There seems to be a lot of misinformation that exists (word of mouth and Internet sources) about the role of pressure when gluing wood. So, let’s look at this topic in detail. (Epoxy adhesives have different requirements, so we are talking about PVAs, PURs, and ureas here.)
First, a wood surface to be glued needs to be quite smooth. The smoother, the better. Why? Because most woodworking adhesives require the two surfaces to be between 0.002 and 0.006 inch apart. (In case you prefer to talk about 1/32 inch and other fractions, 1/64 inch is 0.031 inch, so the maximum gap is closer to 1/128 inch; pretty darn small indeed.) This requirement is if you want the strongest possible glue joint; many times, we do not need a joint with strength of over 10,000 pounds per square inch. In fact, in practice, oftentimes we can be happy with a joint that has 1/3 of its area that approaches this high value and the other 2/3 is much weaker. Stated another way, the strength of the perfect joint is 50 percent stronger than the wood itself; if the joint fails, then, the wood will be the weakest link and so the wood itself will fail and not the adhesive. But, if for some reason we do need the strongest glue joint possible, then we need really flat, true, smooth surfaces.


Find more Wood Doctor at

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.

Second, one of the most common causes of an un-smooth surface is moisture. The smooth surface is excellent when first made, but then over the next 15 to 20 minutes, this surface picks up or loses a few water molecules to the moist or dry atmosphere, causing some localized swelling. For this reason, the strongest joint for wood will occur if the glue is applied within minutes after the surface is properly prepared… and of course, the correct pressure follows promptly. The denser the wood, the shorter the time allowed, as dense woods swell or shrink more than lower density species. (For a thin piece of wood like veneer, the time between surface preparation and gluing can be under a minute.)
Next, having said this, we do have to look more closely at what happens when we make a surface smooth. If we generate too much heat, the wood’s chemical bonding sites for the adhesive can be destroyed. Also, if we use dull cutters (knives, saws or even sandpaper) to prepare the smooth surface, we can squish and tear the wood’s surface (microscopic level) so the surface cells as damaged or no longer well attached to the main piece of wood. In these cases, the adhesive cannot achieve good contact with the solid wood, so we have a weak joint.
Now, when we put pressure on the joint with a hard dense species like oak or maple, even if we use very high pressure, the wood is so strong that we cannot squeeze the surfaces flat. So, with a dense, hard species that is not perfectly flat, when we apply the pressure, we will likely have some spots that are essentially touching and there is no adhesive at these spots (we squeezed it all out) and we will also have some spots that exceed 0.006 inch in distance between the two surfaces.
The wood is too strong to give a little and compress uniformly…that is, close the gaps. However, with softer species, like soft maple, most pines and spruces, cottonwood, and so on, the pressure on the joint will cause the spots that are touching to squish a little bit and the larger gaps will close to a more acceptable level. This is why the softer species are so much easier to glue effectively.
Third, when we spread the adhesive on one or both of the mating surfaces, the adhesive is not uniform in thickness. In fact, sometimes we spread the adhesive in long ribbons, which gives us a longer open time (the time between spreading and applying the pressure). So, prior to applying the pressure, we likely have wavy wood surfaces (on a microscopic level) and we have varying thicknesses of adhesives. So, as the pressure increases initially, the excess adhesive will be pushed around and fill in the larger gaps between the two pieces of wood. (If we only spread the adhesive on one surface, this initial pressure will cause the adhesive to transfer to the mating surface.
Here’s another critical point about pressure. Except with end grain, we do not push adhesive deep into the wood cells. (Wood cells are like miniature soda straws and usually run lengthwise in the pieces of wood being glued.) The adhesive and the joint strength are what happens with the exposed cells right at the surface. Now, if the piece of wood has steep slope of grain, then we essentially have partial end grain on the surfaces to be glued. This means that the adhesive can go into the wood past the surface cells. However, such deeper adhesive does not actually add to the strength of the joint; the strength is dependent on what goes on right at the surface. (Exception: if the joint is tested in shear rather than tension, some penetration might, in a few cases, increase the shear strength slightly. Usually, deeper penetration increases the amount of adhesive used to make the joint, but not the overall strength.)
Using the analogy of cells being like miniature soda straws, it is important that the cells on the surface are not squished but do have partially openings. An analogy for snow skiers is that the straws look like “half-pipes.” This openness allows good surface contact between the wood and adhesive. If the walls are folded in due to excessive pressure when machining or dull tools, then surface contact is limited and the joint becomes weaker. Often times, in my plant visits, I suggest that saws be sharpened twice as often; that is, we are concerned about microscopic events, which is when dull or partly dull saws perform poorly, but such poor performance is hard to notice until gluing.
Note that in all these cases discussed so far, the real problem is with the wood surface and not the adhesive.
It is probably no surprise that when we apply pressure, we squeeze out the excess adhesive (assuming the adhesive is not too cold, thick, or partially cured already). It is common to see operations use plenty of adhesive, ”just in case,” and then count on the pressure squeezing out 100 percent of the excess. This means that squeeze-out is a high quality event. But with larger, wider joints, it is critical that the adhesive be thin enough to move out—a long (maybe impossible?) path for large joints. So, excessive adhesive is good, but let’s not get carried away.
As just mentioned, we are concerned about the viscosity of the adhesive. The biggest factor affecting viscosity is temperature. Many operations maintain the adhesive at a constant temperature year round, 24/7…a great idea. But, remember to also keep the wood temperature constant as well…cold wood can quickly cool and thicken the adhesive.
Now, if for some reason after the initial pressure is applied, the pressure is then reduced even a little bit, there will be some places where the joint opens up (microscopic level) and the joint is short on adhesive (a microscopic gap), so the strength drops substantially. So, it is critical to never let the pressure drop even a slight bit, especially initially. However, as the joint cures more and more, then the pressure can be dropped without such a big negative effect. (Of course, increasing the pressure after the adhesive starts to cure is also a dumb idea.)
Two points: If using manually tightened clamps, it is better to tighten them maybe 80 percent initially and then go through moments later and final tighten them all. If you go 100 percent right at the start, the wood can shift as others are tightened and the joint might open in some spots. Second, with hydraulic or pneumatic pressure, I have seen the pressure come on, close the clamps, and then drop slightly (various reasons why this happens) within a few seconds or longer. The initial appropriate high pressure squeezed out the excess adhesive, so when the joint re-opened, there was not enough adhesive in the joint area any more to make a strong joint.
Soon question will be “Can we have too much pressure?” The answer is “Yes, we can have a starved joint,” from technical point of viewpoint. But due to the nooks and crannies, there will always be some parts of the joint with the optimum spacing. Hopefully, these spots provide enough strength to develop adequate strength. We can starve a joint for other reasons: by not using enough adhesive, by letting is pre-cure, or by having too thin an adhesive. If we have an RF curing process, the lack of adhesive also means different heating and curing for the rest of the joint. Pressure is just one factor in making a good joint; the interaction of pressure and other factors must be considered.
Finally, if we use pressure to also bend a crooked piece flat or straight, we have two issues...a lot of the area will be low on adhesive where the contact area is really high in pressure. Also, when we release the pressure, the bent piece that we are holding straight will try to bend back to its original shape. This puts a lot of stress onto a joint, so the amount of useable strength left (after we subtract the bending stress) before we get a failure is very low...and so the joint breaks at low external stress levels.


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