OK, I adjusted the table dimensions to severely minimize sag, but
still want to go for that stretcherless, trestleless, "minimalist"
look. Using red oak, will edge join (maybe w/ biscuits, maybe not) 16
3"x1"x72" planks for a 72x48x1 top. If I join these right and true,
the top itself should be sufficiently sturdy of it's own accord for
dining table purposes (the wood is about 80 pounds, I tested 280
pounds on the sagulator and still get less than 'visible' sag, with
legs at the very corners. That seemed like not enough sag to be
anywhere near failure, ihmo. Anything really wrong with that logic?
Even so, if we could assume for a moment I could design a such a top,
I would like to place the legs about 6" from each corner. Each leg
will be 3 1x3's laminated together. Since I'm going for no stretchers
or trestles, I have the following idea for joining the legs, which is
kind of a dado/through tenon hybrid:
top view: long side x-section: short side x-section:
(long side) |-----------| |------------|
|-------------- | | | |
| |---| |---| | |
| -------|| ||------- ---|| ||------
|----+++++++--- || || || ||
| +++++++ -------|| ||------- ---|| ||------
| +++++++ |---| |---| | |
|----+++++++--- | | | |
| | |
| | |
|-------------- | |
After laminating the legs into the 3x3 dimension, I'll make 2 1"x1"
dadoes, 1" from the top. The table top plank which this leg is
attached to will actually be 3 planks, one between the legs, and one
between each of the legs and the end of the top. So, the top will
actually be assembled with 2 of the planks being shorter, so it
appears to have 4 notches. The legs will go into these notches, with
one dado facing the end of the short plank. Then the "really" short
plank will go with it's end into the other dado, and it's edges
matching the surrounding 2 planks.
If I match the edges well enough and use good glue technique, I would
think the legs would be pretty darn sturdy for a dining table. Does
this sound reasonable?
The final issue is movement - namely the top planks 'wanting' to
expand/contract (tangentially) as much as .10" more than the legs
(radially) if they are both plainsawn. I don't know if I can negate
these or not.Even if the glue holds, it might cause a split in the top
plank somewhere. Applying enough tensile force to stretch a 3" plank
by .10" seems like enough to split it to me, but that's based on
totally uninformed guessing, not hard calculations - which I'd be
interested in learning how to do if someone could point me to a good
book with the relevent engineering education in it. Also, the legs
will tangentially expand/contract against the dado joints, but with
this design, that movement would only be about .03" unrestricted, as
it's 1" of wood thickness in this plane as opposed to 3", so I "think"
this wouldn't be a problem.