Glue Bond Failure in "Mahogany"

Tropical hardwoods may resist glue absorption, leading to failed bonds. Here are helpful tips on prepping the surfaces. December 2, 2006

Question
I recently made a 28" round table top out of Santos mahogany. Two months later the joints are coming apart. The joints were doweled and glued up in clamps overnight. Glue was Titebond III. I've never seen joint failure like this. I've also never worked with Santos mahogany before. I've worked with Hondo and African mahogany with no problem.

I'm trying to figure out how to repair it, but first need to know if I should be using a different glue. Is there something inherent with Santos mahogany that breaks down normal wood glues?

Forum Responses
(Adhesives Forum)
From contributor J:
I do custom stair treads for local flooring shops. I had this problem once. I let the glueup set overnight. Removed clamps and noticed small separation. Broke glue up with little to no force. Looking at the glue line, it was very shiny. I no longer use the jointer to produce a glue surface for this wood. I have not encountered this in over a year. Glue was TB3 and still use it.



From Professor Gene Wengert, Sawing and Drying forum technical advisor:
First, appreciate that this is not a mahogany. It is not related. It is exceptionally hard, much harder than mahogany, and as such, requires exceptional gluing skill and adhesives, as well as a very clean surface. I think that I would prefer a PUR.


From the original questioner:
I've wondered if the glued edges may not have been prepared properly for this particular wood, since it's so dense and non-porous. Still have to wonder why the dowels aren't holding up, though.

Gene, this is news to me. Santos mahogany is not actually mahogany? What would it compare to? I've worked with quite a variety of exotics, so maybe it's similar to something else I've worked with. Thanks to both of you for the input.



From contributor J:
When I removed the clamps from that particular glueup, I was amazed how easy it broke. I would think that the hardness/density prevents glue from making true bond. I still refuse to use the jointer for a glue line. You did not mention if a jointer was used in this operation. Possibly too much speed at the cutter causes burn/burnish, therefore reducing the effectiveness of the glue to make a proper bond.


From the original questioner:
Nope, I don't have a jointer yet. I used an 80 tooth blade on the table saw to shave off the last 1/32" to prep the edges. The Freud blade I use cuts an exceptionally clean, smooth surface, which could definitely be considered burnished.

So that's for sure part of the problem, particularly with this Santos mahogany. I'm still wondering about the doweling failure, though. At this point I'd hazard a guess that the bit I used to drill the dowel holes may not have been sharp enough to drill a clean hole without some burning, which would also block glue absorption and adhesion. As I puzzle this out, I'm starting to wonder why I haven't had this problem before, considering some of the other dense woods I've worked with. I know I'll definitely keep all this in mind for the future.



From Professor Gene Wengert, Sawing and Drying forum technical advisor:
Here is a quick test for you to run. Take the surface you are gluing and put a water drop on it. If the water drop stays there for a few minutes and beads up without dispersing into the wood, you have a surface that is not likely to glue well.

Another tip is to lightly sand the surface to be glued (sandpaper on a block of wood so you do not create unevenness) just before gluing. This assures a fresh, active surface. With your saw preparing the surface in such a dense wood, this "touch sanding" would be very important. This sanding will remove burnished fibers.

Another point: What part of the saw tooth is actually preparing the surface to be glued? The answer is the side of the tooth. Therefore, you need to make sure that your saws have the correct side-dressing (includes both sharpening and alignment), as well as sharpening of the top of the tooth. Not all saw shops side-dress real well.



From Jeff Pitcher, forum technical advisor:
Just a guess here without seeing the actual glue line, but it sounds like you may have let the glue dry or skin over before bringing your pieces together. The shiny glue line is often an indicator of this. This is often a problem when the weather is as hot as it's been lately.


From contributor J:
Santos can be a tricky wood to edge glue. One practice I would think about abandoning is dowelling any edge glued surface. Dowels and biscuits are most preferable on end grain glue ups. You are basically decreasing the glue/wood contact surface area on an edge glued panel and weakening the joint. We glue Santos every day with an EVA/PVA crosslinking co-polymere and have only had one joint failure to date, out of a couple of thousand. Try to glue your panels within an hour of ripping the board. A freshly ripped edge will always glue better.


From the original questioner:
Gene, I tried your water droplet wood test. Pretty interesting! I tried it on a variety of domestics and exotics. Most interesting to me, though, was the test I ran on African mahogany, Honduran mahogany, Santos mahogany, and Bubinga. I included Bubinga in the group because this was the wood I originally wanted for the table, but couldn't afford at the time.

The true mahoganys were the first to show absorption, with the African in the lead, Hondo close behind, both showing some absorption within 15-30 seconds. Both the Santos and the Bubinga took several minutes, almost 20 minutes for the Santos to show any absorption at all. And the Bubinga water droplet was still just sitting there at the 20 minute point. I was surprised that the Bubinga turned out to be the most resistant to water absorption.

I've worked with Bubinga some in the past, but never with an edge glue up, mostly surface glue-ups in intarsia work. I can't help but wonder if I would have had the same problem with the Bubinga, had I been able to use it in this project. Seems so, based on this test. Thanks for the input, Gene. It certainly has been interesting.



From Professor Gene Wengert, Sawing and Drying forum technical advisor:
Take the slow absorption surfaces (called inactive) and swipe with a piece of sandpaper to produce a new fresh surface. Now see what happens. (You can have poor adsorption probably with any species if the surface is old, oil coated, burnished, etc.)


From the original questioner:
I think I'll try that, Gene. I did cut fresh edges on every piece before I tested them, with the Freud blade I've been using for edge jointing. (The one we suspect is burnishing the edges.) So this time I think I'll create two samples of each wood type; one with a fresh cut and one with a fresh cut and a sand. Will be interesting to see if the absorption rate is better with the sanded one.