Q. I recall that you stated in your column that the adhesives we use for wood are stronger than the wood itself. True? If so, I need to know what factors are critical in making a strong joint. The reason I ask is that we are getting glue joint failures and not wood failures.

A. With the adhesives we use for making furniture, cabinets and so on, the adhesive joint has the potential to be 50 percent stronger than the wood itself. In other words, if a panel or other item breaks, it should never crack (even a small crack) or break along the glue joint. If it does, you are doing something wrong. So here are the main factors in developing a strong joint.

 1. You need a fresh surface for the wood and the adhesive to bond chemically. The strongest joint will be made when a surface is glued minutes after it is prepared. As time goes on, the surface attraction for glue (technically called surface activation) drops. Within a week, there would be a measurable drop, especially if the wood is in a dusty atmosphere. See also No. 3 below.

 2. A glue joint between two pieces of wood starts to weaken when the two pieces of wood are further than 0.006 inch apart. (Note that 0.006 inch is one-fifth of a 1/32 inch; really small indeed.) (Also, if closer than 0.002 inch, there is not enough glue to form a strong joint.) It is critical that the wood pieces being joined be perfectly flat. You should strive to make the tolerance in preparing these surfaces about 0.003 inch if you want the strongest glue joint possible. Of course, with 50 percent extra strength in a glue joint compared to the wood’s strength, we can be a small bit sloppy and still be OK.

 3. After a surface is prepared, there is a risk that moisture changes can occur in the wood strip while waiting to be glued. (Many times the wood is at 8 percent MC but the plant’s atmosphere is only at 5 or 6 percent EMC, especially during the wintertime. Such moisture differences mean moisture changes in the wood and these changes will cause the surface to become non-flat; gaps over 0.006 inch are likely. To prove to yourself that prompt gluing after the surfaces are prepared, walk a few pieces through the system without delay and see how good the quality is. In one plant, their rejects after gluing dropped from about 10 percent to 2 percent when they avoided a delay.

 4. Although it is a bit more critical in handling, I do believe that for many operations using a liquid system rather than powder will give you a more consistent glue mixture and will eliminate a potential problem with lumps, poor mixing, and so on. At the same time, handling the liquid system is a bit more critical.

 5. Both the glue and the wood must be at room temperature. This may mean that the heat, if turned down for the weekend in the wintertime, will need to be turned back up Sunday night so that the wood is warmed up by Monday morning. Be aware that in the hot summer months, it is necessary to use a slightly different glue mix to avoid the thin, more liquid, more “runny” glue because of the warm temperatures.

 6. You must have enough glue in the joint prior to applying pressure. We cannot develop a strong joint is there is too little glue. The glue application and spreading system must be adequate. Employees involved in gluing should be alert to situations where glue spread is inadequate; someone needs to immediately figure out why there is a problem and fix it.

 7. When the pressure is applied, it takes a few seconds for the excess glue to be squeezed out. Once the pressure is applied is is super critical to make sure that the actual pressure does not drop (such as might happen if the hydraulic pump performs poorly). A drop in pressure, even a small drop, will allow the joints to open slightly and because there is no longer excess glue, a weak joint will be formed. Likewise, if a problem arrises when the pressure is first applied and then the pressure is released to realign the wood pieces (or whatever) and then pressure is reapplied, the resulting joints will likely be substantially weaker.

 8. We will get the strongest joint when the wood has a clean, but slightly rough, surface. With the nooks and crannies formed by planing, sawing or sanding a surface with sharp tools or sandpaper, the glue will be able to also mechanically attach to the wood (in addition to the normal chemical attachment). At the same time, we do not want the surface to have loose fibers. The sanded surfaces should be inspected with a 10x to 15x magnification glass to check the surface quality.

 9. It is critical that surfaces being joined were not heated to more than 160 F when they were prepared. Overheating will inactivate the surface and this means poor chemical bonding. The source of heating could be if the strip stalls in the sawing, moulding or sanding operation. Dull tools or dull sandpaper will also generate excessive heat. Measure the temperature of the surface immediately (seconds) after the surface is prepared using a portable temperature sensor gun.

 10. It should be appreciated that the maximum strength of the joint will not develop right when the panel leaves the press, but will develop in the 12 hours afterwards. Avoid excessive stress on the joint until the joint is fully cured.

 11. If one of the pieces being glued is slightly warped, but is then pushed straight in the press, this piece will try to return to its warped shape after gluing. This stress that is created will be subtracted from the useful strength of a glue joint; in other words, the joint will not appear to be as strong. Avoid using warped strips.

 12. Normally, the wood fibers (also called the grain) will run lengthwise (parallel to the length) of a piece of wood. You can get a slope of grain (SOG) when the sawmill is sawing crooked logs, when the sawmill uses incorrect sawing procedures, and around a knot. A small SOG of 1:12 can substantially reduce the strength of the wood, but perhaps more importantly is that such SOG also is prone to weak glue joints as the liquid in the glue is soaked up by the grain before pressure can be applied.

 To help avoid a problem, work with the sawmillers, especially those in smaller mills, to avoid sawing pieces that will have steep SOG. Also, when defecting lumber, especially on a larger knot, do not defect too close to the knot. Allowing a bit more wood around the knot to be defected will lower yield, but the strips produced will be better in quality and will glue better.

 13. The adhesives we use on wood are designed for a specific moisture content, especially PUR adhesives. Wood that is too wet will not develop the maximum strength joint. Also, wood that is too dry will also not develop the strongest joint. Moisture control is a critical issue for strength; 6.5 to 7.5% MC is ideal in most cases. I might add that the only way you will now the moisture in a piece of wood is to measure it yourself; do not rely on someone else’s information.

 14. The pieces being glued must have the surfaces being glued at 90 degrees to the faces. If the surfaces are a bit off of true, then the pressure will cause a panel to bow up slightly. To correct this, oftentimes pressure is applied to the top of the panel; sometimes a hammer is used to pound the panel flat. Either technique is a big mistake. This top pressure actually will open a few joints a few 0.001 inch, and, because the excess glue has already been squeezed out, the open joints will be without enough glue to form a strong joint. Get square surfaces before gluing. One source of un-square surfaces is cupped lumber that is fed to a gang rip saw; work on eliminating cup. Oftentimes planing before ripping will help, but better drying is also important.

 15. Be picky; accept nothing but the best.