why end grain joints are weaker.......
Bill Tindall
Some years ago I happened to meet the chief chemist from Frainklin glue company. This was an opportunity not to be wasted. While weeding(the mind does wander whilst weeding) I pondered the question of why end grain joints were weaker. Pasted below is the answer I received to this question.
As you are aware, the geometry of wood is similar to a bundle of straws. Those straws or tubes are designed to take up the sap and the nutrients that are needed for growth and transport them out to every part of the tree. While the edges and faces of a plain sawn piece of wood are comprised of the sides of those tubes and form a nearly solid layer, the ends of the same piece are comprised of only the open ends of those "straws" or tubes. That results in two different facts that create problems in achieving successful end joints. First of all, the fact that the end grain is so open means that there is comparatively little actual wood surface available for bonding. Perhaps even more significant, however, is the fact that the open tubes that form the end of the piece are highly absorbent. That means that when what would normally be an appropriate layer of any water based glue is applied and the joint closed, the porous end grain literally steals
all the glue from the joint. Taken together, those factors drastically reduce the chances for forming a meaningful bond.
"Thus, whenever possible, an assembly should be designed in such a way as to avoid bonding end grain. Yet, when bonding end grain cannot be avoided, recognizing the factors involved can usually allow at least an adequate bond to be formed. My suggestion is that one of two approaches be employed. First, consider diluting a small amount of the glue that is to be used in a separate container - perhaps a jar lid. Then take that thinned glue (diluted with at least one part of water to one part of glue) and apply a coat over the end grain in question. That thin mixture will be rapidly drawn into the wood and have little effect on the bonding surface, but it will partially plug those pores, and prevent the undiluted glue used in the actual bonding operation from disappearing into the wood. The fact that that "sizing" does not fully coat or seal the end grain allows the real glue to anchor well to the wood and leaves enough porosity so that the glue
can dry. When using this approach, the sizing should be given at least a few minutes to dry, but this process could be also be done hours or days before the actual assembly occurs as long as care is taken to assure that the surfaces to be mated are still well matched. .
The alternate approach merely involves a two step application of the glue being used for bonding, with the initial coat applied and allowed a minute or two to soak in, followed by a reapplication of fresh glue. While the initial coat serves to quench the thirst of the open end grain, the subsequent application assures that there is enough fresh glue left in the joint to achieve some reasonable bonding."



