I need to create a panel that is about 24" square and about 1.5" thick and I want it to be dimensionally stable and not prone to splitting or warping. The material will probably be oak. What is the general pattern to approach the lay up?
What is the general pattern to approach the lay up?
Slice your boards into thin veneers and glue the veneers up in a stack, with the grain of each layer at 90° to the layer below it. Put the stack in a press to cure under pressure. Trim the edges and cut to final size.
Now, you're probably not actually going to make your own plywood, but that's how you get dimensional stability. Solid wood panels expand and contract with changes in humidity, and there's not much you can do to stop it short of controlling the humidity. Allowing for seasonal movement is an important consideration whenever you're making something out of solid wood.
So, if you want a panel that looks like it's solid hardwood, make it out of plywood (or another stable substrate such as MDF), and then wrap the panel in thin pieces of hardwood. Be sure to use end grain pieces where the ends of the boards making the panel would be, edge grain on the other sides, and face grain for the pieces. If you do it carefully, you should be able to hide the joints pretty almost entirely.
Up to this point, I've understood your use of panel to mean a flat piece of wood with no moving parts, but maybe you just mean that you want a 24" square thing made of wood, and you're OK with letting some parts move. In that case, a traditional approach would be to use a frame and panel design, where a solid panel "floats" inside a frame that has enough room for the panel to move seasonally. Since the frame is made up of narrow parts, it's relatively stable.
Another method would be to not glue the boards making up the panel together, and to hold the panel together with e.g. breadboard ends. Usually, breadboard ends are pinned in the middle and the panel is allowed to expand and contract at the edges. However, if you use joinery like lap joints, tongue and groove, or spline joints, and if you don't glue them, and you leave enough space for seasonal movement, you could pin the breadboard ends at the ends so that the panel's width stays the same.
Ok, so you recommend laying up the pieces at 90-degree angles. In butcher block and things like that, it seems like they have the grain going the same way. Jul 29, 2022 at 11:57
@TreowWyrhta I was describing the way plywood is made from thin veneers laid in alternating directions. I was mostly kidding because making your own plywood would be extremely difficult, but also wanted to describe why plywood is stable.– CalebJul 29, 2022 at 15:10
You can't create a dimensionally stable panel in solid wood under normal circumstances.
This is because wood naturally responds to changes in humidity by changing EMC. The length will remain effectively unchanged; thickness too (but only because that dimension is so small any changes due to moisture takeup and loss are small enough they're hard to measure). Width on the other hand will continually vary (but see footnote 2).
No guesswork is needed for how much the width might change, you can calculate the expected movement in service for oak, and alternative hardwoods, for your location using tables published online. See What is the maximum width for a full cross grain glue up? These tables use data usually derived from figures published by the FPL, go directly to the source for coefficients of movement for all commercially important US domestic woods.
As the data will suggest, you could almost ignore movement if you select the right species and with the correct grain orientation. See which direction does wood expand?
In addition to selecting for minimal movement with quarter- or rift-sawn boards (do note the issue with oak in regards to the sometimes extreme figure in QS material however) you can minimise wood movement with species selection, although availability and cost are definitely practical considerations now, due to Covid.
What general considerations do I need to take into account for wood movement?
What is the maximum width for a full cross grain glue up?
splitting or warping
Splitting or warping are separate issues to dimensional stability.
Unless the wood chosen is unsuitable (e.g. too wet before use, or badly dried) it should not split in service, same as almost any tabletop or cabinet panel would not normally be expected to split.
Warping can be similar, in that with good stock selection you help ensure, or at least minimise, the potential for any of the boards that make up the panel to contribute to a specific type of warp. Once a panel is made however it starts to act like a single board to a degree, and there's always the potential for cupping if one side can dry more than the other.
But circumstances and use are factors, especially over the long term.
- How the wood with be fixtured to (as in a tabletop), or held by (as in a floating panel), whatever else it needs to interact with.
- If it's intended to be free-floating or nearly so, stiffeners can be installed on the back if felt necessary. In the past this was often achieved with cross-grain stiffening strips (battens) running in dovetailed slots1. Now steel C-channel might be used for the same purpose and if installed correctly the resulting panel can remain the same thickness, a trick harder to achieve historically.
- Whether the wood will be finished and with what, and to what thickness2.
1 This is one of the classic ways drawing boards were built to stay flat prior to the widespread use of plywood
2 Coat any wood in a thick enough application of any film finish and moisture uptake and loss can be reduced to almost zero. But this 'encased' look isn't to every taste and isn't suitable for many applications anyway.