Making Connections

I have a project for a west coast client to make Japanese joinery models, which I’ve been pecking away at here and there between other tasks. These models are at a large size, and it was tricky finding dry 16/4 materials appropriate to the task. I managed to obtain some Iroko and Bubiniga beams. The joints are made to have contrasting halves, and to be readily demountable. Apparently, the client, who is unknown to me, is fascinated by Japanese joinery mechanisms. We have that in common!

I thought I’d share a few pics of the two models I’ve made so far. The first is koshi-kake kama tsugi, the half-lapped gooseneck joint. This is a splicing joint often used on mudsills and purlins where they are adequately supported underneath. It is a tricky connection to fit properly as the head of the joint has sloped ramps which tighten the connection as it goes together.

The two joint halves, the Iroko on the left, and Bubinga on the right:

The joint assembles in one direction, which is a vertical drop in:

These are a couple of standard proportionings for this joint, and I went with the longer and skinnier necked version. There is also a tapered neck version, and some alternate shapes for the joint head as well, but the above is more or less standard.

As the joints slides down, it begins to tighten up:

A view from the other side:


I took these photos just prior to wrapping the joint for shipment, and did not push the joint all the way together. I had done so at the shop but omitted to take a picture. I resolved to be a little more organized in the picture taking for the next joint.

The second joint is a double wedged locking box joint, or shachi sen hako dome tsugi. This is a high class joint used for decorative alcove floor frame corners, and is seen on some temple mudsill corners. It is also used in framed furniture pieces on occasion, and it is a favorite of mine.

The two joint halves:

A view of the joint halves going together, the assembly rotated 90˚ up:


The two halves going together, conventional orientation of parts:


Completed joint, inside view:

The black locking keys, shachi sen, I made out of Gabon ebony. They have bulbous heads on them to make it a lot easier to pull the pins out. Normally, the wedging pins would not have such heads, and would be driven in with a hammer and likely trimmed flush. Once the shachi are driven in, this joint is tightened right up and the pins would be removed only by drilling them out. Finding the right position with fitting these pins so that the connection would tighten and yet not get too tight added a fair amount of time to the fitting process.

A view of the front, the intended position from which this connection is viewed:

Clicking on any of the images will render them in a larger format.

The joints are all hand planed clean, chamfered and then sent out. No finish otherwise. A very enjoyable project for me, and apparently providing much delight to the client so far. I hope the reader has enjoyed the look at these joints. I have about another dozen to make, give or take, depending. I’ll probably post some follow up entries as these connection are completed. Thanks for your visit!

Update: after discussion with my client’s representative, they would prefer no more blog entries of photographs on this topic. Sorry!

11 thoughts on “Making Connections

  1. Nice job, Chris. I'm curious what your client going to do with those? Possibly for viewing on his coffee table? Cool if it were for perspective clients to see, and then they will better understand and want to order work made accordingly, from you of course.

    For the shachi sen, bamboo makes a good choice. A tough material in thin proportions.

  2. Absolutely beautiful. I am working on a white oak lamp with vastly easier joints. This posting inspires me to make it as good as I can. Thanks.

  3. Dennis,

    as far as I understand it, the client likes these in the same way people like wooden puzzles and decorative objects. I think that maybe their kids will get to play with them as well, but I'm not sure. I am hoping the client might want to order furniture or larger structures from me at some future point, and am perfectly content to have the opportunity to make these joints too!

    Yes, bamboo would have been a good material for the shachi sen, but I had none on hand. The Ebony was a perfect choice, especially since I have some! I didn't want to use bubinga for the pins as it is slightly too sticky in surface quality for purposes of sliding in and out readily. I also considered using lignum vitae for the pins.

    Thanks for your comment.


  4. Anonymous,

    i thank you for your comment and will allow it to stand, but I do require people leave their name when commenting, even if under 'anonymous'.

    Thanks for your understanding.


  5. fascinating Chris. Is there a structural reason for using such complex joints in a mundane application (sill plate) when a simpler design would have similar or adequate strength? Or is it more motivated by a spiritual or metaphysical idea like “complexity for it's own reward?”

    Harlan Barnhart

  6. Hi Chris,

    How much of the finished joint surfaces are left straight from the saw as opposed to pared with a chisel? Do you ever undercut shoulders a hair?

