Joinery for a U-Shaped Wood Countertop

Given the natural tendency for wood to move, joinery for a complex wood countertop requires some careful thinking. December 1, 2010

I'm meeting a client this morning that would like a wood countertop in her kitchen. The layout is U-shaped, so for the continuous top she desires there will be two joints. I was thinking of using miter joints and MAS epoxy, but in scanning WOODWEB, I've seen a couple of warnings about miters in this situation. Why? I would have thought this would be the best way to go.

Forum Responses
(Cabinetmaking Forum)
From contributor D:
If your tops are 24" wide and have a nice 45 degree cut on one end, and they mate together, all is well, for starters. As we know, wood expands and contracts across the grain in response to changes in humidity. Let's assume you used 6% sugar maple in a mix of flat sawn and quartered pieces. Then the humidity picks up in the house as all the extended family come to visit and check out the new kitchen. Cooking, cleaning, showering, etc. all increase. With all the excitement, maintenance on the tops is not done, and the MC increases to 11%.

This increase in MC will change the width of the tops by almost 3/8" - to 24-3/8". This will effectively change the angle of the miter from 45 degrees to something other than 45 degrees, opening the points or toe of the miter. I'll let someone else do the math. The family all leave, and you get a phone call about the "tops falling apart."

Let's assume you anticipated all the above - astute woodworker that you are - and you added a 3/4" thick, 10" wide piece of plywood under the miter and glued and screwed it in place to reinforce the perfect miter. Then, as the humidity increases, each board - each fiber - will attempt to swell but be restricted. Each fiber - board - will then crush just a little bit to accommodate the attempted expansion. Now, when the tops dry back out in time, little hairline cracks are likely to develop all around that area as the now crushed wood shrinks back in place.

If the humidity stays stable at all times, the miters will be okay, but you do not have control over the humidity. Even if you float the tops and use a spline that is short grain - oriented about the same as the grain in the tops - the angle of the miters will change in time.

This is why mitered solid wood moldings are limited in practical use to about 4-1/2" wide and two or more pieces are used to build to wider widths.

From contributor J:
I would use a miter joint for looks. I would not epoxy or glue anything. You might like to use some silicone or Siroflex (good brand) at the joint. I would use dog bones routed into the bottom; just snug them, though. I think this would work real good.

From contributor P:
What contributor D said. You can learn from him or find out the hard way, but there really is no education from the second kick of the mule. All kidding aside, kudos to you because you asked the question in the first place.

From contributor A:
Two butt joints with the dog bone connectors. It looks just fine and it won't fall apart.

From contributor B:
I'm about to build a U-shaped layout curly maple kitchen counter for my place. My plan is to do a sliding dovetail joint at the corners, with the counter sections meeting at 90 degrees. The end section of the U will butt into the front edges of the two sides. The end section will have male dovetails running most of the width of both ends, while the front edges of the side sections will have the female receiving slots in the area where the counters intersect.

My thinking is that the end section of the U, which will be secured at the front edge, can expand and contract at the edge along the wall, with the movement hidden below the maple backsplash. The movement will be accommodated at either end in the unglued sliding dovetail.

The side wall (sides of the U) counters will also be affixed at the front edge, allowing for movement under the backsplash. With the front fixed in place, and locked to the butt ends of the perpendicular counter surface with the tight fitting sliding dovetail, no gap can develop (at least theoretically) between the counters at the joint.

From contributor P:
One other caveat - if you are going to oil the tops, or use any other finish, make sure you finish the under side with as much, or even more material than the top.

From contributor L:
Use miters and regret it! Do like contributor B said. You can't stop wood from moving! Accommodate it.

From contributor D:
Contributor B describes an excellent method of dealing with solid wood tops in corners. He also gives good advice as to where to fasten firmly and lets the rest float under a backsplash.

There is a way to get mitered corners if that is a driving criteria. Like anything, all it takes is time and money. We call it "faux solid" and strip 1/8" thick solids onto both sides of a stable corestock, adding end grain and thick edge bands as needed. Then the balanced panels can be joined as miters, dog legs, etc. But the tops are not solid.

From contributor O:
Cut your boards .75 thick by whatever thickness you want the countertop to be - say 1.25". Now face glue all the boards into your countertop. Because of the orientation of the wood, the expansion and shrinking is significantly reduced compared to edge gluing the boards. This is how workbench tops and breadboards are made. Then you could use a spline in your miter joint and use counter bolts to hold the joint together. The spline will keep it aligned and the counter bolts will hold it together. Because no glue is being used in the joint, what little movement of the wood will be handled by the spline and the counter bolts.

From contributor C:
The suggestion that you rip into 1.25" thick parts and rotate 90 degrees is making an assumption that you are using flat sawn material, and the rotation will thereby change your material into quarter sawn grain orientation. Fine, well and good, as the expansion and contraction properties are reduced when the wood is quartersawn, however, the joint will still open up by a significant amount. There are rarely any shortcuts available in the woodworking field. In short, don't do it for a customer.

From contributor L:
I couldn't agree less with contributor O! True, quartered does move somewhat less, but you can't stop it no matter how many dog bones you use! Back to contributor B's method.

From contributor O:
Not looking to stop movement (can't be done). Looking to engineer something that can handle the movement and not come apart. No glue is used in the joint. A spline and counter bolts are used to keep it aligned and allow for movement.

From contributor J:
Never done it. Doubt anybody here has. There is no way a 24" lamination will change width by 3/8 of an inch. Strong 1/16 tops. Don't do it for a customer is sage advice unless you can afford to re-make top should experiment fail. I would try it if: 1) quartersawn stable hardwood, 2) more than an inch thick, 3) not too firmly attached to bases by joint, 4) see previous post. I really think you will be alright, but that is just a guess.

From contributor B:
I would agree that a quartersawn orientation is going to be much more stable. However, like with plain sawn boards, some species will still move more than others.

I see two issues with this plan, though. First is that it is going to look substantially different from a plain sawn layout, and that may or may not satisfy your customer. Second is that while movement will be significantly reduced, even if it expands or contracts as little as 1/16" over a 2' deep counter, that is going to happen on both sides of the miter. That means that any gap that develops in the front of the miter due to shrinkage, or the back of the miter due to swelling, will be about 1 1/2 times greater than the amount of movement of each 24" deep surface. Plus once opened, that gap will likely remain.

From contributor D:
Remember the two Basic Rules of Woodworking:
1. All solid wood moves.
2. All woodworkers will disagree from this point forward.

However, the science is there, and both flat sawn and quartered will move, albeit quartersawn at about half the rate of flatsawn. Some species are much more stable than others. Thickness of the top/boards makes no difference, unless it is decreased to where it can be treated like veneer - 1/8" or less. Gluing or not gluing splines or dog bones or whatever will make no difference. The wood just doesn't care.