Archive for November, 2009

Ranma III

OK – so here’s the final installment of the diamond asa no ha pattern exercise. The asa no ha, or hemp leaf, is a traditional japanese pattern that appears in many, if not all, of the culture’s traditional arts. In shoji work, the pattern can be fit to most of the kumiko structures.  For the diamond structure of this piece, the pattern consists of three legs, each of which have 60-degree spear beveled ends to fit into the corners of each triangular section.
In order to form these ends, I use a shooting jig with a 30-degree angle at the end:


The first piece in the pattern is a long section, with each end spear-pointed at a 60-degree included angle.  To form the ends, you simply shoot a 30-degree angle, then flip the piece and shoot 30-degrees on the opposite face; the result is a perfect 60-degree spear bevel.

This long piece is then split dead-center to form two of the pattern’s legs – but a very small (under 1/32″) bit of wood is left intact, forming a ‘hinge’ to keep the two sections connected.  To accomplish this, I attach a pair of depth stops to my dozuki – set with the saw resting on three sheets of drawing paper to leave just the right amount of depth below the cut.

Here is the resulting piece

The piece is then gently folded along its seam to fit in the kumiko structure; a small bit of water (saliva also works, I hear) on the rear of the hinge makes the fold a bit easier.

Now it’s inserted into the structure.  The third leg is formed of a simple straight section, again with 60-degree spear bevels formed on each end.  This section keys into the hinged area, and locks the entire structure very securely in place.
Now, repeat the process for every section of the structure that you want the pattern in.  I’ve opted for an arrangement of hexagons, in two rows. Then I trimmed the entire structure, and mounted it in a double frame – the inner frame of matching eastern white pine, and the outer frame of mahogany joined with the mitered mortise and tenon joinery I showed at the beginning of the project.
And that’s it – the piece is finally complete.  The kumiko in shoji is never finished – which would be very difficult, but which would also diminish the natural quality of the wood.  The framework is sometimes finished, especially in the west, but I like the muted colors of the mahogany and pine so much I’ve left both with a kanna finish – right off the plane.
Here is the final piece, which I’ll probably hang on a wall of my home, or perhaps give as a christmas present (to someone who isn’t reading this blog).
By the standards of professional door makers, there are some undesirable aspects to this piece.  Firstly, the eastern white pine I’ve used here isn’t nearly consistent enough for fine work – the variations make the intersections show up as individual pieces, rather than blending into a single mass.  Also, by the extremely fine standards of craftsmanship of the shokunin, the joinery in my kumiko would be considered sloppy.  But my goal with this exercise was to set up and familiarize myself with the appliances and fixtures that are necessary for this work, and to get used to the process so that I can employ it in screen and furniture work in the future — and from that standpoint, I’m very pleased with the outcome.
Like much of japanese woodworking, shoji can be a very painstaking and detailed process — especially with some of the more complicated structures like this one — but the end result is
very beautiful, and it is an excellent exercise in close-tolerance handtool work.  Make ten of these, and I can guarantee that your skill with a saw will improve a pretty fair bit.
Next it’s back to some plane work…

Kumiko – I just can’t quit you!

Before I get back to the planes I’m trying to finish up, I want to go on to a bit more of the Ranma project I’ve been working on this week in between coats of french polish.  The traditional square network of kumiko in shoji are relatively straightforward, with lap joints at each intersection.  The triangular, or diamond pattern, however, is a fair bit tricker as each node of the lattice involves three members to join.  I don’t know of any written information in english on the techniques for this — though there are some sketches in Nakashima’s Soul of a Tree. I pestered a couple of expert shojimakers in the US while puzzling out the details, and I thought I’d put up a not-so-brief summary of the basic techniques and joinery.

I’m going to try to document this as well as I can without taking 2 hours of your time, but please be forewarned that it may be a bit confusing at first.  It may help to have a look at the final piece first, to get a sense of what we’re working toward.  Here is a roughly 24″ x 7″ section of diamond-pattern kumiko that I’ll be using in the final ranma:

In order to make the ‘grid’ work, there are three different members of the lap joint that have to be employed – one for the horizontal pieces, one for the ‘left leaning’ pieces, and one for the ‘right leaning’ pieces.  Here are my test sample pieces:

All of these joints are cut at 60-degree angles, and 2/3 of the depth of each piece is removed.  Exactly how to remove them for a good fit is the tricky part.  In order to make these cuts (I’m cutting at least 60 of each for this piece) I use a purpose-built small miter box with a kerf exactly sized to my dozuki.  I also use small guides clamped to the sawplate to precisely gauge depth of each cut.
For the first piece, you cut a left 60-degree lap (all laps are 3mm in width to accomodate the 3mm kumiko) a full 2/3 of the depth of the kumiko. Then make a right 60-degree lap that exactly overlays the first, and to the same depth… gently remove the waste with your narrowest chisel. Here’s what that joint should end up looking like:
The second joint, gets a left 60-deg lap, but to just 1/3 depth.  Then you flip the piece top-for-bottom and cut the same joint from the other side.  Again – these need to exactly overlap.  The dead center point of each joint will be the same.  Here is what it should look like at the end – notice that although you cut both laps in the same direction, the ‘flip’ means that they’re actually opposed in the final piece.