    They look fantastic.


  7. Harlan,

    good to hear from you and thanks for commenting.

    Two things:

    -On a traditional timber frame structure the mudsill is hardly mundane, but rather is the piece which ties the lower ends of the posts together and distributes the wall load over the foundation. It is an important component.

    -with any join there are usually several options, from 'mild to wild' as they say, and what usually decides the matter of which joint to employ is economy, by which I mean the budget. In certain cases strength requirements or assembly particulars will eliminate certain options in joinery. As for strength, bolted connections will realize higher load bearing capacity, and are comparatively simpler to execute, however they have their drawbacks.

    It's the same here in the west in certain cases, like gypsum board finishing standards. Level 1 might be fine for an access tunnel or crawl space, while level 4 is going to please those looking for as 'perfect' a finished surface as possible. In cabinet making, one can join drawers with nails/screws, glued lap joints, dowels, biscuits, fingerjoints, dovetails, and so forth.

    Looking at the gooseneck, there are more sophisticated joints one could choose. Going one step down in complexity would be, say, a half-lapped dovetail, which is weaker to be sure. Another step down from that would be a simple half lap. Another step down is a butted joint as is seen in most Western stick framing of mud sills. Those simpler joints are quicker to make and associate to average/cheaper work. 'High class' timber connections generally conceal more of the joinery mechanism and let the planed wood be shown in all its beauty to the fullest extent.

    For the corner joint, if an all-wood joined connection is desired revealing no end grain to the exterior, and the budget is there, then the shachi-sen box joint shown above has no equal in my view. And there are more both simpler and more complex versions of that joint yet!

    Carpentry is an eminently practical art for the most part, and Japanese carpenters are no different in that regard. They like to have a beer after work sometimes too. Spiritual/metaphysical ideas do not tend to inform the strategies of Japanese carpenters so much from what I have seen.

    It is not a matter of making something complex for its own sake, but of finding good solutions to problems and learning from what hasn't worked in the past. A drive to make something better.

    Finally, complexity is relative. The tree is a highly complex system which eludes our complete understanding at this time as far as I'm aware, though wood is considered an 'elemental' or 'basic' material. Breathing is a simple in-out affair most of us give no thought to, but at a level of deeper chemistry, well, there's rather more to it…


  8. Tico,

    good to hear from you. I don't consider a sawn surface a finished surface in most cases. I generally pare the joint surfaces unless it is really something just banged together. I might saw a surface of a nearly closed joint in aid of getting a better fit, but even then I am usually approaching final fit by paring. In the case of these model joints, since the two woods are of vastly different density, sawing kerfs between the two joint halves is not a great idea.

    On the undercut question, I think the best thing for structural integrity is full bearing on end grain surfaces for those joints in which end grains surfaces are pressed together by the closing of the joint. So, end grain abutments I tend to work to 90˚, and if I undercut, only very slightly.

    For fitting of side grain portions of a joint to one another, if it is the case that the structural importance is less a factor, and it would be helpful to minimize friction in assembly, then I might undercut a little, chamfer lead edges a little, and do some grain pre-compressing here and there.

    If one suspects that seasonal wood movement might cause a portion of the joint to open up more than might be acceptable, then sometimes undercutting certain areas can minimize that tendency.

    I really don't care much for the type of joinery which is undercut so extensively that while a crisp fit might be achieved initially, at best all one is left with is an eggshell of a connection, and after several bouts of seasonal movement the joints usually look rather poor and are, as they were when first cut, of dubious structural integrity. It defeats the primary purpose of joinery, which is a structural one rather than aesthetic.


  9. Chris,

    A great response to Harlan's question. Probably an enquiry such as that is more likely when one is looking at how things are newly done, as when seeing the stacked up worked members prior to the structure raising beginning, or in this case, your constructions.

    Looking at something like Horyu-Ji temple in Nara, of which one of the oldest and largest wooden structures in the world still stands upright after multitudes of large earthquakes, might also better give a good idea of why carpenters used certain methods that might today seem rather complex.

  10. I would pay money to see a video of the making of one of these joints from start to finish! Please!

  11. Anonymous,

    I thank you for your comment and will allow it to stand, but I do require people leave their name when commenting, even if under 'anonymous'.

    Thanks for your understanding.


Leave a Reply to Anonymous Cancel reply