The Convex Conundrum - The Three Principles of Using Japanese Finish Planes

[Stanley Covington]

As we enter Autumn, the falling leaves seem to have clogged the course of Sawmill Creek a bit, so I am floating a log downstream in the hope that it will liven things up.

Like most non-Japanese, I taught myself how to use Japanese planes. Later, I was given the opportunity to learn from highly skilled Japanese professionals in Japan, but found the experience discouraging at first when I discovered I had to unlearn bad habits.

By the way, none of the old boys I learned from had written a book or magazine article, or starred in a DVD on how to use a plane. None of them sold tools, or taught classes, or shilled for tool retailers. For that matter, none of them had ever even read an article or book, or attended a workshop, on using tools, but had woodworking skills literally pounded into them by foot and fist when they were just smooth-faced boys right out of middle school. They were no-nonsense, 12-hour-a-day craftsmen.

One thing I had to learn very quickly was the importance of using a plane to make surfaces as flat and smooth as possible right off the blade, and when that was not possible, to make them at least appear flat and smooth. Sandpaper was never, ever an option. It's not rocket surgery, but these techniques contradict what most if not all the scribblers in the modern English-speaking woodworking press advocate. The three principles I had to learn are as follows.

The first principle is that only a straight blade can make a truly flat surface. It is of course self evident that a convex blade can only produced a concave or scalloped surface, one that is more likely to be out of square in limited areas, and to induce problems when joining pieces together as in a door or window. This means that the effective width of the blade of a plane used to finish plane a surface should, whenever possible, be at least as wide as the surface being planed, otherwise the blade will create "steps" or ridges in the wood's surface. This point is of course applicable when planing joinery pieces as in frames for doors and windows and architectural trim, much of which is less than 72mm in width.

The second principle is that, in the case where the available plane blade is too narrow to plane a surface in a single pass, so that multiple passes are necessary, the final finish cuts should be made with a blade sharpened ever-so slightly convex, especially at the right and left corners, or "ears," of the blade, so that the blade's corners disappear smoothly into the mouth of the plane without any exposed corners that would otherwise cut ridges or tracks into the wood's surface. This technique is applicable in the case of panels and beams wider than, for instance, 72mm which is the effective cutting width of the standard finish plane. It is also one that I believe is well understood by Western woodworkers nowadays.

To maximize the width of a workpiece that can be planed dead flat in a single stroke, there are standard planes in Japan with blades wider than 72mm, such as 80mm, but a 72mm wide "sunpachi" is as wide as most people can handily use. However, there have always been craftsmen in Japan that wanted the perceived efficiency and notoriety of being able to finish plane an architectural column dead flat in a single pass. For this purpose, blacksmiths occasionally forge blades 300mm and wider. If you have seen the online videos of the Kezuroukai, then you may have seen guys using these extremely wide planes. Not so hard to sharpen, I am told, but very difficult to setup and maintain.

The third principle is derivative of the first two and dictates that a craftsman have at least two finish planes in his toolbox: One with a blade sharpened dead straight, and another with a slightly convex blade dedicated to finish planing panels and wider surfaces. I have found this arrangement to be very efficient and to yield excellent results, especially when prepping multiple pieces to be joined together. But few if any of the modern Western gurus seem to teach what is SOP here in Japan.

If you use a convex blade in your planes for all finish planing situations, have you experienced any difficulty planing pieces flat and square, and creating tight joints?

Is it important for the surfaces you produce to be flat and smooth right off the finish plane blade?

Do you typically sand and finish pieces with some product after joining, or do you leave the bare wood exposed?

Please share your experiences.

Stan

[Frederick Skelly]

Stan,
I use only Western planes at this juncture. I very seldom use sandpaper since I discovered finish planes and how to sharpen them. I do feather the corners sliightly, though I doubt I have a uniformly convex edge. I usually go from the finish plane to shellac or paste wax. The panels I make are smooth but certainly not perfectly glass-like, as real Masters like you work with routinely produce. But I get a little better every time.

As always, I enjoyed reading your post. l especially enjoyed your artistic opening paragraph. You're learning that from your Japanese colleagues too, aren't you?

Fred

[Brian Holcombe]

Stan,

As I progress with Japanese planes I have also found that having two standard sized planes to be my approach. I do find flattening furniture components which can be planed in a single stroke to be best completed with a flat blade.

In addition to what you have stated above I have a 48~mm blade that I use for assembled cabinets, I set this to a extremely fine cut and use it to brighten up assembled cabinets. It makes small scallops, but then I return to cut the ridges down and the finally result is very smooth to the touch.