The third cut is even more interesting.  Start as with the first, with a left 60-deg  lap to 2/3 depth.  Then make a right 60-degree lap, but this time go to only 1/3 depth.  When you remove the waste, you should have the following:
Now, it’s time to gently fit these and see how accurate we’ve been. Start with joint 3 (the odd one) on the bottom, and insert the joint 2 gently in place as so:

If things are fairly clean, you should be rewarded with the following:

If so, you can feel pleased – things will almost certainly work out just dandy. Joint number 1 should fit ever so gently into that slot, and leave you with a perfect, or close enough, three-way lap joint.

Now – if you got this far on your first try – start a blog.  You’re a woodworking hero.  Most of us need to make a few stabs at it before we get everything reasonably ‘right’ for this.  Don’t fret – it will all make sense after the first couple of attempts.  After that, you just cut row upon row of each type, at exact intervals (I’m using 2″ on center for each joint) and you should hopefully end up with something like the ‘sheet’ at the beginning of the entry.
A couple of things to note:  first – you can gang-cut the pieces, but you have to pay attention to the offset of each piece in the stack.  Once you’ve made a couple of successful test joints, you’ll undoubtedly have no problems figuring that out if you remember to compensate for it.

Second – these joints are really made to be assembled ONCE and once only.  Assembling them compresses the fibers at each joint, giving you a fairly nice fit, but they will never go together that cleanly again if you take them apart.  If you look at that final test joint photo, you can see what I mean.  That piece was put together and taken apart twice before this picture, and you can see on the left and right edges of the joint where the piece has been somewhat crushed, making it feel ‘not so fresh’.

OK – so that’s it for the basic diamond pattern.  The ‘infill’ patterns to make the hemp leaf (as no ha) design will have to wait a few days at least. The french polish on my planes is just about ready, which means I’;m going to be peining and making a lever cap in the next few days. 

 I want to pause briefly here to let both of the readers of this blog know that no matter how poor the title of this entry was, you have no idea how much worse it almost was. Now normally I’m a lover of the finer things in wordplay as well as tools, but there are some times that even the most pathetic specimens of wordplay can worm their way into  even the most refined bloggibitionists’ prose. Especially in the title…

Let’s be honest, here — it just ain’t in a man’s nature to gracefully decline the savage temptation of titling a shoji blog entry “Lattice Begin” — to say nothing of its flashier, drunker sister “Lattice pray.” Sure it’s easy to toss “Kumiko Chameleon” onto the scrap heap of really really bad ideas – but who among us wouldn’t be drawn to the trailer trash siren song that is “Kumiko – you get there faster if you take it slow”?  (And to think – Mike Love was the SANE one).  Even strong men must at times swoon ‘neath such strains…

But fortunately, I didn’t foist any of those monstrous succubi onto you, gentle reader.  You may now breathe a sigh of relief.

Tooling for Shoji

I’m waiting for the finish to be ready on a couple of planes I’m working on, so I’ve been planning out a shoji exercise I’ve wanted to do for some time.  I’m going to do a Ranma (transom) piece with a diamond or hexagonal structure and asa no ha (hemp leaf) details.  You can see a fantastic example of the pattern in this screen by John Reed Fox, one of my favorite furniture makers.  This is extremely detailed work, and requires a very high degree of precision to execute well.  I’ve done a couple of basic shoji in the past, but this will be my first attempt at this complex pattern. There are a couple of critical handtool fixtures for this project that I wanted to show here.

