Hi all-
I had several glue joints fail today and was hoping you could help me diagnose the issue. I am laminating 2-3 pieces of roasted red oak to make an outdoor bench and the laminations were clamped overnight. Titebond III was applied to one side and MANY clamps applied (mostly parallel clamps, about 12 clamps per 36″ bench leg). Every leg failed, either partially or fully (pics provided). My basement shop is climate controlled and the glue is less than six months old.
What happened? How can I fix this??
Nathan
Replies
So, first off, I am no expert, but have read up on various forms of thermally modified wood. The roasted red oak has been torrefied, which essentially is a slow bake at around 500 degrees in an oxygen free environment which has the effect of making the wood hydrophobic to a degree. Unfortunately, this does affect glue bonds that use water as a carrier, such as Titebond or hide glue products. Epoxy or urethane glues have much better adhesive qualities when dealing with torrefied woods. You could try freshly scraping the surface to expose fresh pores and immediately glue it up. Cure time with water based adhesives on torrefied products is generally longer than normal, and you could also try leaving the boards clamped up for 24 hours or more. It could simply be an issue of not leaving them clamped long enough. Anyway, I hope that helps some, and I wish you luck!
I tested the hydrophobic theory but the wood seemed to soak up water no problem. It absorbed it very quickly, much faster than other woods. Maybe the wood sucked all the water out of the glue before it had a chance to cure?
I suppose it's possible, but I suspect the process doesn't eliminate the hydrophilic nature of the wood, just makes the cellular structure slightly more hydrophobic, or resistive to taking on moisture or anything into the cell wall as it has essentially been crystalized. As far as water passing through it, well, red oak is very porous, so I wouldn't be surprised that it drinks it up, but I'd be willing to bet it sheds it pretty darn fast as well. You'll have to let me know what you observe, I'd be very interested to know since I haven't actually been able to get my hands on some. The tests you're doing also seem to be a great approach to learning more of this interesting creature. Glad you have the opportunity to use it, and I hope you post pictures of the finished product! I'd love to see it.
If that is torrefied wood, that is likely the cause. I have never worked with it but I had a local dealer who tried to carry it for awhile and from what I was told it wasn't easy to work with and everything you thought you knew went out the window. I was also told it won't accept finishes either and while it is very weather resistant you pretty much have to just let it weather naturally because stains and varnishes don't work.
I'm thinking either a natural oil (Tung) or milk paint. Wife's choice. Think I could have a problem with either of these?
I was told that no finish had been found to adhere to it but I have never tried to work with it. That information came from my dealer when he showed me a new product he was carrying design for outdoor projects.
12 parallel clamps on a narrow 3' part may have just pushed all the glue out of the joint. Those fat handles BEG to be cranked down. I have roasted wood and glued it up with both epoxy and TB liquid hide glue with success. I have always scraped or scuffed the surfaces after roasting / before gluing.
This is helpful thanks! I didnt put a ton of "crank dowwn" pressure on the clamps, but I did have a lot of them on there. I might try polyurethane glue next, with some scuffing.
Try this: Take some hardwood scraps and rub joint them together on edge - no clamps. Then take some of the roasted red oak and do the same, rub joint them together on edge without clamps. If after the glue dries, the scrap joint is good but the roasted red oak joint has failed, it's the wood.
~~~
The Rub Joint: It's a simple joint, really. A thin film of glue is applied to two jointed boards, which are rubbed together back and forth until the glue becomes grabby. The glued-up boards are then left to dry under gravity and atmospheric pressure only.
I will try this, great suggestion. Standby!
I would sand the surfaces to be glued before glue up. Don't worry about color. The roasting goes through the wood, so it's not like sanding off stain.
Roasting dries the wood out. It then want to reach equilibrium moisture with the environment. So it's sucking moisture right out of the glue.
Always, always apply glue to both surfaces, of any glue joint. The joint will hold better. It's a little like sizing end grain on a glue joint. Yes, many times it's belt and suspenders, and you don't need to. Until the joint fails, and you actually did need to apply glue to both faces.
Thanks for this. I will always apply to both surfaces now! And I think the wood was burnished from the planer. I didn't rough it up at all. Two mistakes I will not make again!