Sandpaper does not touch my cabinet work with very, very few exceptions. Most often I am finished with structural components the moment they are dimensioned since my final passes are with a finish plane, with exception to if I am cutting joinery into them then I will take a pass or two after that is complete. The less often one needs to work a dimensioned piece, the better, as once they are cut to size then the only thing left to do is to make them undersized.

I have a 7 jointer also setup with a flat edge and I use it to finish components or edges save for one or two strokes by my flat blade smooth plane. This was the 'zamboni' plane for our NYC Kez, where a few cleanup passes were taken between rounds to give the next person a fair shake. The same approach applies to components, cutting them truly flat leaves only a pas or two left for the finish plane. It has the added benefit of needing to wear out the finish plane much less than otherwise required.

I typically do the majority of my work up until that point with a cambered blade.

[Stewie Simpson]

I follow the 1st and 2nd principal as outlined.

Stewie;

[Phil Mueller]

Stanley, thanks for starting this post...I'm looking forward to the discussion.

A somewhat novice perspective. I've transitioned to David Charlesworth's cambered iron method (I'm using a #5) for flattening and squaring both edges and faces. I finish with a straight edge smoothing plane with slightly cambered corners. Depending on the result, I do a final sand on about 75% of the work at the moment.

Reversing or difficult grain remains my prime "learning opportunity", as I am still challenged to read the grain direction and manage it without tear out. At this stage, I typically get difficult wood as flat as possible, short of any major tear out, and finish with sandpaper.

My goal is to minimize the sanding...I'll get there sooner or later.

[Stanley Covington]

Fred:

Glad you enjoyed the post. It is pleasant to occasionally write something light that does not deal with money, schedule problems, or quality defects in construction projects.

You are right about the opening address. It is traditional in Japan to begin a personal letter or friendly communication with some mention of the seasons and/or the current weather. I find myself unthinkingly adapting it in admittedly strange ways in my English communications too.

One of the key things to creating smooth panels with a finish plane is to set the cut depth extremely shallow. The shavings produced when finish planing light-colored woods should be translucent when possible, to the extent that you can read newspaper print through the shaving. This is not difficult to accomplish with a good, sharp, properly setup plane be it a Japanese hiraganna, a Bailey pattern steel plane, a horned German plane, or a Hong Kong side-handle plane. This is the starting point, BTW, for the Kezuroukai. Brian has an interesting report on his blog you might find interesting if you haven't read it yet. I suppose that the Kezuroukai is to woodworking what NASCAR is to moonshine delivery.

In any case, if the plane has a sharp blade and is setup and adjusted properly to take a very thin shaving, tearout is greatly reduced even when planing against the grain. This is the real purpose of the finish plane in the Japanese tradition. In the Western tradition, the term finish plane seems to refer more to a specific length plane rather than the actual function of the plane.

Good luck in your efforts.

Stan

[James Pallas]

Stanley, Thank you for writing this post. I find these things very interesting. Methods of work for different cultures can be very enlightening. I had a question about Japanese methods. From watching videos it seems that the Japanese workers go for the finish right away as opposed to saving the final prep for finish after the piece is assembled. If this is so how do they go about protecting the work in progress.
Jim

[Stanley Covington]

Stanley, Thank you for writing this post. I find these things very interesting. Methods of work for different cultures can be very enlightening. I had a question about Japanese methods. From watching videos it seems that the Japanese workers go for the finish right away as opposed to saving the final prep for finish after the piece is assembled. If this is so how do they go about protecting the work in progress.
Jim

Jim:

It depends on whether or not the piece will be given an applied finish or be left bare wood.

Sandpaper, or an equivalent, has been used historically for prep before applying a finish such as urushi lacquer, but the job of the carpenter or cabinetmaker or joiner is to make the surface perfect, raw wood when he is done. Where such a finish is applied, of course, temporary protection is not as necessary as it is for pieces that will not receive a finish.

For bare wood (aka "shiraki" 白木 or ""white wood") the surface left by the finish plane is the final finish, so it must be perfect. There are two ways to achieve this. In the case of doors, shoji, tansu chests and other projects where the finish plane can work all sides and all surfaces unobstructed, a final pass with the finish plane will restore the finish to perfection. Remember, the finish plane is removing only a very, very thin shaving of wood with each pass, not enough to mess up dimensions significantly. This is the ideal.

In the case where a final pass of the finish plane is not possible, the freshly planed wood is wrapped in paper (and foam cushioning nowadays) before and after assembly to protect it.

