Questions/concerns about shrinkage/expansion when combining hardwoods

[ear3]

I've been commissioned to make a large floating shelf (11ft in length!), along the lines of this:

[attachthumb=1]

[attachthumb=2]

Wood selection has yet to be determined, but I'm leaning towards Walnut, as I can get this pre-surfaced and so won't have to worry as much about warped/twisted boards milled out of rough stock for such a long piece.  The visible parts of the shelf will be Walnut, including probably the bottom.  But the back of the shelf, which will probably be designed to operate as a hidden cleat to hang the shelf, can be some other, less expensive wood.  Whatever the final design ends up being, I presume that the back of the shelf will be joined in some fashion to the sides using a butt joint and dominoes -- this means, though that I will be joining two different woods with varying rates of shrinkage.  So my question is about what guidelines/rules there are for mixing wood species -- is there some way of using the technical data, like the tangential and radial shrinkage values commonly available, to know which woods are good to mix and which not?  Is something like oak or maple for an invisible load bearing element of a piece always a safe choice, no matter what the main wood is?  Thanks for any advice.

 

[Jesse Cloud]

Yes, you can use tangential and radial shrinkage ratios to calculate an estimate of the amount wood will move with a given change in humidity.  For instance, my 40 inch wide walnut tabletop should move about 1/8 inch with a change in humidity from 10 per cent to 60 percent.

Cross grain movement is much larger than movement parallel to the grain.  Probably a larger issue than mixing species of wood will be the fact that for the piece shown in your pictures, the top will move in a different direction than the sides.

There are many work-arounds to accommodate wood movement.  You can join pieces with screws in a slot, with a sliding dovetail, with  "tabletop fasteners that slide in a groove on the apron, or with a combination of glued and not-glued tenons (glue the center one, let the others float in a wider mortise).

You might consider mounting the shelf with metal brackets that are slotted to allow a screw to move.  If you want to keep the front of the shelf flush with the side, then use a fixed screw at the front and slotted screws further back, leaving a small gap in the back for growth if you are installing in a low humidity season.  Attach the top to the sides with tabletop fasteners.

Alternatively, make the shelf out of plywood or mdf and veneer it.  Not enough movement to worry about.

 

[ear3]

Thanks [member=457]Jesse Cloud[/member]  I think I've come up with a solution as far as the wood goes.  I can get common rustic grade walnut for less than half the price, so provided the boards are straight, I think I can just go and use that for the non-visible surface.

And thanks for the suggestions on the design.  I think I'm going to sink the top, horizontal shelf in a dado in the apron and sides -- I thought about a sliding dovetail, but it strikes me that it might get messy fitting a ten and a half foot sliding dovetail joint if it's a tight fit.  Though I haven't actually used that type of joint since I got my Festool routers, so maybe I'm underestimating the precision I could get.

I'm going to work up a design though and maybe post it later for comments/suggestions.