Kumiko Thicknessing board and Kanna
One of the more critical parts of shoji work is making certain that you have extremely consistent thickness on all of the kumiko (the softwood lattice pieces in the interior).  Any variations greater than a couple of thou are going to show up very clearly as a warp in the structure, or a very sloppy joint.  So for the final thicknessing and finish planing of the kumiko, I’ve made a special plane  known as a Hikouki Kanna, as well as a board to run it on.  Here is the kanna itself:
It’s basically a typical kanna, but with a pair of ‘runners’ on either side, planed to an extremely precise depth that defines the final thickness of your kumiko.  For this project, I’m going to be using two different kumiko thicknesses, so I have two pairs of runners prepared – one for 3mm (~1/8″), and the other for 4.5mm (~3/16″).  It’s a fairly straightforward concept, and is remarkably effective.
There is one ‘problem’ though, which shows up when trying to plane such thin material:  the kumiko  bend rather easily, and can be tricky to keep flat on the board when you approach final thickness. The solution is to build in a gentle ‘hold down’ device to get them flat to the board before reaching the mouth and blade.  Here is the bottom of the kanna:
If you look just in front of the blade, you can see the mechanism – it’s a springloaded half-cylinder captured into the sole. The number and strength of springs can be adjusted to ensure the right amount of pressure.
The other part of this design is a dedicated ‘board’ that works very similarly to a sticking board.  It consists of a lower ramp that the plane’s runners ride on, with a raised portion in the center.  The kumiko are placed on the raised portion, which has screws at the end to serve as stops for the kumiko.
This ramp needs to be extremely flat, and I’ve used Quartersawn wood to help make sure it stays that way.  In my setup, the central platform is raised 9mm (~3/8″), and my two sets of runners are 12mm and 13.5 mm thick – giving me kumiko of exactly the thicknesses I need.
Here’s the concept in action, which should make everything rather clear.

The Miter Jack

For the frame on the ranma, I’m going to use a mitered housed mortise and tenon joint that Toshio Odate describes in Making Shoji. It’s a tricky joint, and the miter itself is particularly critical for a clean fit – I’m using my miter jack.

I think shooting boards are in pretty widespread use by handtool woodworkers, but I’m not sure the miter jack is nearly as common — which I think is a shame.  This is a fantastically powerful tool for any sort of joint that involves a miter.  This is my second  jack – I needed to remake it to work with my new Benchcrafted tail vise.  There is a reprint from some fine plans for a jack at the most thoroughly informative website of Alice Frampton (ALF) in the UK:  Cornish Workshop.

The miter jack can be used with a plane for a variety of mitered joints, much like the ‘donkey’s ear’ shooting board, which I think a fair number of folks are familiar with.  Where the jack surpasses the donkey’s ear, however, is more complicated (and structurally sound) mitered joints – like this one, or the secret mitered dovetail. The two wedges of the jack are perfectly mated, and tuned to a dead-accurate 45-degree angle.  The workpiece is clamped between these jaws, and the ramp surfaces act as a jig for the miter work.  In this case, I’m using a paring chisel to define the miter:

The prepared  rail, ready to be mortised:

I’m reasonably adept with a saw, but I’m not capable of getting anywhere near this sort of surface any other way.  Here’s the joint, ready for assembly:

and the end result:

In short, it’s my opinion that the miter jack is a nearly indispensible tool for a handtool woodworker.  There are other ways to accomplish the tasks it’s useful for, but none of them do the job as efficiently, or easily, as the jack does.  if you don’t one, I highly recommend it as a shop project in the near future.
UPDATE – here’s a shot of the bottom of my miter jack, showing the modifications I’ve made to the traditional mounting block.  Typically, these were made to work best in the opening of a traditional tail vise, where the jack can be pivoted up and down for access.  This design wouldn’t work with the wagon vise style, which has no open clamping section.
For this one, I sized the mounting block to fit in the benchcrafted tailvise, and mounted it on a stationary platform that raises the jack off the benchtop about an inch.  This was necessary to provide clearance for the screw handle, and it also allows me to mount the block in the center of the jack rather than the front or back rail.
The one sacrifice with this design is that it’s not really possible to pivot the jack front to back as you can with the ‘traditional’ tailvise design.  To be honest, this is the one concern I had to switching to a wagon vise style – it’s about the only time I really used the open jaws on my old tail vise.  For the most part, though, this will serve just fine, and in the cases where I really want to rotate the jack, I will mount it in my patternmaker’s vise, which gives it more positioning flexibility than any other system I can think of.

Infills, Kanna, and slouching toward Chimera

I suppose it goes without saying that I’m fond of handplanes. My wife would respond to that statement with the same shoulder-shrug and eyebrow pulse she gives when I wonder how it is our toddler twins still prefer used toilet paper rolls to actual toys.

“What do you expect,” she’s saying. “They’re barely one. They also eat lint.

