Dave MacArthur
03-24-2011, 3:15 AM
The topic comes up fairly often, thought I'd write a blurb up with a pic so I can point to it in future when folks ask about wood movement.
I've been meaning to write this anyways so this may be a useful time.
Wood Movement
1. Boards are usually cut to rough length first, reducing all the bow/curve a long board might have that will cause machining/safety issues as well as wood loss when trying to straighten. Length meaning distance along the grain. Wood doesn't move almost at all along it's grain length. Trees don't get taller and shorter with the weather/moisture.
2. rip boards to rough width next, width being distance across the grain, usually on a flat-cut board. This width can change with moisture, so it's best to oversize. Boards can also curve after cut, so that you have to joint the edges (one concave and the opposite edge convex) straight, thus losing width.
3. re-saw boards to rough thickness. Thickness is usually talking about "quarter-sawn" or radial grain direction, that is from the center of the tree towards the bark; but this is most common only because most wood is "flat sawn" or "plain sawn", making it's thickness radial. However, in a quarter-sawn board the grain directions of width and thickness are reversed.
Who cares what grain direction thickness and width are? Well, YOU do, because this is what causes boards to move. First there can be internal stresses in the wood. However, the center of a board can also have different moisture content than the outside. Consider a slice across the trunk of a tree that produces a board. Most hardwood trees we use for boards now are not that big in diameter, and therefore the width of our boards is not that big. Consider a 24" diameter trunk, a decent walnut. Now imagine a 12" slice, see it going across the circular grain lines... can you imagine that one side of the board has the growth rings going quarter-sawn (outer edge of board), and the other side of board where it passes above the center/pith will have flat-sawn tangential ring grain? Because of this, the two sides of the board will move differently as the board dries.
As wood loses moisture, it must shrink both radially (center to bark) and tangentially (around the circumference of tree and any growth ring). The wood shrinks much less radially than tangentially. I had some rough math-based calculations here, but as some disagreed with the exact numbers, I'll reduce it to something not arguable that still explains the point: wood shrinks much less in a radial-grain direction (quarter sawn) than tangentially around the growth rings (flat sawn). While this applies of course to the change across the whole trunk, any "slice" of that trunk must and will display the exact same differential changes. This is the basis for 'quarter sawn stability'--the "width" of a quarter-sawn board is all radial grain, while the "thickness" is tangential grain, so the QS board will change width only much less than a flat-sawn similar board. It will change thickness much more than a flat sawn board, but most boards are only 3/4" thick in use or less, so it's not as apparent or usually critical to our joints.
Now, when you slice across a trunk to make a board, you can see that one side of the board will be more "quarter sawn" than the other, this results in different movement rates. In a THICK board, the different movement rates due to radial/tangential grain differences can be significant just across the thickness of the board. Draw a picture of the end of a tree with rings on it and start slicing out thick boards on paper and look at the ring/grain orientations on different edges/faces of the board, it becomes obvious.
So, you have a board. It dries. It moves, and soon all it's forces shrinking one part, expanding another, twisting here and there due to different shrink-rates, all balance out and it stops moving. But does it stop because it's dry and won't move again? NO. It stops because a layer of wood on each side of the board is exerting it's own force, in fact every small discrete chunk of wood in the board is exerting it's own force vector, and they all add up and move until ... balance. Someone squares that board up roughly and sells it to you.
Now, you take a saw and slice off half of the total balancing forces, and maybe at the same time even expose some wood of different moisture content that begins drying and exerting new forces! So, board twists and warps until all the discrete forces are equalized again. Small milling from all sides of a board will tend to result in small removal of forces that are close to equally balanced, resulting in little or no board movement. Slicing a thick board in half can not result in differential changes, as the thickness of these boards is a significant distance in comparison to the critical grain-line changes.
I've been meaning to write this anyways so this may be a useful time.
Wood Movement
1. Boards are usually cut to rough length first, reducing all the bow/curve a long board might have that will cause machining/safety issues as well as wood loss when trying to straighten. Length meaning distance along the grain. Wood doesn't move almost at all along it's grain length. Trees don't get taller and shorter with the weather/moisture.
2. rip boards to rough width next, width being distance across the grain, usually on a flat-cut board. This width can change with moisture, so it's best to oversize. Boards can also curve after cut, so that you have to joint the edges (one concave and the opposite edge convex) straight, thus losing width.
3. re-saw boards to rough thickness. Thickness is usually talking about "quarter-sawn" or radial grain direction, that is from the center of the tree towards the bark; but this is most common only because most wood is "flat sawn" or "plain sawn", making it's thickness radial. However, in a quarter-sawn board the grain directions of width and thickness are reversed.
Who cares what grain direction thickness and width are? Well, YOU do, because this is what causes boards to move. First there can be internal stresses in the wood. However, the center of a board can also have different moisture content than the outside. Consider a slice across the trunk of a tree that produces a board. Most hardwood trees we use for boards now are not that big in diameter, and therefore the width of our boards is not that big. Consider a 24" diameter trunk, a decent walnut. Now imagine a 12" slice, see it going across the circular grain lines... can you imagine that one side of the board has the growth rings going quarter-sawn (outer edge of board), and the other side of board where it passes above the center/pith will have flat-sawn tangential ring grain? Because of this, the two sides of the board will move differently as the board dries.
As wood loses moisture, it must shrink both radially (center to bark) and tangentially (around the circumference of tree and any growth ring). The wood shrinks much less radially than tangentially. I had some rough math-based calculations here, but as some disagreed with the exact numbers, I'll reduce it to something not arguable that still explains the point: wood shrinks much less in a radial-grain direction (quarter sawn) than tangentially around the growth rings (flat sawn). While this applies of course to the change across the whole trunk, any "slice" of that trunk must and will display the exact same differential changes. This is the basis for 'quarter sawn stability'--the "width" of a quarter-sawn board is all radial grain, while the "thickness" is tangential grain, so the QS board will change width only much less than a flat-sawn similar board. It will change thickness much more than a flat sawn board, but most boards are only 3/4" thick in use or less, so it's not as apparent or usually critical to our joints.
Now, when you slice across a trunk to make a board, you can see that one side of the board will be more "quarter sawn" than the other, this results in different movement rates. In a THICK board, the different movement rates due to radial/tangential grain differences can be significant just across the thickness of the board. Draw a picture of the end of a tree with rings on it and start slicing out thick boards on paper and look at the ring/grain orientations on different edges/faces of the board, it becomes obvious.
So, you have a board. It dries. It moves, and soon all it's forces shrinking one part, expanding another, twisting here and there due to different shrink-rates, all balance out and it stops moving. But does it stop because it's dry and won't move again? NO. It stops because a layer of wood on each side of the board is exerting it's own force, in fact every small discrete chunk of wood in the board is exerting it's own force vector, and they all add up and move until ... balance. Someone squares that board up roughly and sells it to you.
Now, you take a saw and slice off half of the total balancing forces, and maybe at the same time even expose some wood of different moisture content that begins drying and exerting new forces! So, board twists and warps until all the discrete forces are equalized again. Small milling from all sides of a board will tend to result in small removal of forces that are close to equally balanced, resulting in little or no board movement. Slicing a thick board in half can not result in differential changes, as the thickness of these boards is a significant distance in comparison to the critical grain-line changes.