Japan has an abundance of excellent softwoods. Not as many huge trees as there once was, but still lots of volume. The cedar varieties such as Yoshino shugi and Akita sugi smell wonderful when freshly planed. There is nothing like it in the US. Yoshino sugi, and to a lesser degree Akita sugi, develop a beautiful shine when finished properly. Hinoki (cryptomeria) is another excellent wood that planes leaving a beautiful surface and a wonderful smell. A hinoki bathtub is wonderful thing to experience. Hinoki is similar to Port Orford Cedar, but smells better IMO. It planes and works similar to Alaskan Cedar but does not have the same mediciney smell.

Hinoki is a great wood in all respects. Naturally rot resistant like many cedars, it is highly unusual in that it develops maximum strength in bending, tension, and compression approximately 300 years after the tree has been cut down. Most varieties of wood begin losing strength gradually soon after the wood has dried to equilibrium. This is one reason why Japan's shrines, built almost entirely of unfinished hinoki, have survived so long.

Another interesting thing is that, regardless of the variety, wood that that has been properly planed and left unfinished prior to exposure to sunlight and moisture is much more durable than the same wood that has been sanded and treated with a varnish or paint. This is because the cleanly cut fibers (which are are mostly tubes) do not allow as much water to penetrate into the wood (due to hydrogen tension), while the fuzzy surface left by sanding tends to pull much more rot-inducing water into the wood quickly and deeply. Also, the fuzz is quickly oxidized by UV rays further promoting cellular degradation.

Surface finishes such as varnish and paint, always fail after UV exposure. Water then penetrates into the wood where the remaining varnish or paint traps the water. Unless the varnish or paint is religiously maintained, it does not tend to make exposed wood last longer, especially when applied on top of a sanded surface.

The point of all this meandering is that the woodworkers of the Japanese islands have always had wood that responds well to planing, and when properly planed, develop a beautiful, fragrant, extremely durable surface that ages gracefully. The Japanese have a deep, genetic love of cleanly planed, bare wood, and the finish plane is an important tool in maximizing these traits.

Stan

[James Pallas]

Thank you again Stanley. The paper or padding makes perfect sense. It brings more questions. Are card scrapers or similar part of a Japanese workers kit? In what cases is end grain left exposed? Is air dried material more typically used for fine work. I know, I know, too many questions but you have piqued my interest.
Jim

[Stanley Covington]

Thank you again Stanley. The paper or padding makes perfect sense. It brings more questions. Are card scrapers or similar part of a Japanese workers kit? In what cases is end grain left exposed? Is air dried material more typically used for fine work. I know, I know, too many questions but you have piqued my interest.
Jim

Jim

Card scrapers are not traditional tools, but are seeing some use nowadays,especially for truing plane soles. Not sure where you would go to buy one in Japan, though.....

Exposed end grain is OK for architectural and structural applications, but is considered unrefined in cabinet, casework, and joinery. Of course, hiding endgrain entirely can be expensive, as in the hidden dovetail mitre at casework corners, so large pinned fingerjoints with exposed endgrain are common too for less expensive work.

I am not aware of any use for green wood other than for crafts and baskets and sich. The best wood is air dried, of course, because it has abetter color and scent, but kilns are used too. I have no idea what the ratio of air-dried to kiln-dried might be.

Stan

[Patrick Chase]

Another interesting thing is that, regardless of the variety, wood that that has been properly planed and left unfinished prior to exposure to sunlight and moisture is much more durable than the same wood that has been sanded and treated with a varnish or paint. This is because the cleanly cut fibers (which are are mostly tubes) do not allow as much water to penetrate into the wood (due to hydrogen tension), while the fuzzy surface left by sanding tends to pull much more rot-inducing water into the wood quickly and deeply. Also, the fuzz is quickly oxidized by UV rays further promoting cellular degradation.

I'd always wondered if this is the case. It makes intuitive sense - A cleanly cut surface exposes fewer fibers, which reduces wicking (which is what I think you're referring to as "hydrogen tension"?). Less exposed fibers means less surface area, and therefore less oxidation.

Surface finishes such as varnish and paint, always fail after UV exposure. Water then penetrates into the wood where the remaining varnish or paint traps the water. Unless the varnish or paint is religiously maintained, it does not tend to make exposed wood last longer, especially when applied on top of a sanded surface.

Indeed. A film finish doesn't do much over the long term unless you periodically replace it.

EDIT: Added "fewer" as that was what I was trying to say.

