Temperature, Surface Prep, and Glue Joint Failure
I glue up straight and bent laminations with powdered PPR glue (in this case Dural brand marine glue). My logic is that a stiff glue line won't creep and show gaps over time like pva might if there is internal stress in the wood. Is this necessary? Should I just use pva glue, or consider polyurethane glue?
When working with PPR for laminations, I mix glue according to instructions on can, let it slake for about 15 minutes. Then, I do a double spread with a short nap roller, getting an evenly glistening surface. I work as fast as possible, usually no more than 15 minutes. Once clamped (in this case pipe clamps both sides, every six or so inches, clamping 8/4 cherry, five pieces, at four inches wide, five feet long), I make a tent out of poly and set an electric heater with a fan inside to crank the heat. (I've wondered if this is a good idea to blow warm air at the wood for so long)I leave it for minimum six hours.
I should mention that the glue was new. I use a card scraper and lightly scrape the jointed surfaces to get rid of tracks from nicks on my jointer/planer. The wood tested at 6-7% mc. The wood came into my shop cold, was machined over about two hours, and then glued up. My shop fluctuates in temp in the winter - thus the heating and hoarding. Some of the cut-offs I tested refused to break at the glue line, while others nearly fell apart in my hands. All the joints had a fairly large bead of squeeze out.
Does anyone have any ideas? Maybe the wood was still cold, and insulated itself from the heater? Could it be a bogus batch of glue? I'm stumped, and now gun-shy to use this stuff again.
From contributor J:
Dural's website doesn't specify a minimum application temp, but the temperature issue sounds important to me. I doubt that two hours of machining time would be sufficient to thoroughly warm 8/4 lumber even if your shop was at a good temperature for gluing, and since you built a heating tent for the glue-up I'm assuming that the shop was pretty chilly. I can't see cranking the heat inside a tent after clamping such a massive glue-up together as being very effective. If the glue's curing is retarded by cold then the water that's necessary for the chemical reaction has plenty of time to soak away into the wood before the reaction is complete. At the same time, the extreme, 'cranked' heat may be causing extra stresses in the assembly.
If the air temps in the shop are down around 50 degrees (or lower), I think you'd be pushing your luck with almost any type of glue. I keep my shop thermostat set to 55F even when I'm not there, because otherwise it takes too long to warm up the materials to the point that glues and finishes work the way they're supposed to.
You said that some of the joints seemed solid while others nearly fell apart. Did you notice any pattern in which fell apart? Were the ones towards the middle of the glue-up more or less fragile than those towards the outside?
From contributor G:
I don't think the temp has much to do with it. I believe the scraping you did to the wood glazed the surface making the surface to smooth for the glue to get the adhesion it needs to hold. After you scrape the lines from the jointer (unnecessary I believe unless it was horrific) you should have roughed the surface slightly with sandpaper.
From Jeff Pitcher, forum technical advisor:
I just went through a similar scenario with a customer and we determined that the problem was almost certainly a combination of temperature and moisture content. It's a condition we call "dry out" and Contributor J described it quite well in the last post. UF resins need a minimum of 70 F to cure properly and using cold wood is a sure fire way to be on the ragged edge of this even if your ambient temperature is ok. With an MC of 6-7% the water in the glue is going away before a cure takes place.
From the original questioner:
After reading the responses, and re-reading my own posting, it does seem fairly obvious that the material wasn't at the proper temperature throughout, and that fanning hot air on a thick block of wood is going to do little to heat the inside of the glue line. From now on, I'll get the entire shop cooking before I begin, and give the wood plenty of time to reach temperature.
As for the use of the card scraper, I should have elaborated: I use it to get rid of ridges from nicked knives, and to reduce/eliminate the corduroy texture left by the planer etc. I may be wrong, but I was taught that a smooth planed surface is better for gluing than a sanded surface because it has fewer torn fibers, and creates better specific adhesion as opposed to mechanical attachment. I would be curious to see if there's a consensus on sanded vs. planed glue surfaces.
After reading the archived topic “getting a good glue joint on cherry” (nearly identical failure) and conducting a water drop test, it was conclusive that the sanded surface absorbed much faster than a jointed surface, and both absorbed faster than a scraped surface. It seems the card scraper leaves the surface too burnished for proper glue absorption. I believe it was a combination of poor temperature control, and inappropriate surface prep.
From Gene Wengert, forum technical advisor:
I agree 100% with Jeff and would add that if you used too much pressure (you did mention the large amount of squeeze out; pipe clamps can be over-tightened easily) then with dry wood (which is common in the wintertime), the moisture left in the small amount of adhesive left in the joint will rapidly disperse before the adhesive can cure properly. Cold adhesive (or warm adhesive on cold wood) will also not spread well into the small nooks and crannies in the wood leading to poor joint strength.
Would you like to add information to this article?
Interested in writing or submitting an article?
Have a question about this article?
Have you reviewed the related Knowledge Base areas below?