In looking up the Titebond site to see how thety explain two major problems with their glue, I came across an old post in Google from a FHB reader complaing about a massive failure of TBIII on Western Red Cedar.
This problem is recognized by TBIII since it was their glue tech that told me about it. WRC has resins that repel water. In about 20% of current Cedar harvest has a high enough resin/oil amount that TBIII (Or II or I) will not bite and hold. The oil repels the water in the glue. The glue looks as if it did not dry (indeed, it cannot).
What can you do? Wipe the surfaces with acetone until no oil comes off with paper towels. Both surfaces. Thoroughly. What about those 122 mortises? Start wiping….
Obvioulsy, this is crazy. They know this but will not tell us. This ‘part II’ will help shed more light:
We make exterior doors with a 5 year warranty. It is crucial our joints stay tight, especially in the panels. About 1.5 yrs ago, we started to get calls about panels coming apart. Sure enough, they were. Most were not all the way thru, but the exterior surface had open cracks – enough to stick a shet of paper into them. We reaired, replaced, cussed and swore. About $12,000 worth. But….what was the cause?
Again, I called the glue techs at Titebond. All these panels had been glued for width with TBIII under ideal conditions, but these kinds of problems make one wonddr what the hell is going on. One other thing the failed panels had ib common was color – they were all very dark. Another thing in common was an Eastern exposure. The morning sun hit them full on, at almost 90 degrees, and the surface of the door could get up to 190 degrees! Here is where the glue comes waddling in: at elevatec temps, TBIII looses strength. I’ll let you look it up, but it is something like losing 60% of its strength at 175 degrees, and 90% at 200 degrees.
That is huge!! The photo below shows a hairline crack in the exterior surface of the wood. This 1-3/8″ thick Mahogany panel has loost over 50% of the glue line – one can flex the panel on the joint and it opensabout halfway thru the thickness. Other panels had almost 80 or 90% no glue hold.
Why is this not known or discussed? Are we really that much of a red headed stepchild that no one wants to be honest with us?