This post is regarding a job we are bidding that was actually completed by another company where their product was a complete failure. These parts are structural components for a Euro style post and beam construction (Winter garden or Window wall). The current components where constructed with an LSL core and 1/8" thick Qtr. Sapele veneers. The widest parts are 14" and the narrowest around 6" and all are 2" in thickness. What has happened is there are cracks everywhere down the length of the components and the veneers are raising at these cracks. As I understand these parts were laminated during the summer months and the failures did not happen until a few weeks ago with the onset of very cold dry air. So I assume there was a moisture issue here. We have been asked to bid this remake job and am researched methods to do this correctly. Possibly similar construction for structural reasons with assurance MC of the LSL is where it needs to be and the addition of cross banding before face veneers, or a solid sapele lamination would also be a possibility.

Wondering if anyone would have any insight into this situation as far as what would be the best method of construction.

One could always go back to the historical precedent that worked. The solid post and beam worked and worked well. It set the standard for that type of construction.

I suspect that the LSL/Sapele is too far removed from that precedent and the LSL and Sapele will both move differently under different MC. The adhesives in both the LSL and the laminations may be failing as well.

I would never consider LSL unless I had engineering specs from the LSL maker that said it was acceptable in that use, and that someone else (architect, contractor, etc), with financial responsibility, insisted I use it.

Bob, Do you know what type of glue, and clamp method was used on the failed parts, and how the core was prepared?

Everyone knows this is a moisture problem but how to control what you get with LSL material is beyond me. Are these exposed to the weather on one side? Gene where are you?

Moisture problems aside is it possible to use a 1/8" thick veneer at the 14" width? Seems like 1/8" acts more like solid wood.
Joe

I just received some actual photos showing what is really happening. Also the cores are not LSL rather PSL Parallam beams which to me are less stable than a good LSL material.

View higher quality, full size image (2232 X 2232)

View higher quality, full size image (2232 X 2232)

View higher quality, full size image (2232 X 2232)

An additional picture...

View higher quality, full size image (2232 X 2232)

Those types of products are manufactured by compressing the would fibers under extreme pressure. Any introduction of moisture, even seasonal swings, will cause expansion way beyond normal solid wood products. I would use a finger-jointed stave core for stability.

Along with the obvious moisture related problems, I can see glue line failure between the face veneers and the corestock. Also, the face veneers were not edge glued, or the wrong glue was used, leading to failure. Edge gluing is necessary in thicker 'veneer' stock. I would be very interested in the logic that was used to determine this type of construction, and especially the glue used.

I would avoid any use of such 'engineered' wood in the coming rework. I would even question the use of metal brackets that apparently stand in for joinery. There is no substitute for the real thing: true mortise and tenon joinery. Moisture contents and expansion/contraction of all the materials must be matched for compatibility. Glues must be appropriate and not marginal in any sense.

Expectations of the customer may have be adjusted. It is not reasonable to expect post and beam work, exposed to the elements to look like fine cabinetry.

David,
I talked to the original builder about this and according to him he was forced to go this route by the architect because they did not think solid wood was structurally sound. Myself, I would go back and push for solid instead of using garbage like this core.
By the way those curtain wall connectors are very good and part of a well engineered cladding system.

I think you are right about improper gluing.

Bob,
If they want to use a engineered core take a look at the LVL cores made for door work. Supposedly dry and stored inside. National in Denver carries these.

Does anyone know how much movement across the grain if any LVL has? I would still be concerned about going 14" wide with these.
Joe

Joe - Interesting that the architect did not think solid wood was structurally sound How could he end up there?

You are right about those cores, though. Stave built cores of the (untrusted) solid wood, or the commercially available door cores would be a far better choice than that crap.

(Rant...) It just shows how those of us in this business are such bad communicators. That those who know are ignored in favor of some marketing materials aimed at design professionals. Far more money is spent on selling than there is on the actual correct specifications. How is it that the people skilled at putting themselves at the top make the decisions and the most money, while those who know are in the shops or drawing boards, making things that really work, and work well?

To me, it looks like the first and third picture show LVL core and the third picture is either PSL or LSL,

Whatever way you choose to try I would make a four foot long section of the 2" x 14" and then try your hardest to make it fail.

R&D on this project missing?

I meant to say the fourth picture shows either PSL or LSL. I think the first and third are LVL,

I believe you are looking at two different issues. Both are related to moisture content changes after manufacturing. Note that any moisture checks now will not tell us what the MC was at the time of installation, prior to cracking, or even at the time of manufacturing. The defects would result if the pieces were exposed to high MC after manufacturing and then dried out quickly. It would be worse if the core was very dry initially. Careful examination and measurement should potentially allow a better determination of what has occurred.

Gene,
Moisture issues aside I am curious to the expansion - shrinkage properties of LVL door core like in the picture. It appears to be edge glued from thin strips of flat sawn fir. Would this make it behave like fir?

We use this without issue on 6" or so door stiles but have never thought gluing thick skins to wide pieces of this a good idea.

Joe

View higher quality, full size image (1632 X 1224)

Thank you to all who had insight into this matter. The latest developments on this project have put our remake scenario on hold. The clients have decided to fill the cracks and refinish in hopes the cracking will not continue. We are doubtful of that happening so for now we are putting this on the back burner.

The response of the client makes me wonder why we even bother sometimes.