Report back after your tests. There are a lot of elements to this and I'm curious!
Response from Titebond:
-----------------------------------------------------------
Your joints below look like they are wide open. I assume that these are joints you pulled apart? The glue looks stretchy in the pictures which would indicate it did not dry properly.
Note on roasted wood (Thermally modified wood): heating causes the wood to carbonize. The carbonizing gives the wood better resistance to moisture. This carbonizing also causes the wood’s chemical properties to change and decreases adhesion of most adhesives. The higher the heat treating (carbonizing) the lower the adhesive bonding. Also, since the wood is more moisture resistant, water-based wood glues such as Titebond III will dry slower.
Cleaning with acetone is preferred over mineral spirits for cleaning wood surfaces.
If you have questions about the quality of the glue, take some unmodified wood and glue it with the Titebond III in question. Clamp overnight and break with a hammer. It should give a high amount of wood failure.
-----------------------------------------------------------
This is for an exterior application so some of this doesn't apply but instrument builders have been using torrified wood now for some time. I have no experiance with it myself but have read about it. Many instrument builders talk about gluing problems. Gluing things like bracing and bridges to torrified tops have resulted in many failures. Fish glue is claimed by many to be the solution and others say polyurethane glue. Long clamping times are universally agreed on. Almost all say that water-based glues are where they have the most trouble. Some say they have successfully used waterbased glue. So,there you go! This cooked wood is hydrophobic apparently. I have a drought in my backyard. One would think that the water starved earth would just soak it up when you water but it doesn't .If the earth is super dry water just bounces off!
I've taken an attitude to torrified wood. Its my attitude so leave it alone if you dont agree! They cook it at high temperature in a low oxygen environment until they create a fossil ( my take, no science) of wood. The wood looses all the properties of wood and they actually change the cellular makeup. Carbonized they say...we all go back to carbon eventually! This is great for the industry - they are now able to use wood that they formerly left standing because it had little or no commercial value. If you built a deck out of poplar it would last a couple of years at best. They now offer torrified poplar decking. It can't rot because there is nothing in it to rot! A fossil---
Back to glue....outdoors fish glue is out. People say rough up the surface but what would that do if in fact it isn't truly bonding to the wood. It might create enough adhesion to get you out the door but then blow up on you at a later date. We clean oily wood with a solvent to remove enough oil temporarily to get a glue bond. There is no oil ot contaminants present. I think, well I don't think, but since there are claims that polyurethane glue has been successful I would probably go that route....with an extended clamping period.
Yeah it was nonporous. I have glued up 10 samples and will follow up tomorrow on the results.
1. Roasted oak with TB3 (same situation that failed)
2. Roasted oak with TB3 with good prep (sand, acetone, water)
3. Roasted oak with TB2 and no prep
4. Roasted oak with epoxy
5. Roasted oak with old polyurethane glue (gorilla)
6. Roasted oak with new polyurethane glue (Titebond)
7. Roasted oak with thin CA glue
8. Roasted oak with thick CA glue
9. Poplar with TB3
10. Poplar with TB2
I used rub joints on all. Face grain with about 1”x1” glue surface. Not scientific but I will learn something and pass it on to the forum.
Awesome. Looking forward...
I broke my 10 test samples, and the most wood failure (strongest joint) occurred with the Titebond III, which is in complete disagreement with my initial hypothesis (TB3 was a bad choice for my application). The new polyurethane, epoxy and thick CA glue also performed very well for these small samples. The "old" polyurethane glue did the worst, followed by the thin CA glue. I did not notice any appreciable difference between the "prepped" TB3 and the "unprepped" TB3 samples (reminder: prep included sanding at 150 and cleaning with acetone). I have three more large samples for polyurethane, and both prepped and unprepped TB3. I will fill you in on the results.
Having worked with dreaded gorilla glue Ive found you won't get a strong joint without proper clamping pressure. It expands and that same foamy thing that happens outside the joint that cuts like bread is the same thing that will be in the joint if not properly clamped. It likes water to kick off so dampening the joint prior to clamping, and I don't know but might be necessary with with cooked wood, is a good idea. It loves greenwood outdoor construction. Anyway with gorilla glue a rubbed joint won't tell you much.
This forum post is now archived. Commenting has been disabled