Gateway (27)

It’s striking to me, on the one hand,  how there exists the basic and fundamental Japanese technique of kerfing a timber, sewari, a method used to control the tendency for the material to check when drying, and on the other hand, how largely unknown this practice is in the west. Virtually every woodworker who happens by my space immediately notices the kerfs in the timber and expresses worry that something has gone drastically wrong.

“Yep, I put those kerfs in there. The sticks came from the mill without those cuts.”

The usual look at this point is a sort of bewilderment. Then I explain the rationale, that the kerfs have widened out as a result of the timber drying, and the usual head nodding ensues, or, as often, more blank staring.  The information is taken in, but not quite digested somehow. I find this response surprising.

How can something so basic and fundamental – and easy to do – seem to be so completely out of the range of Western thought and practice? Westerners know technically/scientifically what happens to wood as it loses moisture, and that larger pieces can crack and check, and we used to kerf floorboards and door sills to mitigate the same issue, but the idea never seems to have gone any further. It’s a curious thing. I don’t know the reasons, but I have noticed that sewari seems outside the scope of general understanding. Well, dammit, hopefully by showing in more detail how the process works I might spread the good word a little further. Yes, it’s time to get evangelical.

Well, maybe not so much.

Anyway, the past couple of days have been involving work to patch those kerfs. A patch, inlay, or more specifically a shim in this case, is termed umeki in Japanese.

I try to match the grain of the umeki as closely as I can to the piece into which it will be fitted. Depending upon one’s luck and fussiness, the result can be reasonably seamless – or not. We do the best we can. Matching an area with flatsawn grain, as in the following example, means that the umeki must be quartersawn stock and therefore the viewed surface of the shim, also flat grain, can be quite difficult to match:

I achieved a decent result with this one I think:

Of course, a few weeks in the sun and the subtleties of the grain pattern will become muted.

A closer look at the end grain shows the configuration of the umeki more clearly:

Here’s the same process underway with the other nose piece:

The stub tenons and end grain abutments of the piece are still untrimmed.

My shop is unheated, however the washroom has a small space heater to keep the pipes from freezing, and that makes it a good place to do a glue-up:

Since the last blog entry, I have moved the nose piece tenons along:

Another view:

The next task was to complete some dimensioning work on those nose pieces – the sticks are about 1/16″ (1.6mm) too tall. I elected to do some rough material removal with my portable power plane:

Clean up afterwards by hand:

Checking that the result is square:

More umeki placements on the main posts:

The main posts are never going to make their way into the washroom, so I hit upon an idea. I purchased an electric blanket to place over top of the glue-up to keep the pieces warm:

I use epoxy for this process, which will cure fine down to 20˚F (-6˚C), however the cure time is then extended to a full week, and I prefer to move things along a bit faster than that if I can help it. The electric blanket seems to be working well, and is long enough to make it possible to effectively glue in longer umeki on the other post faces.

Here, I’ve cleaned out another kerf in preparation for patching:


The umeki are made about 0.1mm fat in thickness and then eased slightly on the insertion side with a plane. They are then driven in after glue has been slathered into the slot.

I noticed that one of the nose pieces has a slightly curled down tenon, so I corrected the surface with a plane until it was decently flat:


The opposing edge of the tenon has yet to be trimmed to the line.

Here, on this nose piece, the kerf can not so readily be dressed clean by routing, as the kerf moves along a bowed line:

These nose pieces came out of the large beam which bowed during drying, which explains how, after the S4S machining, we end up with a curve in the kerf.

In this case, I make a slim umeki that is tapered to fit as best I can and judged that the fit should be okay as the kerf was in decent shape:

That piece was also dragged over to the washroom for gluing. It will spend the night there, warm and cozy.

On to working on the kabuki end tenons, which are pretty dang long:

After the rip and crosscut, a little handsaw action to separate the waste:

Well, at 4:00 pm today my shop was at 3˚C (37.4˚F), which was starting to feel a bit cold. Tonight the forecast is calling for extreme cold, down to -22˚C (-8˚F) with windchill taking things down to scary numbers I’d rather not mention. Due to the charms of the polar vortex, I’ve been bringing chisels and planes home to sharpen, and plan to stay at home for most of tomorrow sharpening. That’s how I started today. I might pop to the shop in the middle of the day for a couple of hours though. We’ll see how it goes.

All for today – thanks for your visit. Post 28 anyone?

5 thoughts on “Gateway (27)

  1. Chris,
    Twenty years ago, I worked for a firm doing structural reinforcement of wooden structure by epoxy injection, we would close the room around the beam being injected, put an electric heater inside and let it cure for a week minimum. If you're interested by the technic, I can scan a few pictures.

  2. how deep is the kerf supposed to extend into the timber? I saw another guy kerfing down to the heart, which seems too far, but I dont know much about Sewari. Your kerfs seems to go not to the heart? JD

  3. JD,

    well, look at it this way:

    -for controlling checking, the greatest movement is on the outer portion of the tree, so kerfs to relieve drying stresses need not run especially deep
    -for the purposes of drying the timber more evenly, the kerf allows for moisture relief from inside the log/cant, so the kerf ought to run deeper on that account.

    As a general rule perhaps, if kerfing to control both checking and drying, the kerf should run in 2/3~3/4 of the way to the pith. If kerfing already dry timber to control checking in service, as with flooring, then the kerf can be relatively shallow.


Anything to add?