I'm making a solid wood bar-type counter that will float above an island for bar-height eating. The width is 14" (two 7" wide boards) with 7ft. and 3ft. legs. Plan is to miter across the width, and I'm already committed to this approach. In theory, a wide miter like this will open up as the planks either expand or contract. So I want to glue it as solidly as possible and hope for the best. I'm ready to rout a stopped groove in the mating edges and use a 1/2" thick solid wood spline. I'd probably use epoxy to give a little more glue-up time. The spline will only be about 1" wide, with 1/2 inch in each side. Alternatively, I could also use my horizontal mortiser and rout 3 or 4 separate mortises that could be considerably deeper. Is grain direction important for the spline or loose tenons? The length of the miter joint is 20". If I knew the miter would open a little eventually, I might leave a little V-groove on the surface, but I'd rather have a flush, unbroken surface.
(Architectural Woodworking Forum)
From contributor B:
It seems to me that the miter glue line should stay intact with seasonal movement just as well as a side grain joint, particularly if you throw some spline or biscuits or loose tenons in there. It will shrink and swell as it breathes. Trick is to allow the angle of the L shape to open or close as necessary. Let the top float to move around a bit. The reason miters open up is that they are locked in on one side or the other. This doesn't allow for movement. Simple fact that it is a miter glue joint doesn't mean it will behave differently. Heck, a biscuited and glued butt joint will stay together, won't it?
I let the lumber sit in the shop for a week, then ripped 1 7/8 strips. Face glued immediately into 13 in wide boards. Jointed and planed to 1 ½ +, then immediately glued up the 13 in boards. Dressed and sanded to 100 gr, cut mitres and slots the next day. Let sit three more days before final sanding and mitre assembly.
On all three I used a splined mitre and structural epoxy. I routed a ½ in stopped slot and inserted a slightly undersized hardwood spline, grain oriented perpendicular to the mitre. Glued up the mitre with structural epoxy – the reason for the undersized spline is the structural epoxy I used doesn’t like to be crushed. I considered contributor A's idea of drawtites, but couldn’t on one, and found it unnecessary on the other two. I’d use them in the right situation.
The biggest problem is the mitre, as you say. Either you get a monster L, which is unbelievably ungainly, or you have to assemble on site. Personally, I like the monster L approach, as it allows me to control everything at the shop, before the install. The seam gets finished in a controlled environment, and it minimizes the steps I have to take with a customer looking over my shoulder. My biggest concern, when first using this approach, was cracking/breaking the mitre during transport/install. Not a worry in the world using structural epoxy. We’ve transported two monster Ls – one 25 inches wide and 12 feet one way and 7 feet the other – without a hitch. I’d do it that way every time if I could. Gotta be darn sure of your template, though.
On another we physically could not have gotten a monster L into the kitchen, so we had to assemble on site. The slightly undersized splines were a godsend, as the space was tight and the ideas I had to draw things together would not have worked as planned. Thing is, we had to shim a bit to lines things up, which was fine, but we forgot to account for the drawer opening under the spline (inset drawers – spent two hours remedying that mistake, but it had nothing to do with the top).
Just noticed that your L will be big, but not a monster. I’d seriously consider pre-gluing in the shop. I would advocate for my method only because it will transport more safely than biscuits, but once it’s attached, it won’t matter. What will definitely matter is how you attach it to the cab. No amount of epoxy, drawtites, glue or biscuits will keep it from cracking if you don’t allow for movement.
Bottom line is, there seems to be several ways to skin this cat, but in all cases you have to start with good fundamentals – good lumber, account for movement. After that, you should be okay.