Question on "Structural Miter" Joints

[Kesh Ikuma]

Hi, I got a question regarding a leg joinery for my ongoing desk build (my first standalone non-shop furniture).

I came up with an overall design then decided to work out the details as I build along. (This I'm quickly realizing was a risky decision...) The leg design is heavily inspired by Aled Lewis' Wish Desk (which was on the backcover of a fairly recent FWW issue). Being a FWW backcover story, the details of how to build the desk was not fully elaborated in the magazine (only how to attach the legs to the top). So, I took a stab at it and came up with the leg joinery:

343010

Ideally, I should've made these long loose tenons integral; but the designing-as-building bit me hard in this case as I already cut the leg pieces to have only 1"-long tenon. Literally within a second after the saw separated the last offcut, I realized 1" is probably too short for what it is (legs are milled from 6/4 stock). That is how the mixture of a long (3-1/2" - 4") loose tenon next to 3/4" integral tenon came about.

Am I overcomplicating this joinery? Any alternate ideas?

BTW, the front and rear legs on each side will be joined together by some yet-to-be-determined method, and there is a horizontal stretcher right behind where the legs are crossing.

Thanks for your inputs!
Kesh

[Wayne Lomman]

That's an engineering challenge. Your approach is fairly complex. If you have already machined everything and don't want to do it again, use a high strength glue and hope for the best. If you are prepared to do it again, consider a half lap joint. You only see one face so make the face side with the mitre line and the bulk of the joint behind. Good luck. Cheers

[Kesh Ikuma]

I did consider half-laps during initial design, but I wanted to stay true to the look of the Lewis' desk. The outer leg joint is exposed from above.

343017

Also, I wanted to mill the two parts as closely together as I could for the best grain continuity. These are the reasons why I chucked the half-lap idea.

I guess my question boils down to this. Do the integral tenons in my design do much? If The long loose tenon and splines can provide enough strength, it would simplify the construction substantially.

[Michelle Rich]

pleasedon't think I'm raining on your parade, but boy, have you bitten off a big chunk. You say you are a beginner, basically. That desk is complicated and an engineering marvel. I have been wwkng 30 yrs and make windsor chairs, and would find this table a huge challenge. the legs are the simple part. It is the ability to get those legs attached to the tabletop, considering the movement of the top getting all the mortises to line up would be daunting for most of us. I sure wish you luck, and post pictures of your building experience.

[Kesh Ikuma]

Since my question was on the leg joinery, I left out my biggest cheat :D The top will be made of appleply, which I purchased 4 years ago when I promised my wife of her desk (surely you know how that goes, right? lol). Ply top will simplify top-leg joinery greatly, and I'll use M-T joint (you can see the tenon outlined in my drawing) with a couple screws to join top and front legs.

Being an engineer by trade, I'm drawn to well-engineered piece like Lewis' desk. His sliding dovetail to join the top and front leg is a genius design, and the most trickiest part of the build, had I copied his design entirely. Eliminating this and breaking down the project into parts, I decided this build is manageable with my current skill and tool sets. We'll see. I'll try to snap a few pics along the way (it's hard though)

[Al Launier]

That is a beautiful piece of furniture, delicate with strength. What are the O/A dimensions? Your wife must really be pleased with the design.

Some questions that will display my ignorance of this design:

Will the stretcher be attached to the rear of the drawers?
What is the extension behind the drawers for?
And whate is the rectangular hole in the top for?


Keep the pictures coming - interesting project!

[Kesh Ikuma]

Thanks, Al. But let's not get ahead of ourselves here. There is still a possibility of her ending up with a pile of scrap wood instead of a desk. haha


What are the O/A dimensions?

Pretty standard desk dimension: 5' x 2' x 30" The depth is governed by that of the ply in hand.




Will the stretcher be attached to the rear of the drawers?



No, the stretcher and the drawers will be completely detached. The drawers are suspended from the top with a threaded rod and its side is attached to the front leg with screws. Here is the shot from under the desk:

343023



2. What is the extension behind the drawers for?


It is intended to host the subwoofer of her computer speaker system.



3. And whate is the rectangular hole in the top for?


That is the recess to hold a power strip and for cable management.

Mind you, the plans for the drawer assembly and the recess are nowhere near final. I figured I will wait to make the final say until I have a dry-fitted desk (top + legs) in front of me. It will give me much better sense of their proportion and function.

[pat warner]

I'd rout a seat for some aluminum, (http://patwarner.com/images/metal_wood.jpg) a mending plate, maybe on the back side if you don't like hardware.

[Kesh Ikuma]

@Pat - That's an intriguing idea. Although having a mending plate on a show side quickly got shot down by my dear customer, I'm contemplating to use it along the arched underside (say 1" wide plate on 1-1/2" thick wood). A couple questions.

Do you think there is enough strength if I replace the loose tenon with the mending plate? The short tenon would be extended in place of the loose tenon in this scenario.

