Wednesday, Woody Wednesday – Understanding Shrinkage

Well, that title sounded a bit more salacious that I intended. No really, I’m writing about wood properties here, not the unfortunate results of skinny-dipping in a cold pool.

In this, our second edition of Woody Wednesday, we will discuss the one wood property that causes more ruined projects and gnashing of teeth than practically any other – its propensity to expand and contract with  changes in moisture content.

Wood is hygroscopic and anistropic. Hygroscopic means that it has an affinity for water – very helpful for a living tree, which must conduct thousands of gallons of water tens to hundreds of feet above ground through the pores of its wood to reach its leaves. Let’s pause for a moment to consider what a marvel this is. This video will help:

Those water-conducting pores are what makes wood anistropic – meaning that its properties are different depending upon direction. You can think of wood like a bundle of straws. The properties at the top of the straw bundle are very different from the properties along its sides. The opposite of anistropic is isotropic – materials that have the same properties in all directions. Examples of isotropic materials are metal and glass. On a microscopic level, you could correlate them to a jar of sand, rather than a bundle of straws.

Wood vs. Steel

It is this combination of hygroscopic and anistropic properties that causes wood to shrink and swell with seasonal variations in moisture content. When we saw or split wood from a log, it contains water. Lots of water. Some of this water is what we call “free” water. This is the water that was moving freely up through the pores, from the roots to the leaves. The wood also contains “bound” water – water that is chemically bonded with the cellulose and lignin that make up the cell walls.

When the wood begins to dry, the free water tends to exit the wood rather quickly. Since it isn’t chemically bound, it is free to evaporate. The point at which no more free water remains in the wood is called the fiber saturation point. Only bound water remains. This occurs at 25-30% moisture content.

What exactly is moisture content, by the way? You’ve probably heard it mentioned, have you ever wondered what it means? It’s pretty simple: moisture content, or M.C., is simply the weight of the water divided by the weight of the woody, fibrous material, expressed as a percentage. So a 100 g sample of wood at 26% M.C. contains 20.63 g of water and 79.37 g of actual wood (20.63/79.37 * 100 = 26%).

As the free water dries, the wood changes very little in size and shape. After the bound water begins to dry, however, funny things start to happen. When water that was previously bound to the cell walls evaporates, the cell walls contract. This wouldn’t be problematic if the wood shrank evenly, but wood is anistropic, so the shrinkage occurs differently depending on what plane we’re looking at.

When scientists talk about wood, we discuss three different planes: transverse, tangential, and radial. As woodworkers, we tend to refer to these planes as end grain, quartersawn grain, and flatsawn grain. There’s also a term for a plane that is in between quartersawn and flatsawn, which we call riftsawn. I’m not an artist, but here’s a drawing to wrap your mind around it:

Ellery 043
Three wood planes: transverse, radial, and tangential.

Wood reacts very differently along these three planes. It shrinks very little or not at all along the grain. It shrinks 2-6% in the radial plane (from pith to bark), and about twice as much (5-10%) in the tangential plane (along the growth rings) from the fiber saturation point to air-dry. This calls for more custom artwork:

wood 002

This differential shrinkage has a profound impact on the stability of the wood as it dries. Quartersawn wood tends to be the most stable, but it will shrink in thickness more so than flatsawn wood. Flatsawn lumber has a definite propensity to “cup” as it dries due to the differential shrinkage between the tangential surface in the middle of the board and the riftsawn grain on the edges. Something that is turned round from green wood will end up egg-shaped as it dries. Again, a picture should help to visualize these effects:


Wood is considered “air-dry” when the moisture content of the wood reaches equilibrium with the relative humidity of the air. Scientist call this the equilibrium moisture content (EMC). It can be as low as 5-6% if you live in the desert, or as high as 15-16% in a rainforest. A more common range for air-dried wood in temperate climates is 10-12%.

Now, wouldn’t it just be dandy if we could dry our wood down to the equilibrium moisture content, and then have a stable, predictable material that we could glue and screw to our hearts’ desire, without any consideration for dimensional changes over time? That would make everything so easy!

There’s just one problem. Wood will always, always, reach equilibrium with the air surrounding it. So unless you live in a temperature-and-humidity-controlled laboratory, you can expect weather patterns, seasonal changes, and modern heating and air-conditioning to cause the moisture content of your wood to vary by 3-4% in normal use. This might not sound like much, but if you don’t incorporate allowances for wood movement in the design of your furniture, the results can be catastrophic wood failure.

Putting It Into Practice

Consider the sassafras table that I built a few weeks ago. The top of the table is 48″ wide. Sassafras is a pretty typical domestic hardwood in regards to dimensional stability. It changes by about .003″ (three thousandths) per inch of width for each percent of change in moisture content. Three thousandths sounds like we’re picking nits, but the numbers multiply rapidly. If we assume that the MC starts out as 12% and drops to a minimum of 8% in the dead of winter with the heater running at full blast, that’s a 4% change in MC. Multiply .003 x 4 (% MC change) x 48′ (width of the panel in inches) and we get 0.576″ in shrinkage along the width of the table. That’s over a half-inch!

So, since the wood does not shrink along it’s length, if I had screwed down the battens without making allowances for wood movement, it’s obvious that the result would be, at best, a severely bent tabletop, or more likely an ugly split right down the middle. Wood will swell as it absorbs moisture and shrink as it releases it. It is as certain as the sun rising in the east and time slowing down as you approach the speed of light.

Since I am fully aware that my tabletop will shrink, I was able to design the battens with this in mind. I drilled elongated holes into the battens and screwed the top on using washers, which will allow the top to shrink and swell with the seasons, all the while sliding as it wishes along the battens without bending or splitting.

