Directing Wood Movement with Fasteners and Glue

[TomInNC]

Over the course of the past year, I've read a ton about solid wood construction and wood movement. Something that frequently comes up is the concept of "directing" wood movement. For example, gluing the center points on frame and panel to direct the direct the expansion and contraction evenly within the frame. I have yet to come across, however, a definitive resource on exactly how different types of restraints affect the direction of wood movement. For example, if you screw down the end of a shiplap panel, does that direct the full amount of expected expansion to the unscrewed portion of the board? Or does this simply limit the travel on the end with the screws? 

Because of something I am working on now, I am primarily interested in how restraining 2 parts of the board would affect movement. Specifically, one side of a board for shiplap is screwed to ledger strips (far left side), a drawer slide is mounted in the middle with screws (another restraint, I think), and the right side is unrestrained. Does adding the additional restraint in the middle of the board force all of the movement to the right of the location where the hardware is mounted, or will I have issues with cracking in the part of the board between the far left ledger strip and where the screws secure the hardware in the center?

Anyone out there aware of resources on exactly how wood movement works when you start constraining portions of the board?

 

[gee-dub]

I have only fixed one end or the middle of a given piece of material.  I assume a fixed end, a fixed middle and a loose end will have a smaller version of the same problem as fixing both ends.  That is the material will move as it will.  When it hits a restriction that is too great it will react.  If the movement amount is great enough I would expect it to bow, split, or force an adjacent joint apart.  If the movement is mild due to width, environment, or species the movement may be innocuous.

[Tpt life]

There are calculators online for how much each species is expected to move along and across the grain. You just need to allow for that distance in your fastening. Check how people elongate holes in breadboard ends as the fasteners go out from center. A little planning can save a ton of effort. 

[Chestnut]

Wood movement can get complicated. I'm not entirely following your written description. If it's possible to draw out what your explaining with a specific portion of the project in mind it may help get better guidance.

For a general answer, it depends and is complicated. Lots of factors come into play like the thickness and width of material, the type of material etc, the environment the piece will live in etc. For directing the wood movement and different fastening methods. Assume they are all the same. A fastener or glue is a pin that holds the material stationary at that point. If a wide panel is fastened at 2 different points separated by a great distance problems will occur. If a panel is framed a problem will occur.

Some items like drawer guides fastened with screws will have some inherent wiggle room that will allow for their use on solid wood with out issue, but It's best avoided. Attach the drawer guides to internal frames that span the panel and attach on the front and rear legs. In frame and panel construction the panel is there for astehtics and not structure. The frame is the structure and where attachments should be made. If attaching a panel or wide board it should have 1 attachment point, front center or rear depends son the piece of furniture and the design.

[Bmac]

Always a great topic to discuss as I think movement of wood can be complicated and tricky. When confronted with fastening wood cross grain with each other I rely on a few simple rules. 

First, the width of the board plays a big role, a narrow cross grain attachment is not a big deal, a wide board is a big deal. As Chestnut stated above, fastening a wide board a at two points some distance from each other will not end well. 

Second, if it's a wide board than make just one point a fixed point and allow for movement of the rest of the board. That fixed point can be glued, a tenon (domino also), or a metal fastener. There are probably other fixed point fasteners I forgot. The fixed point can be located on one side or in the middle, and the choice often is based on which part of the board you don't mind the movement in. So with a panel in a frame you can fix the middle and allow for movement of the floating panel within the frame width wise on either side of the fixed point. With a bread board edge if you want the front of the piece to stay flush than put your fixed point in the front part of the bread board edge. So you can decide on which area is best for your project to stay fixed. 

Thirdly, with a wide cross grain situation, you often need additional fastening. Any other point should allow for movement and that can be done many different ways, but does not include any glue joints. It can be as intricate as a sliding dovetail or it can be as simple as a screw placed in a elongated hole, elongated in the direction of the wood movement. 

In your case that you were describing I would look for one fixed point and one point that allows for movement, you decide where you want it fixed and where you are okay with some movement.

All solid wood moves, and the take home message is you can control wood movement by allowing it to move.

 

[JohnG]

Drawer slides often have some slotted holes. If you are attaching a drawer slide to a solid side panel, use the round hole at the front and then horizontal slotted holes for the rest of your attachment points. This will allow the panel to move without the metal slide restricting it. As mentioned previously, the holes also have some inherent wiggle room as well. 

[Coop]

John, good point and one I had never thought of. I always figured the horizontal holes were for people like me to adjust the slides if and when I drill the holes incorrectly.

[JohnG]

On 4/14/2022 at 11:47 AM, Coop said:

John, good point and one I had never thought of. I always figured the horizontal holes were for people like me to adjust the slides if and when I drill the holes incorrectly.

Well, that’s what I’ve always used them for, too. But that doesn’t mean that someone else can’t use them to allow for movement!