What thickness Al stock do you recommend?

Thanks!

[Mike Cutler]

Kesh

I agree with Pat about using a blind mending plate. That would put a lot of strength into the joint(s)
I would use both the mending plate and the tenon, but nothing says that tenon can't be replaced with 1/4" drill stock epoxied in place. ;) I think you'll find that the 6/4 legs are going to actually be stronger than you think. Of course people tend to lean on desks, so you will need some extra "beef".
I made a piece for my wife, a library shelf system, and used blind, and recessed, drill rod and all thread. The piece has an "airy" look to it, but it's pretty stout.

It's a nice desk

[pat warner]

"What thickness Al stock do you recommend?"
***********************************************
1/4 or 3/8 is plenty. Its strength will depend on your fasteners, how many & where they're placed.
I tap (http://patwarner.com/images/pwinsert.jpg) and use 1/4-20 cap screws.
Its nest should be sweet and without much play.

[Wayne Lomman]

It's a bit hard to describe but the half lap still works. To me it is intuitive but that's the way I design. The best I can say is that you stop the inner half lap just below the top edge.

I have doubts about the longevity of your proposed design as there seems to be too much end grain gluing which is not so good.

If you use the embedded steel or aluminium plates, make sure the back is roughened preferably blasted and epoxy glue in place. Metal to timber will fail over time if not fully bonded as they are such dissimilar materials.

Composite construction like this is good. I am currently working on a dining suite for a client where this is the only way to do it. The table is 3 metres long on a single pedestal. Steel inside plywood core is the construction method so that it is slender and strong which is what you are trying to achieve. But remember, the reinforcing plate must be bonded to work. This is where flat pack furniture really fails among other reasons of course. Cheers

[Kesh Ikuma]

Thanks all for the great suggestions.



tenon can't be replaced with 1/4" drill stock epoxied in place.


This might work even better than mending plate for me! I've been thinking about how to jig up to flush mount the mending plate on the arched bottom surface (and how to bend the plate to the right radius). I came up with an idea but more complicated than I like (essentially a rocking router base). Metal rod, on the other hand, means I just need a drilling/doweling jig, which is a whole lot simpler proposition.

Because I'm a paranoid, I'd like to use a thicker aluminum rod (3/8" maybe?). Knowing myself, I'll hit the rod with a router bit when I create mortises on these legs for the stretcher.



If you use the embedded steel or aluminium plates, make sure the back is roughened preferably blasted and epoxy glue in place.


Thanks for this tip. I never thought of this. Mind elaborating a bit on "roughened"? Is sanding the rod surface to 120 grit sufficient?

[Wayne Lomman]

It should be rough enough to file your finger nails on the metal surface. If you burnish it, you will have wasted your effort. Use sharp paper and change it often. Abrasive blasting is best as you get myriad peaks and valleys about 60-70 microns in height and the resulting increase in surface area is astronomical. Spent far too of my life in the abrasive blasting and industrial painting industry where this is a crucial starting point for any job.

By the way, can you not put a flat plate over the joint on the inside face instead of inside the radius or drilling pins in place? Cheers

[mark mcfarlane]

@Pat - That's an intriguing idea. Although having a mending plate on a show side quickly got shot down by my dear customer, I'm contemplating to use it along the arched underside (say 1" wide plate on 1-1/2" thick wood). ...

Kush, If the boss is against metal/wood, and if you are a member of FWW, show her some of the pics from this recent article. http://www.finewoodworking.com/how-to/article/marriage-of-metal-and-wood.aspx. It changed my life (and got me researching milling machines).

[rudy de haas]

We have a piece of furniture we call "pretty desk" - Danish teak from the 1940s or early 50s. The front legs on it are angled and the drawers "float" 3 inches or so below the bottom of the desk surface. The maker finessed the issue you face by moving the back legs to a more vertical position (so not like your design that way) and by extending the side of the drawer box up and back a little bit relative to your design. If you did that, the side of the box would transmit force downward to the lower two thirds of the leg while preventing sideways movement from stressing the joint. You do not say in your original post what wood you're using, but if this is a reasonable hardwood like Santos Mahogany that should let someone sit on the front of the desk without damaging it.

Further (and please forgive me if you think I'm over-stepping here) that desk design looks like it fits an older PC style computer user. My son's new imac has one small cable, for power: networking, the keyboard, and the mouse are all wireless. Speakers are built in, but wireless speakers are now pretty much standard for external use too. Since Macs are now just PCs with more expensive logos running BSD, I assume you can get comparable wintel world products - and planning for fewer cables and no keyboard tray should let you simplify that design a bit.

[Kesh Ikuma]

@Wayne - Thanks for the info. I'll see if I can source somebody with sandblaster (that's the equipment right?) locally.

