Wood movement question

[Tom Porter]

My sister gave me a great live edge English Walnut slab about 4 feet long, 15 inches wide and 3 inches or so thick. I'm going to build a bench, which will be a new experience, having thus far made smaller projects or things a little more utilitarian. It's early days in the planning but as I consider the design, I have a question regarding wood movement. I'm experienced enough to know one would should account for wood movement but not experience enough to know much about it. Ideally, I'd like to have legs with stretchers between them at the ends with a single stretcher running lengthwise and attaching the legs to the top with mortise and tenons. However, I don't think this would allow for wood movement and eventually there would be stress on the long stretcher. I suppose I could attach the legs to an upper stretcher (if that's the proper term-something more like an apron) and then screw said upper stretcher to the top with slightly oversize holes to allow for wood movement.

Like I said, I'm just at the planning stage but am I on the right track? Any advice? Other ideas?

[Jamie Buxton]

I'm not sure I understand, but... Wood expands and contracts across the grain, not along the grain. That is, the 4 foot dimension of your slab will not change. And the length of a stretcher running that four foot direction will also not change.

With a 3 inch thick slab, I'd use four legs, and mortise each one into the bottom of the slab. I'd make the legs kinda chunky. I would not add any stretchers; the slab is going to hold them well enough.

[Tom Porter]

Good point. I didn't mean the long stretcher; mistyped I guess. I thought about legs without stretchers but have little confidence in my ability to get them straight and square. I figured stretchers would give me a better chance to get them all square. Could be wrong about that...

[Andrew Seemann]

Assuming the board is already at moisture equilibrium, mostly you would need to worry about seasonal expansion/contraction across the grain, which over 15 inches of walnut could be around 3/8 to 3/4 or so, depending on the grain orientation, growth ring width, and your seasonal humidity change. The other thing to worry about is the board doing some cupping/uncupping, again due to seasonal moisture change; because of the curve in the growth ring the board will arch slightly towards and away from the growth rings due to the rings (its a geometry thing that is kind of hard to explain). That might be around 1/16 to 1/4, but that is a rough guess, having not seen the board. Feel free to post pictures of it, top view, side view, and end grain.

If you use stretchers, I'd recommend putting them lower on the leg, so that the leg has enough length to flex. Attaching them with screws or something that would allow a little flex rather than a rigid joint wouldn't be a bad idea either.

[Jamie Buxton]

Good point. I didn't mean the long stretcher; mistyped I guess. I thought about legs without stretchers but have little confidence in my ability to get them straight and square. I figured stretchers would give me a better chance to get them all square. Could be wrong about that...

Why do the legs have to be square and vertical? I'm seeing legs that are somewhat slanted out as they go down to the ground. They don't have to have exactly the same angle -- they're four feet apart, and nobody will ever know.

Part of my mental picture is based on furniture from early America. A bench was a tree trunk split in half, with legs mortised freestyle into the bottom of the half-trunk. A three-inch-thick slab -- with live edges -- is darn near a half of a tree trunk.

[Jim Becker]

Alignment of the grain direction on the legs with the grain direction of the top will help mitigate seasonal movement issues if, especially if the same species, whether the legs are narrow or a full "slab" type leg. I've built several benches/tables like you describe and there have been no issues because of that alignment.