The Technique of FURNITURE MAKING

Shrinkage

A full understanding of the reasons for the shrinkage of wood, the probable extent of the movement and the direction it is likely to take (wood is not a homogeneous solid and the movement is not equal in all directions) is essen­tial if furniture is to be soundly constructed, for it is impossible to lock wood fibre permanently against its natural inclinations, and its strength is such that it will eventually overcome every effort to confine it. Furthermore, the increasing use of artificial heating means drier atmos­pheres and greatly increased shrinkage values, and this equally applies no matter how old the wood is, for even antique furniture will rapidly disintegrate in over-heated surroundings. It should be pointed out, however, that heat alone is not inimical to wood: it is the degree of dryness occasioned by the heat which is the deciding factor.

The probable extent and direction of the shrinkage is best understood by visualizing the tree as a compact cylinder composed of innumerable smaller cylinders or annual growth rings fitted tightly within each other (Figure 15).

As the tree dries out so the shrinkage will take place in every direction except in the length, the circumference of the outer cylinder will grow shorter and the cylinders within will each grow progressively smaller and more tightly packed (1,2). This circumferential shrinkage along the length of the annual rings is always greater than the shrinkage between the rings, and while a precise ratio is not possible (for the figures depend upon the density of the wood species and the oil and resin content of the fibres), it is generally assumed that the amount of shrinkage between the rings will only be a little over half that along the length of the rings.

The effect of these shrinkage values is illustrated in 3 and 4. In 3 the greater amount of shrinkage in a tangentially cut plank will be across the width, as shown by the heavy arrows, with about half that amount between the rings as indicated by the smaller arrows; whereas in radial or quarter cut wood the main shrinkage will be in the thickness, which is relatively un­important (the amount shown in the drawings is exaggerated for illustration purposes). Thus, the radial or quarter cut plank (4, 5 A) is always

5 -

fill 11 (I III 1111.1 A t ft

15 Shrinkage of wood

more stable, for not only is there less shrinkage across the width, but the pull is fairly equal from one annual ring to another, and therefore there is no tendency towards distortion. In 5B the annual rings in the tangentially cut plank are of different lengths, and the pull on the underside away from the heart will be much greater than on the side nearer the heart. A tangentially cut plank will, therefore, pull in, rounding as it dries, as shown by the arrows, while circular sections (6) will go oval, and square sections (7) diamond shaped. Where the grain direction is irregular as in 8, then the plank may twist in several directions across the width, while a further complication in tangential timber is that when it is cut from the log the saw cut is parallel to the long axis of the tree and, therefore, diagonal to the growth rings. It must be accepted, therefore, that wood is always somewhat unpredictable in its behaviour, and only very generalized rules are possible.