Shrinkage/Expansion tables-nearly worthless
Bill Tindall, E.Tn.
It is well to know where the data in these tables come from. The ones I know of are a result of averaging the shrinkage of numerous test specimens in drying from fully saturated to zero moisture content. The value reported is the average of this shrinkage between saturated and zero moisture content, averaged again for the number of test specimens. . Seldom reported is that the naturally occurring variations in the tree result in a 30% (from memory) variation in this value between trees of the same species. The tabulated values are useful for comparing stability of various species, but useless for predicting what a board in hand may do because....
1. Shrinkage is not linear between 30% and 0% moisture content.
2. Your board may differ substantially from the average board.
3. KD lumber becomes more stable as a result of the typical KD process.
Actual data:
As a result, I find actual stability is much better than these calculations predict. I have 3 test pieces that are cut-offs from panels. They contain the usual variation of grain orientation of flat to quartered. The humidity where they live varies from 40% in the winter to 65% in the summer. Reported are the extremes of sizes I have measured over the past 11 years.
Walnut 25 5/32" to 24 8/32" range 3/32"
Walnut 15 7/32 to 15 9/32" range 2/32"
Elm (reported to be very unstable) 11 42/64 to 11 45/64" range 3/64"
Had I made a 5" tall drawer from the elm it would have changed in height less than 1/32" from season to season. The drawer would not have changed at all in length.
Moral of story- fit length close to prevent racking when operating. Use judgement in fitting height taking into account where the drawer will live and under what conditions it was fit.