Now I’ve tried to include her in the fun. I bring the occasional nifty wood shaving device to the dinner table to show her how cool it is. I invite her to come down and turn perfectly good bits of wood into fluff with me. She is ever-so-kind when she declines. “Mmm… I think I’m going to clean the bathroom” she says. Without a hint of irony. (I married very very well.)

Not only that, but she refuses to refer to my sanctuary as I have asked. I call it ‘the lab’. She calls it ‘the shop’ or, even more horrifying, ‘the garage’. Ugh. It’s not a garage. It hasn’t had a car in it in most of a decade. And a shop is where little old ladies fawn over doilies in clouds of musty perfume. A Lab is where serious men create awesomeness, and wear cool monocles and develop bone-chilling laughs.

“Lab, garage, whatever. Have fun’.


I assume, though, that if you’re reading this blog you know better. We know that making furniture and tools and changing stuff into other, slightly more organized ‘stuff’ is how we get to keep playing with our legos and erector sets long after we’re supposed to be adults. Right?

So you’ll understand why I find it sad when someone ONLY uses western planes, or ONLY uses Japanese planes (kanna). That’s like those annoying lego purists, who think erector sets are too clunky. Or the erector commandos who think legos are kids’ stuff. A good lab needs BOTH! And here’s why:

For softwoods, kanna are capable of a surface that is unlike anything I think western planes are capable of… kanna users often refer to it as ‘kanna finish’. In my view, Cherry is sort of the borderline wood. I think both infill planes and kanna do a spectacular job on cherry. Anything ‘softer’ – I tend to use kanna. Harder woods? I generally use western planes.

I’ll do some more talking about the differences between the eastern and western approach to planes in future entries, but for now I just want to share a little bit of the ‘why’ I like both. All these pictures are the same piece of not-very-straight-grained cherry. This is a castoff chunk. I wouldn’t use it in furniture.

First, a small kanna. (For those interested in details: it’s a 36mm Rakuzan ‘super blue steel’ blade by Yokosaka in a dai I made from Macadamia. Technically it’s more of a block plane, but I’ve set it up as a final smoother.)

Here’s the surface it leaves…

it’s really hard to photograph, but there is a glow to the surface that I can’t quite match with any western plane. Why not? Well, there are a few things, but one definite difference is that in a kanna the blade is the last thing to touch the wood. There is no burnishing of the surface as with the heel of a western plane.

Now – here’s an infill smoother on the same piece:

Nice shavings. Here’s the result:

Nice smooth surface, but it doesn’t quite have the same surface quality as the kanna. As soon as I place a finish on it, the surfaces look identical, but nonetheless I just want to show that there is more going on here than just push vs. pull.

So where is the infill advantage, you may ask? Well, woulde it help if I mentioned that the shaving above was taken against the grain? With a 47.5-degree bedded blade. A well-made infill isn’t immune to grain direction, but it might as well be in most woods – Cherry included.

Now for my fellow mad scientists out there, who may be thinking ‘wonder if you could make a hybrid of some sort’ I can tell you t
hat I am not completely sure. But hasty preliminary research indicates some cause for optimism.

New toys, new plane

I’ve been playing at single parenting this week while Jenn is out of town on business… I don’t think the PTSD is too severe, but it’s pretty much wiped out shop time. I did manage to steal some time over the weekend to get my new auxiliary bench up, and install a couple of new tools:

First up is a new patternmaker’s vise. This one is the new clone that a number of places sell. I’m hoping to acquire an old Yost version of the Emmert in the near future, which will replace this, but I got a really good deal on this one. I have to say that so far I’m very impressed with the vise. The castings certainly aren’t pretty, but functionally it’s really fantastic. It’s going to make shaping totes a much faster and easier process.

The second addition is a Grizzly compact bender, which you can see mounted behind the vise. I bought this to simplify the bends such as the one at the rear of the small miter planes. It’s a really simple contraption, but it’s a very welcome addition, and makes it MUCH easier to get precise bends without having to do much tweaking after the fact. It’s a tool I basically use for about two minutes every few weeks or so, so it lives under the bench, and mounts in under a minute.

I also made a start on the rear infill for the steel miter I’m working on. I put the recently-completed bronze miter to use tuning up the bed and shooting the end. The miter board in this picture, by the way, is Evenfall Studios’ deluxe shooter — a fantastic design, perfectly executed. It’s a much better board than I would ever have bothered to make for myself – and there’s something to be said for that.

Rough tuning the bed is done with a panel plane, which gives me a good head-start at flat along the bed’s length. I have a camber on my panel, though, so I followed it up with the miter plane to eliminate any hollow lingering from the panel.

Now I can start fitting the plane and hopefully get to the peining this weekend.

Return top