This is but one of many examples of designing furniture to allow for seasonal wood movement. The most famous example is the ubiquitous frame-and-panel door. The narrow members of the door experience very little movement, but they house a wide panel in a groove that is free to shrink and swell as need be without affecting the dimensions of the door. If doors were designed from solid wood, they wood stick shut as they swelled in the summer or leave ugly gaps as they shrunk in the winter.


Okay, so now you have a basic understanding of why it’s necessary to design your furniture with wood movement in mind. It’s beyond the scope of this (already long) article to discuss any more specific techniques for addressing wood movement, but a basic understanding of wood movement and a little common sense can go a long way to avoiding furniture failures.

I will tell you that I typically take a very unscientific approach to addressing wood movement myself. I know that my wood will shrink when I move my furniture from my un-heated, un-cooled workshop to inside my house, so I plan for it. I typically figure that the wood that comes out of my shop will be around 11-12% MC, but I never measure it. I have probably used wood as high as 14% MC, but since my furniture is designed with wood movement in mind, this is never problematic in practice.

The woodwork inside my house is generally around 9-10% MC. Maybe a percent drier in the midst of winter, or a bit wetter when the windows are open during the spring. Figuring on a maximum change of 5-6% is generally pretty safe. But again, I don’t bring any numbers into my figuring. Wide panels get more allowance than narrow members. Experience is my guide. If you are uncomfortable with this seat-of-the-pants design, then Popular Woodworking has a great online resource for calculating wood movement. Give it a shot.

Also understand that the numbers for EMC in my neck of the woods may be vastly different in your part of the world. Check out this page for typical EMC in different cities in the United States.

Finally, if discussions of equilibrium moisture content, hygroscopicity, and transverse planes has really whet your appetite, the Wikipedia page on Wood Drying is actually quite excellent.



Extra Credit

It’s one thing to design your furniture such that we avoid problems with wood movement. But what if we could design furniture that actually takes advantage of wood movement? Wouldn’t  that be something?

In fact, woodworkers have been doing exactly that – for centuries. But that’s a topic for another day…

5 thoughts on “Wednesday, Woody Wednesday – Understanding Shrinkage

  1. I’m a bit disappointed you mentioned shrinkage without this clip:

    More seriously, the video you posted, while very well done and interesting, doesn’t quite line up with what I learned in school. The question of how trees pump all the water they use vertically 10-150′ is that scientists can’t quite account for that amount of energy. While there is a great amount of proof that the water column within trees is under a great amount of tension, the tension alone can’t lift all the water the tree needs.

    Am I being nitpicky? Yes and my apologies.


  2. Thanks for the comment, Clark. Of course, that Seinfeld episode is the reason no one can so much as utter the word “shrinkage” without eliciting smirks and snickers. Thanks for correcting my oversight 🙂

    As to your comment on how trees pump water through their xylem hundreds of feet above ground level – well, it’s been 10 years since I took Advanced Tree Physiology, but I don’t recall Dr. Teskey having any consternation about the amount of energy required. The video falls pretty much in line with what I was taught.

    I did find this article from Scientific American that answers the same question. Their answers are basically the same.

    No worries about being nitpicky, it’s one of my strongest traits. Do you have any references that discuss the problems in accounting for the energy required for trees to lift water? I’d certainly be interested to read them if you do.


  3. It’s been about 10 years since I took tree physiology and I only recall the professor claiming that there wasn’t a complete accounting for the energy needed to move all the water to the crown of the tree. Most of the energy could be accounted for but it wasn’t enough to actually fuel the entire trip to the crown. At least, that is how I recall him saying it.

    I’d love to post some references but even a quick search told me that it has been too long since that class and I don’t know the words and phrases I need to search this out.

    The other way of posing this question is if the energy of the sun and the evaporation that occurs at the boundary layer of the leaf are ultimately what cause the negative pressure or tension within the water column and that is what moves the water upwards, how does the process start in the spring when there are no leaves?


    1. Your last paragraph states the problem very succinctly and sent me searching for additional resources because I couldn’t answer it. All of the world’s tallest species are evergreens (redwood, eucalyptus, sequoia, Douglas-fir, Sitka spruce, noble fir, auracaria, etc.), which don’t have to contend with this issue. No doubt, that’s not a coincidence.

      How the deciduous species resolve this conundrum is more complex than I expected – evolution has provided apparently provided many different solutions so there is not one right answer. One tree that has been studied extensively for its dormant season xylem flow is maple (for obvious reasons, i.e. that’s where we get maple syrup!).

      Here’s a passage from an interesting link on the topict .

      “Maple sap flow during the leafless season is physiologically unique in that it is largely independent of root pressure and only occurs on occasions between October and April when warm days follow freezing nights. Maple winter sap flow is caused by pressure in the stem generated by alternating daily cycles of night freezes and warm days. Cool evening temperatures generate negative pressure from the dissolution of gases in the xylem, which were seeded in from adjacent parenchyma and intercellular spaces. The negative pressure replicates the effect of transpiration, which draws still-liquid water from the soil into the roots. As the night freeze deepens, water freezes along the inner walls of the hollow fiber cells adjacent to the xylem and in intercellular spaces. Eventually vaporized water on the surfaces of all cells freezes. The ice formation compresses and traps gases in the stem. The heat of the day melts the ice and causes expansion of the compressed gases, which generates positive pressure in the stem that pushes the sap up the stem and out the nearest exit, if one exists, such as a maple producer’s spile.”

      Anyway, good discussion – thanks for you contribution to the topic!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s