Gateway (40)

Post 40 in an ongoing series describing the design and construction of a kabukimon, a type of Japanese gate. This is a project for the Museum of Fine Arts in Boston. Post 1 in this series can be found here if you’d like to start at the beginning. Each post links to the next at the bottom of the page.


Today marks the 800th post on the Carpentry Way.


Another day, another day in the shop. Today’s work started off with finishing up the joinery on the ends of the magusa, a pair of ‘T’-shaped tenons. Here’s one end:

And the other:

The magusa is upside down from its actual orientation when installed.

Then I moved on to the hikae-bashira – the rear support posts. These are connected to the main posts with a pair of nuki, and are at a wider post spacing than the main posts:

The point of making the rear posts wider is that with the gate doors open, there is a larger viewing window of the garden as one approaches. Since the gate doors are open during the museum’s operating hours, the employment of wider rear support posts makes sense. If the normal condition of the gate were to have its doors in the closed position, and/or there was no special view revealed when the doors were both fully open, then employing rear posts with the same post spacing as the main post would be logical.

As the opening is effectively flared, there is the matter of dealing with the connections between the rear posts and the nuki. There are two common ways to do this. One method is to rotate the rear posts so they are normal (orthogonal) to the stretchers. This allows the nuki to pierce through the rear posts centered on the front and rear post faces. The drawback is that by rotating the post it is out of plane with the front part of the gate, main posts, etc., so that difference in orientation does create a certain amount of visual discord.

The other common approach is to keep the rear posts orthogonal to the front posts, etc., and have the nuki penetrate them without being centered on one of the faces, or indeed, being off of center on both entry and exit faces. That’s how it was done in the old gate. The rear post’s orientation makes it fit better with the rest of the structure, however the off-center mortises for the nuki create a less polished appearance.

My solution to this is to make the post parallelogram in section. This way its side faces are orthogonal to the nuki, and its front/rear faces are orthogonal to the main posts, etc., and thereby the nuki penetrate the post faces centered on entry and exit. This presents the most seamless look, and I expect most people won’t notice it at all.

Making the rear posts into parallelogram section sticks involved doing some end cuts, cleaning up the end grain, choosing the orientation and position of each post, and then marking the parallelogram section on the end grain. From there I sliced a piece of stock in the bandsaw with the table tilted to create a tapered piece which would support the rear posts as they went through my planer. A few additional pieces of plywood were used to guide the post through.

Here’s a video I took today showing the process from there on out – planing several passes to create the first sloped face on the sticks, then making some adjustments to the surface with a hand plane to get the angle dead-on, then a final series of passes through the planer with the angling shim removed to plane the opposing face parallel to the first:

Fortunately, the pieces came out exactly as I wanted them, are dead nuts on the desired angle, and are 0.01″ over dimension in both axes to allow for a final round of hand planing. I also allowed for 6″ excess length, so I can trim off any snipe which occurred. Next time in the shop I will lay out and cut the mortises in the rear posts for the nuki and for the threaded rod’s pocket access to put the nut on the rod.

In other news, the custom-made bronze decorative hinge leafs, or hassō kanagu, arrived from Japan today – they’re beautiful!:


The back sides of the two pieces:

All for today my friends. Thanks for coming by the Carpentry Way. Post 41 is up next.

5 Replies to “Gateway (40)”

  1. Happy 800th anniversary.

    PS you anti spam cannonade always makes me smile. “Is this how you want to spend your time on earth?” Absolutely brilliant.

  2. Chris,

    I'll second Jonas' motion: Happy 800th!

    While I'm at it, I have questions about the planer shim setup you used on the rear posts. I'm used to planer support jigs that amount to a tray that travels with the workpiece through the planer. IIRC, you described one such back in the halcyon days of the “Irregular Situation” series (ah, here, in post III).

    Here it looks like the support shim and plywood braces remain stationary at the infeed side of the planer, it appears by using cleats that hook over the infeed table. Is that correct? Here's where things get fuzzy for me: Is the workpiece supported on the outfeed side? The beam appears to feed smoothly, but it seems that a stationary platform like this would prevent the lower rollers from engaging. Could you elaborate a bit on this setup?


  3. John,

    thanks for the comment and questions.

    Yes, you are correct in surmising that the support shim and plywood spacer remain stationary and use a cleat on the inboard end to keep them from sliding. The support shim runs the full length of the planer table so it supports the stock as well as the regular table can. My planer has double outfeed rollers which also improve stock feeding and support.

    I never use the lower rollers (mounted in the table) on my planer. They are intended for use on roughsawn and wet stock, and I rarely if ever feed such material through the machine. The rollers are at their lowest position, which means they spin when the stock moves over them but they are not actually carrying the stock. I find the rollers when raised also contribute significantly to snipe, so they're not something I make use of. Stock feeds better on the rollers, however regular table waxings do the job in that regard.

    So the stock is simply fed over the shim. The tilt of the shim tends to encourage the stock to move downhill, and that is where the plywood spacer comes in. The edge of the table has a raised lip against which the plywood abuts. The other piece of plywood keeps the shim from squirming outward under pressure, and I used a piece of double stick tape under the far end of the shim, 'just in case'. I physically took care of the rest of the outfeed support. Simple, and it seemed effective.

    The set up was accurate to +/- 1/32″ at worse, and any inconsistency was easy to take care of by hand planing. The shim was simply bandsawn, and I counted on the hand plane work to deal with the outcome from any roughness or irregularities in the beam surface.

    Make sense?


Anything to add?

error: Content is protected !!
%d bloggers like this: