A Ming era table with Chinese and Japanese joinery

Gary.......welcome to WC.......that is a very impressive piece of woodworking.......I'm certainly looking forward to the joinery explanations and illustrations that allowed you to make this beautiful table without glue.......masterful.

Welcome aboard! Looking forward to your posts about joinery. A tall table like that seems likely to be subject to racking. Joinery that will stay tight without  fasteners will be interesting.

I burned all of my Madrone, two trees that dried well while standing. It's the highest BTU producer in our region. Too bad "ornamental" fireplaces have the poorest efficiency. A candle provides better warmth.

I would love to see how Madrone works out for furniture. I see Offerman, of Parks fame, works it and he admits it's detraction as a mover. I am earmarking the continuation of your commentary!

Gary,  I really like this table.  In particular, I like the top.  Thick, little overhang, and the space between it and the rest of the table.  Nice touch.  Can't say enough.

Thanks for the welcome, folks. It's nice to be here.

The inspiration for this table and its joinery was a series of posts by the late Chris Hall at his blog, The Carpentry Way. After reading through his build of an exquisite Ming style dining table I knew I wanted to try something similar some day.

The basis for Hall's table (and mine) is this Ming table:

and its joinery:


What looks like a solid top is actually a frame and panel. The wide frame has mitered corners held together by inner dovetailed tenons. The thin panel sits flush and tight with the short arms of the frame and is captured in a groove there. The panel sits on top of the long arms and is held down flat by a series of sliding dovetailed battens and dovetails on the frame arms.

My table top is similar except for the corner joinery and that I used two panels separated by a narrow expansion gap. The corner joint is a Japanese locked miter. Inserting the tapered parallelogram shaped wedge (called a shachi sen) is what locks the joint. Here is what a half-joint looks like. The opposite side is identical but inverted.



And here was most of the frame before shaping the dovetails hold downs for the panels to match the battens. You can see the tiny mortises in the corners for the shachi sen.

And here are all the parts during assembly:


More to come.

Added later 33 min 25 s:

I wanted to add more about the panel(s). The original Ming designs all used a single wide panel that is mitered to fit tightly to the short arms of the frame. This would be a big no-no in Western woodworking since seasonal movement would either crack the panel or open the miter joint. I puzzled over this for a long time. How could these table tops not crack or blow up? I've never found a good answer. It may be that the woods they used (mostly rosewoods now long unobtainable) were unusually stable. Maybe humidity changes were less in China before widespread central heating. Maybe many tables did blow up and we are seeing only the survivors in museums under controlled environments.

In any case, both Hall and later I did not want to risk disaster so we both included a small gap in the center between two panels. Any seasonal changes occur there. In my case, the half panel is only 7 inches wide. I left a gap of a thin 1/8 at assembly in January, middle of our wet season where I live in Oregon and where indoor humidity is a fairly constant 60-65%. In the summer indoors the humidity is a fairly constant 40%. The table has now been through two winters and most of two summers and the gap has changed only slightly during that time, probably perceptible if I put a caliper on it but otherwise not.

That also speaks to Bruce's question about madrone being squirrely. I have read and heard stories locally of the same thing. The stock I used was vacuum kiln dried and has been rock solid. I suspect the deal is that it moves a lot during drying but once dry is OK. And maybe vacuum drying is the way to go if you have access to it.

I will also say that the madrone I used, while very expensive, worked absolutely beautifully with hand tools and was a joy for the kind of joinery and beading I was after.

Added later 12 h 50 min 37 s:

Leg to apron joinery. I used a Chinese triple mitered joint. There are many different versions of this, some simpler and some more complex. I used a complex one that incorporates a tapered dovetail to resist tension.  The stub tenons on the top of the leg insert into mortises in the bottom of the top frame.




Here is what they looked like while making them and before shaping the aprons or legs.




As you can see, the aprons insert into the joint from above and are locked in place by the internal tapered dovetail. The apron tenon inserts into a partially open mortise. That inhibits the apron from twisting. A locking mechanism that I'll show in  the next post holds the top onto the legs and prevents the apron from lifting up out of the joint.

Hi Gary

That is absolutely first class work! I love the joinery, the precision, and the results. Altogether spectacular work.

Welcome to WoodCentral. I hope we see a lot more of you here.

Regards from Perth

Derek

Very nice work, Gary. I've done a couple versions of that tri-mitered joint over the years, but none as sophisticated as yours. Impeccable workmanship.

On another note, what is the purpose of the housings you've chiseled on the inside top corners of the legs? 

