My kitchen is separated from a main room by a pier with a curved base:


The main room is all white cedar (walls, ceiling) and ash flooring -- it's a log cabin. I decided to replace the white vinyl baseboard molding with rough sawn cedar to match the rest of the room and house. Steam bending cedar is difficult (as are most soft woods), so I machined veneers (about 0.040" thick x4) and molded them without using steam to a form, Then, I added a top "finish" veneer (about 0.050") to match the surface finish and thickness of the trim that will be used on the long flat section. The top veneer is extended well past the bend and is intended to overlap a thinner (0.160") backing so as to hide the joint between curved and straight sections while matching the thickness (about 0.210") of the straight section of molding.




My question is whether that backing strip should abut the laminated curved portion or whether to leave a small gap, such as 0.125" to 0.25". That joint will always be a stress point. I am thinking that no gap will make it stiffer and failure, if it happens, will be due to tension on the outer, finish veneer. On the other hand, a small gap will be more flexible, but also yield to compressive failure.

Opinions? Gap or no gap?

John
PS: Those clamps you see were not used for bending. I used 1/2" screw band clamps around the whole form for that. They are only in place to hold another 12 hours for full cure.

 

Gap or no gap?

Personally I'd go with a small gap, 1/16" or so. Wood no matter how well you varnish or urethane it will still move with the humidity. Think of the new interlocking hard wood flooring instalations, they want it to not be hard against the wall, to keep it from buckling as the seasons change.

 

Opinions? Gap or no gap?

I agree with @shortbus. I’d leave a small gap. Expansion and compression at the joint will either buckle the veneer (if the joint is glued strongly, such as if it is epoxied) or stress the glue joint until it fails (popping the end out) and the end separates.

A joint on a curve is always difficult. If the are adjoining straight runs and I have a sufficient length of veneer, I’d use one long piece in the curve and make joints on the straightaways.

 

Wood no matter how well you varnish or urethane it will still move with the humidity.

Tell me about it. My log home moves a lot.

I cut all the pieces, including laminations, from a single board using a bandsaw. I suspect any joint beneath the top laminate will eventually "telegraph" its presence in some lighting and humidity conditions. My aim is to minimize that, and my bias is the same as yours. That is leave a small gap. I just ordered some flexible epoxy from PC-epoxy.com (PC-Fast-N EZ epoxy). The user datasheet says it is Shore D60, but PC's demos seem to show it is more flexible than one might assume that to be. I was thinking of filling the joint with it.

I considered rabbeting both pieces on the back and putting a laminate over the joint, but that would be a lot more work. What do you think?

 

Did you glue the lamination's when you bent them? One thing you do have going for you is, from what I think I see, is it's at floor level, so it won't be that noticeable. And DJ's idea seems like a good one if at all possible to do, a single strip that goes completely around the base.

I haven't done much veneer work but do know some people use scarf joints when joining instead of a butt joint.

 

I agree with @shortbus. I’d leave a small gap. Expansion and compression at the joint will either buckle the veneer (if the joint is glued strongly, such as if it is epoxied) or stress the glue joint until it fails (popping the end out) and the end separates.

A joint on a curve is always difficult. If the are adjoining straight runs and I have a sufficient length of veneer, I’d use one long piece in the curve and make joints on the straightaways.

The visible joint will be on the straight part. The joint I am talking about is where the curve joins the straight section. Here's a drawing (not to scale):

Right now, I am at the point of gluing the backing to the straight portion of the face laminate. Why did I do it like that? I used a concrete form for my mold and strap clamps for making the curved laminate. Had the width of the thing been smaller, I might have made a form that would have laminated both the straight and curved sections at the same time.

Edit: Added more detail to drawing.

 

@shortbus
Yes, the curve backing is 4 laminations that were glues and cured. The "backing" is the same thickness as those 4 laminates (after gluing). The the whole thing is covered with a face laminate. That way, I was able to use a concrete form that was very close to the exact diameter I needed.

Edit:
On straight pieces, I have used scarf joints. On a curve, they are much harder to make. I consider the face laminate as a compromise.

 

UPDATE
I decided to bite the bullet and made mating rabbets on both sides. The flat piece was easily done with my saw and only needed minor adjustment. The curved edge was a little more difficult and involved more hand work, but not as bad as I had feared.

Now, it will be a hare bit stronger. Glue is setting. Final pictures will follow.

John

 

Done
Here are some pictures.

1) Full outside view as it will mount under cabinets (skew is mostly from camera perspective):

2) Inside of lapped rabbet joint (left a little space to hopefully avoid telegraphing the joint):

3) Inside showing slight excess top and bottom for trimming, if needed:


It sits pretty flat and curved section has very little skew, if any. The cabinet base to which it attaches is a different story. Fun project. Got a 1/2" 4 tpi blade for my bandsaw to get the rough sawn look on the re-sawed stock.

Curved end will be trimmed about 1-1/2" in length to meet existing molding. Will attach with #6 flathead brass screws. All I need is some 1/2" Tee molding for that junction. Easy to find in large rolls -- I only need 4".