How strong are these legs?

[Prashun Patel]

I'm making a dining room table and the client is in love with these table legs. What is the strongest way to make these legs?

My instinct is to do a bridle joint.

How would you do it?

Sorry it's rotated!!!

[Jamie Buxton]

There's three joints. On the big L-shaped pieces, a bridle will do. Then there's a mortise-and-tenon connecting the L-shaped pieces to the crossbar. But then there's a challenging one: there are Y-shaped joints connecting the four little pieces of the crossbar to the long piece of the crossbar. I've done something like that with multiple Y-shaped loose tenons.

[Prashun Patel]

Thanks a lot, Jamie.

I think the Y stretcher is narrow enough that I can make that out of a single straight grained piece. I am pretty sure I can minimize the grain run out on the branched sections. In fact, I may just laminate that stretcher from 2 pieces.

That three-way joint in the leg bothers me. I don't want to compromise the tenon on either piece or make the leg mortises too large.

Also, I should have mentioned that the top will be 6/4+ walnut about 40" x 108". Do you think I'll need a third leg assembly in the middle?

[Jamie Buxton]

My guess is that the bridle-plus-M&T isn't the challenge. The L-shaped members are beefy, and the bridle joint can have lots of glue area. The mortise sticking into the bridle can be substantial, but the bridle is big enough that there will still be lots of strength in it. I'm more concerned about the Y-shaped joinery. All the pieces joining there have long leverage, and could really stress that joinery.

I have one additional concern: rotation. Imagine a vertical axis through the middle of the table. Have the table sitting on the floor, so the "feet" don't move. Now attempt to rotate the top around that vertical axis. I can imagine that the structure is going to flex enough to make the table feel shaky. If I were building the thing, I might prototype it first in poplar, or at least not buy the glass until I've established that the base works.

Third thought.. The L-shaped leg structures in the pic look like they're wood. However, it might be that that the stretcher structure is metal. It seems to be a different color than the wood. Metal would be stiffer so there would be less rotation shakiness, and the Y-shaped joinery would be much stronger.

The glass, incidentally, could be expensive. Glass shops will custom-make thick glass tops to your dimensions, but they charge a lot. A more cost-effective approach is to identify a "replacement top" (google) which somebody is already stocking, and build the base to fit those dimensions.

[Jamie Buxton]

Here's a thought.. Replace the five-part stretcher structure with single piece of wood -- say a 3x3 timber. The big L-shaped leg structures fasten directly to the 3x3. This would fix my worries about the shakiness, and the complicated y-shaped joinery. Visually, I don't think it harms the piece. You still get the big L-shapes under the glass top.

[Jamie Buxton]

The more I look at your pic, the more I think that the Y-shaped area is metal. Look carefully at the Y-shaped area that is farthest from the camera. There are color shifts from the stretcher to the L-shaped legs, but there is also a color shift a few inches along the long stretcher. It might be that the Y-shaped joining piece is a three-armed piece of metal, and that the legs and long stretcher are hooked to it by bolts that somehow install from below.

[Prashun Patel]

Jamie-
The table is an inspiration for shape only - not materials.

The entire final project will be in wood. Walnut top, walnut stretcher, walnut legs.

[Jamie Buxton]

Jamie-
The table is an inspiration for shape only - not materials.

The entire final project will be in wood. Walnut top, walnut stretcher, walnut legs.

Good, but I'm thinking that the thin y-shaped pieces may work okay in metal, but be too flimsy in wood unless you beef them up a bit.

[Steve Baumgartner]

I'm inclined to agree with Jamie - in the photo there is an area in the Y that looks like it might be a weld. I'd be really concerned about the strength of that Y joint. It takes stress in multiple directions.

Steve

[Prashun Patel]

Great to know. What if I laminate that stretcher so that each ply contains the entire Y? The laminations would stacked, vertically.

[Jamie Buxton]

Great to know. What if I laminate that stretcher so that each ply contains the entire Y? The laminations would stacked, vertically.

That would certainly help fend off the short-grain breakage issue, which is good. I'd probably give the stretcher a larger cross-section, to reduce worries about shakiness.

[Frank Drew]

Great to know. What if I laminate that stretcher so that each ply contains the entire Y? The laminations would stacked, vertically.

Prashun,

I'm not wild about the design of that understructure, but if that's what the client really wants.... then I think I might laminate the stretcher but in the opposite way to what you've proposed, stacking horizontally, if you will, and I'd do the lamination in two halves, each containing half of the Y at each end, then I'd glue the laminations together back to back. Either mortise the ends of the Ys for loose tenons, or incorporate tenons in the laminations (include extra length in the middle layers, or saw tenons in the finished shapes.)

It's going to be a complicated assembly, no matter how you do it!

[Prashun Patel]

I've done a smaller coffee table with a similar 'y' stretcher, and because there is no apron, the assembly is actually very easy. I'm not crazy about the design either.

She's open to alternatives. Need to put some drawings together. Dagnabbit Googlesketchup! I just never had the patience to learn that thing. Boy do I sound old.

[Gene Davis]

Buy a nibbler and a TIG welder. If you like working in wood, you'll love working in steel.