Aliphatic resin gives me the creeps!
Heat Testing Glue Joints
© Frank Ford, 4/25/98; Photos by FF, 1997


I'd heard so much discussion about the various qualities of glue joints that I finally decided to do a little testing of my own. As you may know, my goal is to make luthiers aware of the tendency of aliphatic resin and white glue (I tend to use Titebond as a generic term) to creep under even mild heat.

My testing methods were not as precisely controlled as I might want before publishing a monograph in Fine Woodworking, but I think I got some pretty clear results.


I glued and clamped a plain lap joint in some clear heart redwood. In all the samples the joint measured the same length and width. Then I made up a very simple testing fixture which held the joint vertical as I loaded the end with lead bricks. I have these lead bricks which weigh in at about 3 lb. each, so I simply clamped one on to the end of each stick:

Pretty simple, eh?

I tested the samples at room temperature and found that if I loaded seven pounds on each, the joints would break immediately, splitting off wood fully across the face of the joint. I'll talk more about that shortly.

OK, on to the heat stress test. My kitchen oven's lowest setting is very nearly 200 degrees Fahrenheit, so I chose that for my test. When I can sometime, I'll run this test at about 150 degrees. I used a good standard thermometer to calibrate my oven setting over a couple of hours, so I'd be sure it was very close to 200 degrees for the test.




I opened up the door, slipped in my fixture, and waited. In only a few minutes I heard a loud "clunk" as one of the weights hit the board:

Within seven minutes the Titebond joint failed utterly and cleanly. (By the way, I did repeat this test with the same results.)



Then I figured I'd measure how long the hide glue would take to fail at 200 degrees:

That's right, it didn't. After waiting all day (Mind you, I was doing other things) I finally couldn't stand it and had to try a higher temperature, so I cranked the oven up. I had to stop the test with an oven setting of 375 degrees because the wood charred and my smoke detector went off!


So you could say that hide glue resists heat better that Titebond and its aliphatic cousins.

But if hide glue doesn't melt at 375 degrees, then how come you can take joints apart with heat? Well, I think it works like this: If you're stressing the joint nearly to the breaking point, then hide glue gives up in heat as well. There are no joints that are under this stress in a normal stringed instrument, because just the slightest touch and they'd explode.

When we're prying joints apart, we're stressing the immediate area to the max, and if heat is applied, then hide glue joints will part.

I repeated my test just like above but loaded the joints with two bricks each. All it would take is a tiny bit more and the joint would break even while cold. In fact, I had to be very careful in setting the board down on the oven shelf because a slight bump would cause the joints to break. That happened twice as I was running my tests.

Within 5-7 minutes, both joints failed. In fact they failed at nearly the same time.


Look at the results:

Titebond failed utterly and cleanly. The hide glue joint started to lose a little strength and then the wood broke. The joints looked the same each time I repeated the test.