Q. I know you have probably written about this before, but I ignored what you wrote as it was not a problem until this past year when the switch was made to water-based finishing. The problem is that immediately after finishing, we can see the glue lines rise up. It is a thin little bump -- hard to feel but easy to see when the light is right. It is not continuous along the length of the joint, most of the time. It is worse with denser species. We have conducted tests involving sanding just before finishing (might have made it worse in fact). We waited five days before finishing after edge gluing. And so on. Can you help?
A. I am going to assume that you are perfectly accurate when you say the glue line itself is popping up. If it was the glue line and the wood immediately around it, then the answer would be different. Also, it is a key point that you notice this immediately after finishing.
My diagnosis is that you have a glue line that is a bit too thick in spots...maybe 5 to 10 mils instead of 2 to 5 mils. When you sand the panel, the heat causes the adhesive to expand; wood is not affected by heat.


Find more Wood Doctor at FDMCdigital.com

Gene Wengert, aka The Wood Doctor, troubleshoots wood related problems, and explores lumber and veneer qualities and performance, species by species, in Wood Explorer, inside FDMC's Knowledge Center.

With a thinner glue line, there would not be enough glue to expand that much. Likewise, the heat from the finishing ovens will soften the glue and make it easy to bulge out of the joint.
So, what would create pressure on the joint so that when the glue is softened, the pressure would be large enough to push the glue out? Answer: The addition of moisture. High RH can do it in a panel days or longer after manufacturing. In your case, the moisture is coming from the base coat which is water-based. This effect might be aggravated by the presence of drying stresses (also called casehardening).
So, what can you do?
First, when you find this glue ridge only in spots, as you reported, then the glue line is a bit too thick in spots. In other words, your surfaces for gluing are not perfectly flat but vary by 5/1000 of an inch or so--saw wobble. I would suspect that with the denser woods that the heat that is created when sawing causes the saw to move a few 1/1000ths of an inch. A saw is designed to avoid wobble, but in addition to heating, you might have the wrong saw or a dull saw. I prefer high speed steel as it can be sharpened better. To check this saw wobble effect, you could joint the edges after sawing to get them perfect and then see if the resulting joints are better or defect free.
Next, it is the sides of the saw’s teeth that actually touch the surfaces that you will be gluing. So, side dressing of the rip saw blades is really important. You might try sending one batch of saws to a different shop and see if they do a better job than what you have now. A better sharpened saw will mean less heat.
Of course, the saw blade could be fine but the saw feed is not correct so the lumber moves while being ripped. Do you have a machinist flat edge to check the surfaces for flatness?
Also, the ripsaw does need a flat reference surface, so the lumber should be skipped planed prior to ripping. Also, the lumber drying procedures should be producing fairly flat lumber.
Have you checked your glue spread rate and pressure? We do not want too much glue or too little pressure, giving us a thick glue line.
Sometimes in the wintertime, the adhesive is cool and thick so it does not flow and squeeze out well, so you have a thick glue line; keep the glue warm. Or, you could have cold lumber that quickly cools the warm glue after it is spread and the same thing happens; give the lumber plenty of time to warm up to room temperature.
So, once you have a tight fitting joint with proper glue spread and pressure, one additional step is to use a urea adhesive or maybe a 50-50 urea and pva adhesive. With water-based finishes, we do have to have a higher quality panel, which means better gluing as I have outlined.
I looked at this phenomenon under the microscope and it was very clear that the glue actually bulged out of the joint. It was also clear that if the glue line were just a few mils thick (under 5 mils, I believe), this was not an issue; in other words, it seems to occur only with a thick glue line. My conclusion is that heat in sanding causes thermal expansion of the PVA. If the glue line is too narrow, then there is not enough glue to expand to the extent that it can be felt or seen. We find it most troublesome on flush rims glued to a panel. After finish, the bump may be apparent even from heat of finish ovens if the glue line is wide.
I have also reproduced this phenomenon when a dry, glued panel is exposed to high RH conditions, around 80 percent RH or higher. When the RH goes up, on wide joints the PVA actually seems to "grow" out of the joint. Again this only is an issue on wide joints, say above 5 mils and pretty common if the joint is 10 mils. This moisture effects seem a bit more puzzling because I am uncertain if high RH can actually make the PVA grow as much as I have observed. The growth or bulge is very significant. I theorize that the compression on the surface of the wood, coupled with the softening effects of high RH, squeezes the PVA out of the joint.
There are two solutions that have worked. The best solution is to make a tight-fitting joint; my column has certainly advocated that approach over the years. A second solution is to use urea or urea/PVA blend adhesive with perhaps 50 percent urea.

Have something to say? Share your thoughts with us in the comments below.