On the mending plate, now, I think about it, yeah. I can route the mortise flat bottom, variable depth. Fit an aluminum bar thick enough for the deepest parts (ends), and pattern route the aluminum to match the wood. I did a quick check and using 4"-long Al stock yields 3/16" depth difference between the middle and ends. So, 3/8"-thick bar stock should suffice.

This being said, it's tough to justify the mending plate route (at least to myself). The process is much more involved, but it offers neither aesthetics advantage (can't see'em unless you crawl under it) nor structural benefit over metal dowels + splines. Here is why I think it doesn't offer additional structural strength. These joints in everyday use are subject to shear stress (sliding) and cleavage stress (pivoting about the bottom edge) and a very little of tensile stress (pulling apart). To my eyes and intuition, the structural advantage of mending plates over dowels is against the tensile stress, which we don't really need to worry about (unless I bolt down the legs to the floor and try to lift up the desk ;))

Yes, I'm over-thinking this...

[Kesh Ikuma]

@Mark - Good call. That article probably was the reason I could accept Pat's idea without any knee-jerk reaction. However, the response I got was "Meh" when showed her the article this morning. lol

[mark mcfarlane]

@Mark - Good call. That article probably was the reason I could accept Pat's idea without any knee-jerk reaction. However, the response I got was "Meh" when showed her the article this morning. lol

It's hard to argue against a 'Meh'.

[Kesh Ikuma]

@Rudy - I did consider various leg placements when I was designing the overall look. The vertical rear legs, without any doubt, would have made things a lot simpler, but also takes away the iconic look of Wish Desk.

> Drawer placement

Hmmm, I cannot quite visualize what you mean by your description. Mine is also about 3 inches or so below the desktop.

> iMac

She's a MBA user for some time (it's her 2nd MBA; I've been trying to bring her back from the dark side but has been unsuccessful so far lol) but prefers to use separate monitor, keyboard, mouse, and speakers. (She only uses it as laptop when she goes off to conferences.) The idea is for the MBA to sit on top of the drawers, and all the cables are to be neatly tucked away under the desk (power, video, audio, & USB). The desktop "cable access panel" is to give her an easy access to the power strip if she needs to use a wall-powered device at her desk. I'm not 100% sold on the current implementation though; so, it'll most likely change by the time I actually get to it.

BTW, my other hobby is building PCs (and future planned hobby of building speakers); so, I won't let her drift off to those all-in-one's too easily :D

> Keyboard tray

This actually is a requirement. She has been using a keyboard tray/drawer all her life and does not like to type at the normal desktop height. Since this will be her desk for all purposes, lowering desktop was not an option. I already had to talk her into a single-rail tray instead of a keyboard drawer. I wouldn't like at all the look of a keyboard drawer with this style desk.

[rudy de haas]

[QUOTE=Kesh Ikuma;2598141]
> Drawer placement

Hmmm, I cannot quite visualize what you mean by your description. Mine is also about 3 inches or so below the desktop.

Assuming the photo attachment worked.. you can see that the designer used the full sidewall as a design element. Your design is, I think, much nicer than this, but the point is that most of the sidewall on pretty desk is decorative rather than functional. Looking at your diagram I'd say you could use just enough of a sidewall to extend part of the drawer cabinet side up and back to connect the drawer, the top., and the joint where the legs meet together. That would detract a bit, but not much, from the floating drawer look while adding considerable strength.

[Kesh Ikuma]

@rudy - I see what you mean now. Since your photo (thanks for posting BTW) is only from the side, I'm guessing that it has a "floating" look in frontal view because the top drawer is located where the front edge of the sidewall is perpendicular.

I'm reluctant to make the side to go all the way up to the desktop, purely for the aesthetics reason (I like the open look of the current take). Also, I suspect it won't be as effective dispersing the weight as much as in your desk because the sidewall would be only attached to the front leg (the front and back legs are offset).

[Wayne Lomman]

Kesh, the joint will have tensile stress on the outside. It's the whole problem to be resolved with this design. As you would know, timber is not great in tension and the plate underneath does nothing to resist that tension. So I agree with you that it is hard to justify doing it. It's why I wanted a flat plate on the inside but it will be seen so pins are probably tour best bet. Go half inch at least. Cheers

[Kesh Ikuma]

@Wayne - Ah, sorry I misunderstood you. So, you are suggesting using a metal floating tenon instead of a metal dowel. I was so entrenched in figuring out the underneath mending plate, I misinterpreted your comment.

BTW, a mending plate does offer tensile stress relief. Not the plate itself but the screws/bolts.

Anyway, I think I'm pretty set to go with these joints:

343129343130

If I see a thick aluminum bar/plate stock at local borg, I'll consider your suggestion. (Otherwise, I have a leftover 3/8" Al rod that I can use.)

Also, just if anyone's curious, here is the current state of the leg pieces (and their templates)
343131

And the bottom side of the desktop
343133343134