Sandwich'ing

[Joe Craven]

Hey everyone,

Long time, no post (or real participation)...sorry about that.

I have an idea for an entry table that the wife has requested. She hasn't requested the sandwich'ing that I referring to, but it's something that I want to try. I plan to make two of these: one for us, the other as an auction item for a charity organization that has an upcoming gala event. I think I can convey this question without providing a diagram.

The concept is a circular table which is something like 28" wide (diameter of top) by 28" tall, with probably a shelf somewhere near the middle. The challenge that I have created for myself is to sandwich some lighter wood between darker wood, almost like an Oreo if you looked at it from the side. I have plans to integrate this same theme into the legs, so that the outer and inner woods kind of meld. That's not where I'm needing help, though.

Let's call these "layers" like outside Oreo cookie Layers (1 and 3) being walnut, and the inside cream Layer 2 being yellowheart.

My question is this: if you were just making a circular top of a table where you want to do this sandwich thing, what's the best way to glue disparate species of hardwood? My thoughts focus on three different methods...

Method 1: edge-glue panels of the same woods at [whatever appropriate thickness], then face-glue those full-width panels together to make the sandwich. Followed by cutting the circle shape on the bandsaw (or router jig, or whatever). I like this one the least because it requires consistent flattening of each "layer" of the sandwich before gluing up the whole thing. And then making sure you've got consistent pressure on the glue-up, across a wide surface.
Method 2: similarly, face-glue the disparate woods plank-by-plank, then joint (S2S) the edges, and edge-glue the planks together, very carefully. The problem with this approach - which is shared by the last - seems to me that the appearance of the end-grain (esp. for the lighter, middle layer of the sandwich) will look so different than the side grain, that the piece wouldn't look presentable unless you're able to place it in an acutely-shaped corner...something that doesn't exist in most homes. And, for either of these two methods, do I need to be conscious of grain direction or anything else?
Method 3: this is where I need experienced advice, I think; glue up the outside panels (Layers 1 & 3), and sandwich them around MDF as Layer 2...the cream. Cut the MDF layer maybe 1/4" less diameter, and then just band (edge-trim, in a circle) the lighter wood that's been planed to a thickness that will bend on that radius. This solution only suffers from the line where the band joins its ends, and I suspect that having MDF in the middle of the sandwich might add long-term stability. I just don't have enough experience with mixing MDF with "real" wood to know if this is a good idea.

I'd be really interested in any response, even if just academic.

Thanks,
Joe

[Wayne Lomman]

The key to this design working is that the different materials need matching shrinkage rates. This data is readily available if you do a basic google search.

Method 1 is straightforward if you have a thicknesser. It's the simplest and the one I would choose.

Method 2 will only work if you have, again, accurate thicknesser and accurate edge gluing for the second stage.

Method 3 works with veneers but not so well with thicker overlay as the mdf is too stable.

What is the problem with seeing end grain? If something is round you have end grain unless you veneer it at which point it no longer appears to be solid timber. Cheers

[Jim Becker]

Wayne is absolutely correct...for best results, you'll have to select your "sandwich" species very carefully to minimize wood movement challenges and also keep the grain direction for the layers absolutely consistent with each other. I see no way to avoid end-grain unless you build it solid and edge band to simulate the layers.

[Joe Craven]

Thanks Wayne - this is very helpful. My thinking w/ the end grain is that it's going to be a different color than the edge grain when I apply the finish: specifically, darker on the end grain with wood like yellowheart. I was hoping for consistency all the way around. I will do the research on shrinkage rates and go from there. I do have a planer and I agree that it's the better method of the first two to make sure that the layers are consistent throughout the piece.

Thanks again!

[andy bessette]

How about veneered plywood with solid edge banding?

[Wayne Lomman]

Colour difference can be managed with sanding and staining techniques. More later. Cheers

[Wayne Lomman]

Following on from above, if the job is clear coated, you can selectively bleach darker areas before polishing. If you are staining, finer sanding will absorb less stain than coarser sanding, so you can control how dark your job gets this way. Cheers

[Joe Jensen]

I would not sandwich because of the differing expansion and contraction of different woods. There is an excellent book titled "Understanding Wood" that explains this very well and they have tables with all the woods and the radial and tangential expansion rates. I live in Arizona where it's never super humid. We go from maybe 6% humidity at the low end to maybe 70% when it's really humid for a few months. Indoors my humidity swings are much less. maybe 20%-40%. I have a flat sawn red oak kitchen table I made that is 37" wide. The width changes like 3/8" over the course of the year. Flat sawn moves much more that quarter sawn. If you want to build to last get the book and study.

I would do a solid wood top, and then do the band you want with solid wood as a ring around the edge along with another ring of the wood used for the top.

[Joe Craven]

OK so it turns out that this sandwiching technique is not impossible, but I think that it requires a press rather than simple clamping. I learned this at the end of my glue-ups when the stock that I was using was narrow enough to only need a few inches of width. So rather than using the clamping jig that I had devised for most of the project, I used the infeed table on the jointer and then essentially pressed the boards together given the narrow width. Those final glue-ups were more consistent for sure, because I had a dead-flat surface and plenty of downward pressure, I think.

At any rate, here's the finished prototype after one coat of BLO. It's for the wife and she's not too picky, thankfully. Thanks to everyone for your help.

FirstCoat.jpg

[Erik Christensen]

it's kinda of a bummer when somebody's prototype looks better than my finished project