[TomInNC]

Thanks for the replies. My question here came up as I was reading Chapter 4 and Chapter 5 of the Hoadley's Understanding Wood. In Chapter 4, Hoadley writes: "wood shrinks or swells due to loss or gain of bound water from the cell walls.... The orientation of the long-chain cellulosic structure in the cell wall is nearly parallel to the long axis of the cells. As water molecules enter and leave the cell walls, the resulting swelling or shrinkage is mainly perpendicular to the cell walls and does not influence their length. Similarly, pushing marbles into a straw broom would make the broom head wider, but would have little effect on the overall length of the broom head."

My general takeaway from his discussion is that because of the cells respond to moisture changes, dimensional changes along the grain (longitudinal) of a kiln dried board are so small they can be ignored (0.1% of length). Tangential and radial changes in a board's dimensions, however, can be significant, and these are the changes that can cause problems via shrinking and swelling. The nature of the movement will vary by species and with how the board was sawn.  Is this more or less what the major takeaway the primary source of wood movement should be? Just wanted to make sure I was starting from the right place for these questions.

Now regarding restraining and directing wood movement, if you happen to have the book, on page 89 Hoadley talks about how much an 18 inch panel should move and says "since the panel will be pinned at the center, we can assume symmetrical behavior and look at either half (of the board)." The source of my confusion is why we know that we can make this assumption. Is the idea that, in the center of the board, the pins are strong enough that changes in the wood cells near the pins is redirected away from the pins? I've tried to summarize how I'm envisioning this in the attached drawing, the red pins at the center redirect movement of the cells at the center of the board. These cells expand, pushing on the cells next to the, which then expand pushing on the cells next to them, and so on, so the ultimate movement is in the direction of the green arrows. 

In the case of my shiplap question (second picture), say you have 2 rows of restraints (red circles) on the board (glue, screws, etc). I'm thinking that the column of pins on the right direct cell expansion to the right (blue arrows) and left (green arrows). The column of pins on the left end of the board directs cell movement towards the center (green arrows). If this is what is going on, at some point the cells are going to hit one another (red box). What happens at that point? Will this generate some sort of crack? Or would this "bumping" cause the cells to "turn around" and redirect expansion towards the ends of the boards?

 

 

[TomInNC]

On a related note,  I really like the mix of science and woodworking in the Hoadley book. If anyone has references for sciency woodworking books like this, please send them along.

[Mark J]

19 minutes ago, TomInNC said:

Is this more or less what the major takeaway the primary source of wood movement should be?

Yup!

[Mark J]

26 minutes ago, TomInNC said:

"since the panel will be pinned at the center, we can assume symmetrical behavior and look at either half (of the board)." The source of my confusion is why we know that we can make this assumption. Is the idea that, in the center of the board, the pins are strong enough that changes in the wood cells near the pins is redirected away from the pins?

Yes, the expansion/contraction of the wood panel in the region of the pin will not be zero, but it will be so small that it can be ignored.  More simply put, it isn't enough movement to cause mischief.

I think I understand your ship lap question.  Each board or lap will expand/contract with humidity changes.  Just as in your drawing movement can be toward (or away) from where two laps meet (your red rectangle).  But there is a small gap between each pair of boards that accommodates the change in dimension.  In other words ship lapped boards are not butted up against each other.

[Chestnut]

1 hour ago, TomInNC said:

The source of my confusion is why we know that we can make this assumption. Is the idea that, in the center of the board, the pins are strong enough that changes in the wood cells near the pins is redirected away from the pins?

The assumption is that each side of the board from the pin will expand and contract similarly. Which in general is not true but is true enough for woodworking measurements. With a single attachment don't think of the pins as being strong just think of them as a reference point. The edges of the board as referenced by the location of the pin will move. Alternately if the reference point is the front edge then the pins and the rear edge are moving.... wait this is making things more confusing . I really enjoy physics.

When there is only 1 pin on a board/panel there isn't anything to push against. wood movement is an issue when the wood is constrained by 2 or more points.

1 hour ago, TomInNC said:

In the case of my shiplap question (second picture), say you have 2 rows of restraints (red circles) on the board (glue, screws, etc). I'm thinking that the column of pins on the right direct cell expansion to the right (blue arrows) and left (green arrows). The column of pins on the left end of the board directs cell movement towards the center (green arrows). If this is what is going on, at some point the cells are going to hit one another (red box). What happens at that point? Will this generate some sort of crack? Or would this "bumping" cause the cells to "turn around" and redirect expansion towards the ends of the boards?

Ship lap and also tongue and groove flooring is designed in a way that there are essentially small gaps between each board. When you pin on one side of the board the other side will expand into that small gap. Alternatively it will contract and make the gap larger. Ship lap and flooring are typically narrow strips so the gap is tiny and hard to see. I think it's important to recognize that ship lap properly installed is not a continuous panel. It is a combination of multiple individual panels or boards.

Also wood movement is a percentage game. 1% contraction of 4" is 0.04" or not really enough to make a difference. 1% contraction of 45" is 0.45" and needs to be accounted for.