After seeing a few blogs about wood shrinkage, I thought I would throw in my 2 cents. I was considering building a 44”-wide solid maple countertop in my non-air-conditioned shop. It had to fit snuggly over an existing island countertop made of granite. I was concerned with what would happen when the wood acclimated to its new air-conditioned environment.
I needed a ballpark answer to what I thought was a relatively simple question. How much shrinkage did I need to account for? I found there is a lot of information available; ranging from peer-reviewed papers to opinionated forums. But none of them answered my questions and most of them just raised more questions. Frustrated, I devised a small experiment of my own to find the answer I needed and to see first-hand what I was reading about.
See data here: http://www.acuteimagination.com/WoodShrink.htm
I assembled a number of samples of wood acclimated to my non-air-conditioned shop including plain and quarter-sawn walnut, mahogany, and maple (OK so they are not exactly described correctly, but you can see from the grain pattern that some of them have grain almost parallel to the top surface while in others it is at least 45 degrees if not perpendicular to the surface.) I planed the pieces to an identical thickness of 0.185” and cut the pieces to a width of 2.297-2.302”. I then further divided each of the samples into 3 pieces and labeled them “C”, “W”, and “D”.
Pieces labeled “C” were taken into an air-conditioned environment and after acclimating, measured for changes in the 2.297-2.302” width. This would be the basis for the ballpark answer I was looking for.
Pieces labeled “W” were submerged in distilled water for 3 hours, and then placed in a zip-loc plastic bag (without drying excess water) to prevent the loss of and to allow for equalization of any water already absorbed. Any changes in the 2.297-2.302” width from this treatment (approaching 100% moisture content) would be purely a curiosity.
Pieces labeled “D” were placed in an oven for 3 hours at 250 degrees Fahrenheit. After removal, they were immediately placed in a zip-lock plastic bag to prevent them from re-absorbing ambient moisture as they cooled. By allowing the samples to cool before measuring, any changes in the 2.297-2.302” width from heat expansion would be eliminated. Any changes in the 2.297-2.302” width from this treatment (approaching 0% moisture content) would be purely a curiosity.
Considering that this experiment has a very small sample size and only a single repetition, one should not read too much into the data; however:
1) I would be comfortable thinking that I could expect somewhere between 0.35 and 0.61% shrinkage. Over a 44” surface, that translates into 0.15 to 0.27 inches of shrinkage. I would probably build a little extra fudge into it.
2) It appears that 2/3 of the shrinkage that occurs after 3 weeks of adjusting to an air-conditioned environment can be measured after just 3 days.
3) It appears that “quarter-sawn” wood shrinks less than “plain-sawn” wood of the same species.
4) It appears that once a wood is soaked, it does not shrink back to its original size.
5) It appears that once a wood is dried to 0% moisture, it does not expand back to its original size.
If you are concerned about wood shrinkage in a project you are working on, you might be able to predict the shrinkage by preparing a sample as described above and moving into its new environment for as little as 3 days. Multiplying the result by 1.5 might give a good estimate of the shrinkage in 3 weeks. Assuming the thin sample can acclimate in just three weeks, one might infer that thicker, full-size samples of the same wood would over longer periods shrink accordingly.
This was done in the Tampa bay area of Florida during July of 2007. Of course results will be different for your location and time of year. Give it a try and let us know how it works out for you.