Thanks for posting, and welcome...

Ellis Walentine, Host

Thank you, Derek. From what I've seen of your work I thought you might like something like this.

Ellis, good eye. That housing is for the next joinery, which is my favorite in this piece. I stole the idea from Chris Hall.

In Ming versions of this table the top is held on by a bit of glue (fish glue, or sometimes just lacquer) on those little stub tenons before they insert into their mortises on the bottom of the top. Gravity does most of the work keeping the top on and the stub tenons register the top relative to the legs and aprons. The glue keeps things together if the table is lifted by its top or by the aprons.

In Western furniture the top might be attached by clips to the aprons. But since the aprons are not restrained vertically in that tri-mitered joint, clips won't work.

Hall had the insight and imagination to borrow from Japanese timber framing and use a yatoi hozo (雇ほぞ), which is a floating tenon secured by shachi sen and which have a dovetail on one end. These are much stronger than a dab of glue on a stub tenon and can be disassembled fairly easily for any future repairs. For my table, they look like this:



And they fit into that housing on the inner corner of each leg like so. The gap at the bottom of the housing is important to allow the yatoi hozo to pull down tightly when the shachi sen are tapped in.


The dovetailed end is captured by two half-mortises on the undersurface of the top frame.

And then the frame is assembled and the corner joint key is tapped home to lock the yatoi hozo in place.  Here for the test fit I had left the locking key way long to make it easier to remove.



There is a lot going on in those legs!


When everything is assembled and in place on the base frame it looks like this. When I did the final assembly 18 months ago I left all of the shachi sen a little long, just enough so I could tap them in further or pull them out if necessary. Since they aren't visible except to folks who care enough to look, I think I will leave them this way rather than cut them off flush.




The last bit of joinery is that for stretchers. I'll do that next post.

Added later 3 h 11 s:

Before getting to the stretchers here is one more photo of the top during assembly. All the corner joints are closed. All that remained was tapping in the square pegs (komisen in Japanese) for the draw bored battens.



Funny story: after finishing the pegs I was of course eager to set the top onto the lower frame. I carefully wiggled all four yatoi hozo into their housings on the legs, gently tapped in their shachi sen, and carried the completed piece into the house to proudly present it to my wife. She came out and oohed and ahhed as I hoped. As we were admiring it I started to realize that something was wrong but I wasn't sure why. After looking a few more minutes I realized that I had put the top on backwards! There are some minor defects and color mismatch on one side of the top and bottom that I intended to both be in the back). No worries. Back to the shop, remove the pins, reverse the top, reinsert the pins.

Try that after a glue-up!

The stretcher joinery is one common to Chinese and some Japanese pieces. It doesn't have a strict name or translation to English but I call it a "spear point" tenon. I used the one on the right.



This joint has several structural features that recommend it. One is that it is an almost a double tenon with lots of glue surface. It can also be pegged but I chose not to because I didn't want exposed joinery. The 45 degree mitered/mitred abutments provide a little more racking resistance than 90 degree shoulders of similar dimension. One other feature is that the mitered abutment allows one to carry a chamfer or beading around a corner.



The downside is that it takes patience to get them fit. Accurate layout is essential and caution with the last few paring cuts. Practice helps a lot. And sharp paring chisels. And did I mention patience?

Here is the start before fitting.





Almost there. Pause, take a breath. Should the next paring cut be on the tenon or the mortise/mortice?



Checking all four stretchers together. You can see that they can't be trimmed either too long or too short without changing the fit of the aprons to the legs which would change the fit of the top to the aprons.



That's it for the joinery. If it isn't too much, I can show how I shaped the legs and aprons and frame, and added the beading.

Gary - The only part of your post that is familiar to me is when you put the top on backwards. And of course, I only find things like that after they are glued in place. My mind can't comprehend the patience it must have taken to build that table. Trying to visualize and layout joinery like that would drive me to drink.

Pat, I won't minimize the work involved, but broken into smaller chunks over a period of months it wasn't so bad. I think I had about 500-600 hours in the piece, including research, design and finishing time. That sounds like a lot but is based on about five hours per day over four months. And it was a lot of fun. 

I also didn't show the SketchUp models, several mockups, practice joints, do-overs, and artisanal firewood created. I made only one serious layout mistake that required entirely re-making two aprons. I have the luxury of not being under time or economic pressure. I was, though, under some pressure to get it right because of the expense of the wood. As my wife helpfully said when I told her how much it would cost, "Well, OK. But don't (mess) it up!"