Kanawa-tsugi Joinery Models Part 1

---

March 2024 has proven to be a jammed packed month. After a somewhat slow start to the year a few different projects and craft fairs all came together this month, making for an intense period of work. One interesting project that has been occupying my time the past couple of weeks are some kanawa-tsugi (金輪継ぎ) joinery models. These models, 50 in total, are destined for the Huntington Library where I had the fortune of doing some work with the Somakosha crew the past couple of years.

The joints are all made from cherry which I milled in to 40x40x175 mm blanks (approx. 1 9/16″ x 1 9/16″ x 6 7/8″). Each blank makes up one half of the joint, so for 50 joints total, I made 100 blanks, plus a few extras.

Each half of the joint is identical in form, so I performed all the same layout and cutting steps on each blank. For reference sake, here’s a model drawn in Blender of one half of the joint.

Typically the outline and cut lines of kanawa-tsugi joints would be fully drawn out, but given the quantity to save time I kept layout marks to a minimum, and instead decided to rely on jigs and machine setups to make repeat cuts at specific dimensions. There were a few areas where layout lines were necessary though, so for those areas I used a set of ball pen kebiki.

Here are the layout lines for the end stub tenons, which are the first area to tackle when cutting these joints.

The orange highlighted area below shows the stub tenon cutout area on the 3D model.

For roughing out the stub tenons I modified an existing router jig, which allowed me to hold each piece vertically for routing. I like this kind of setup for cutting tenons. It’s accurate and safe.

Each piece was clamped in the jig using a toggle clamp and set of stops.

Then from above it was a simple matter of setting the correct depth on my router and carefully routing out the waste.

Here are the completed blanks with stub tenons roughed out.

I routed as close to the line as I felt comfortable. Ultimately this will be one area of the joint that requires a bit of chisel work to square up the corners of the stub tenons along with some final paring to cut right down to the ink line.

Next I made a couple of cuts at the sliding table saw to establish the inner shoulder and wedge shoulder.

Here are the locations of the table saw cuts shown on the 3D model:

Then it was over to the bandsaw to rough out the bulk of the joint. At the point it starts looking a little more like a kanawa-tsugi joint.

One more detail to take care of was some drilling to rough out the waste for the inner grooves that will accept the stub tenons cut earlier. I probably should have done the drilling before the table saw cuts and before bandsawing out the bulk of the joint, to prevent the drill from drifting due to lack of support on the exposed side of the shoulder. But this way worked out nonetheless. The drill drifted in a few spots but nothing too serious. I also thought about making these cuts with a hollow chisel mortiser but in the end I decided that drilling would be less frustrating and risky, given the quality of the hollow chisel mortiser currently available in my shop.

The next phase of work was the part that I was most concerned about, cutting the long sloping faces of the joint.

On typical timber frame scale kanawa-tsugi joints, these faces are usually roughed out with a hand saw or circular saw, followed by careful chisel work and planing. That much hand work on 100 pieces seemed pretty daunting so I worked up a router jig to make easier work of leveling out the sloping faces.

Here’s the piece clamped in the jig. The mdf bed that the joint half is sitting on is angled such that the top edge of the jig and sloping faces of the joint become parallel. That gave me the ability to run the router on the upper edges of the jig to rout a clean surface at the correct slope.

After a couple test cuts to dial in the correct depth on the deeper of the two sloping cuts, I set the router depth stop and locked it down tight. To deal with the offset upper sloping face I used a 5mm piece of brass as a reference to reset the depth.

So each joint involved clamping the joint in the jig, plunging the router down to rout the deeper sloping face, followed by resetting the router depth using the 5mm brass reference and cutting of the upper sloping face. Kind of tedious but all the cuts came out really cleanly and consistently.

And that was it for the machine work for these joinery models. From here on out it was all chisels and planes.

I started by cleaning up the inner grooves to recieve the stub tenons.

This skinny little detail chisel is perfect for cutting in narrow spaces, and made it easy to pare the side walls of the grooves.

I made good use of a simple bench hook for the hand tool work. I haven’t regularly used a workbench with vice for probably close to 10 years. Occassionally I need to clamp down oddly shaped or large pieces for planing or chisel work, but most of the time I prefer to use simple stops for holding work. I find it gives me a lot more freedom of movement and makes the overall process more fluid.

A little vertical board, which I screwed into my bench hook, came in especially handy for use as a stop for paring the end grain portion of the inner grooves with the joint standing in a vertical orientation. When you’re doing lots of hand tool work, especially for days on end, you’ve got to find a way to get comfortable, and with this bench hook setup I was able to comfortably sit and do all the chiselling and detail work needed.

One other area that needed a touch of cleanup with the chisel was the wedge shoulder. First chop…

… then slice out the waste.

And lastly was some cleanup on the stub tenon ends.

Once again I used chisels to shave right down to my layout lines and clean up the corners.

And then it was the moment of truth, bringing two halves together…

After an initial test fit each joint required a touch more fine tuning to fully close, with some joints requiring more work than others. I did all the fine tuning on the stub tenon and groove ends. The really tricky part was making adjustments such that the joints were as clean and gap free as possible while making the joint easy to assemble and dissassemble. Determing where a joint is binding up, and where wood needs to be removed is a tricky task.

Here’s one joint fully closed after a bit of adjustment.

… and several days later the all 50 kanawa-tsugi joints were adjusted and fit!

Some people may find this kind of work tedious but I really get into it. I love benchwork and hand tool work. I also think mini-production runs like this are great because you have the ability to really hone the process and your technique. Usually by the end you’re way better than when you started.

Well that’s all for now. Next week I’ll share the final details which includes finish planing, making wedges, dying and finishing.

~Jon