not use the glue to transfer shear. The beam never de-laminated but experienced checking in the wood fibers.

That sounds strange. If the wood/adhesive bond line doesn't provide shear resistance, you might as well stack the laminations dry, allowing them to slide across each other.

You keep repeating 'the beam didn't delaminate' but, while that may be technically correct, one of your pictures shows a feeler gage going in full depth very close to, if not on, the glue line for what appears to be a substantial portion of the length. And the allowable checking, shown in a later post, is exceeded so remediation is required.

Does the beam currently show any sag? Does the remediation procedure include jacking to remove some load while the screws are being installed?

 

[Pettrix]

That sounds strange. If the wood/adhesive bond line doesn't provide shear resistance, you might as well stack the laminations dry, allowing them to slide across each other.

You keep repeating 'the beam didn't delaminate' but, while that may be technically correct, one of your pictures shows a feeler gage going in full depth very close to, if not on, the glue line for what appears to be a substantial portion of the length. And the allowable checking, shown in a later post, is exceeded so remediation is required.

Does the beam currently show any sag? Does the remediation procedure include jacking to remove some load while the screws are being installed?

I think what the engineer was saying is that when they do load calculations, the glue is not given a "value", only the wood itself is given a load calculation value. Of course the glue is key to bonding the beams but the glue itself is not given a load value when calculating for shear resistance. They calculate the loads/strength using the boards. This is probably done as an extra safety factor in designing the beam size. If the glue was given a shear transfer value then they could reduce the size of the beam but they error on the side of caution and oversize the beam.

When the feeler gauge goes in, it does not go in straight, it will turn up. In other words, the check is not a clean line but rather a jagged line with wood fibers that prevent the feeler gauge from going in straight. A de-lamination would result in a clean break in the board, right at the glue line. There would be no broken wood fibers and the feeler gauge would go in cleanly on a straight line. The checking is a jagged line that contains wood fibers throughout. If it was a de-laminiation (wood glue) failure, that would be a totally different issue and fix.

There is zero sag in the beam. No jacking is required when installed the screws per the engineer's calculations. The beam did not fail or break.

I am not an engineer but I can't see them cutting corners as their liability is extremely high. Engineer's will always error on the side of conservative caution to prevent any future issues. They don't want a roof failure and be held liable for providing a fix that was doomed to fail. The engineering repair was stamped and sealed by a state licensed engineer that is well-versed in glulam beam repairs.

 

[rScotty]

Engineer also mentioned that the glue in the glulam does NOT transfer shear, per their calculations. The wood fibers transfer shear while glue adheres wood together but for calculation purposes, they do not use the glue to transfer shear. The beam never de-laminated but experienced checking in the wood fibers.

By not including glue shear, he is making the assumption that the glue lam can be considered as one solid piece of wood. That works as long as the glue doesn't shear, AND as long as no deep cracks form.

An analysis might not include glue, but it has to include deep cracks - even if he doesn't want to do so. Reason is that when a crack goes through the beam deep enough, then what you are now looking at are two stacked beams, not one solid beam.

You are probably concerned more with deflection than simple shear.

 

[LD1]

Yes, the glue isnt accounted for because they are "assuming" the glue is at least equal to if not stronger than the wood. Therefore they are treating it as if it were one solid beam....without glue.

And yes, proper glue should pull wood fibers apart before the glue separates

 

[Pettrix]

Yes, the glue isnt accounted for because they are "assuming" the glue is at least equal to if not stronger than the wood. Therefore they are treating it as if it were one solid beam....without glue.

And yes, proper glue should pull wood fibers apart before the glue separates

That is one thing the engineer emphasized is that the glue did not fail as the checking is in the glue fibers.

 

[Pettrix]

Is the engineering fix a proper fix? Or do some people here think it's not?

 

[ning]

Is the engineering fix a proper fix? Or do some people here think it's not?

I think you'd have to engage another engineer to get a good answer to that.
It's good that you've got all this documented now in case there's failure going forward, but I'd bet that no failure will be compensated fully for the resultant stress (between now and the eventual failure) and then dealing with the results of the failure at the time it happens (if it does, of course; I'm not saying it will, just saying that I doubt you'll be well-compensated if it does).

Damn tough situation.
Makes me wonder if you could somehow get a piece of steel U-channel under that beam from one side to the other as backup.

 

[ponytug]

Is the engineering fix a proper fix? Or do some people here think it's not?

I have no idea if it is or not, but you have the advice of the manufacturer's engineer, written, and documented. If it were me, I would do as they suggest, and document that, for yourself and them.

Then I would forget about it. Between the original engineering reserves, and your repair, I doubt that your actual loads will ever present any problems.

All the best,

Peter

 

[Pettrix]

 

[Pettrix]