[Derek Cohen]

One of the key things to creating smooth panels with a finish plane is to set the cut depth extremely shallow. The shavings produced when finish planing light-colored woods should be translucent when possible, to the extent that you can read newspaper print through the shaving

Thanks Stan. Intriguing.

Your comments take us full circle back about 4 years, when fine shavings began to be displaced by the thicker shavings under a chipbreakered smoother. Very fine, translucent - read through - shavings have long been the hallmark of a well set up handplane. However, in recent years they were seen to be unnecessary if planing could be done without tearout, and indeed fine shavings were considered slow and cumbersome in terms of speed and efficiency. Of course, a tearout-free surface is not necessarily a finely finished surface. One factor that promotes faster as better, in Western woodwork, is that a finish is nearly always applied. Most, if not all, finishes will obscure differences between planed and sanded surfaces.

Posts like yours - where information provided adds a fresh understanding why we do something, rather than just what we do - can influence a lot of people to work better, with more confidence in what they do. There is so much apparently contradictory advice on the 'Net that I am sure that many fail to understand that methods are not black and white.

Regards from Perth

Derek

[Stanley Covington]

One of the key things to creating smooth panels with a finish plane is to set the cut depth extremely shallow. The shavings produced when finish planing light-colored woods should be translucent when possible, to the extent that you can read newspaper print through the shaving

Thanks Stan. Intriguing.

Your comments take us full circle back about 4 years, when fine shavings began to be displaced by the thicker shavings under a chipbreakered smoother. Very fine, translucent - read through - shavings have long been the hallmark of a well set up handplane. However, in recent years they were seen to be unnecessary if planing could be done without tearout, and indeed fine shavings were considered slow and cumbersome in terms of speed and efficiency. Of course, a tearout-free surface is not necessarily a finely finished surface. One factor that promotes faster as better, in Western woodwork, is that a finish is nearly always applied. Most, if not all, finishes will obscure differences between planed and sanded surfaces.

Posts like yours - where information provided adds a fresh understanding why we do something, rather than just what we do - can influence a lot of people to work better, with more confidence in what they do. There is so much apparently contradictory advice on the 'Net that I am sure that many fail to understand that methods are not black and white.

Regards from Perth

Derek

Derek:

Thanks for your insight. Sharp as always.

I don't follow the chipbreaker thing from 4 years ago. Chipbreaker debate threads on SC are always so silly, it seems to me, kind of like arguing whether it is better to travel to the moon by balloon or rocket, that I simply skip them.

Unlike the Kezuroukai, thin shavings, in my experience, are not the objective in making a fine cut, but are a by-product of producing a smooth, apparently flat surface with a handplane. One can of course achieve a smooth finish while cutting relatively thicker shavings, but thicker shavings mean larger changes in dimensions, and a higher likelihood of tearout, and bigger, more easily discernible ridges when using a convex blade for multiple parallel passes on wider members.

So it is most efficient to have a couple of dedicated finish planes on hand, one with a straight blade, and one with a fine convex blade, both setup specifically for taking very fine shavings. Planes setup for thicker shavings are used for dimensioning, truing, and flattening as quickly as possible. These planes will need to be sharpened more frequently, and their soles will wear and get out of wack quicker than the finish planes. The finish plane is used much less so it stays sharper longer, and the blade's setting (which take some effort to get just right, as I'm sure you know) and the fettle of the sole are ready to rock-n-roll longer. Since the shavings it takes are so thin that the first few passes won't produce full-width shaving as it smooths down ridges and cleans up tearout. But when the finish plane's work is done, the surface will be as close to perfect as a hand-powered tool can manage.

I'm certain you already know this, Derek, so sorry for preaching to the choir, but for those who find what I am writing strange, please realize that the effectiveness of taking translucently thin shavings on the last few strokes is especially clear when using the convex-blade finish plane on wide panels. Since the plane's blade cannot plane the full width of the board in one pass, steps or ridges are unavoidably generated regardless of the shape of the blade. Of course, the sharply defined corners of the straight blade will produce sharply defined ridges, while a convex blade will create less-apparent ridges where plane passes meet. But when using a convex blade for multiple passes, so long as the radius of the convex curve (or radius just at the blade's corners, if you prefer) is very large (a very shallow radius), and the shavings taken are very thin, most people in most circumstances are unable to see or feel the ridges.

Where the convex blade is not an issue, thinner shavings are simply weaker than thicker shavings and have less leverage to induce tearout, producing a smoother finish.

I hope this makes sense.

Stan

[Derek Cohen]

Of course, the sharply defined corners of the straight blade will produce sharply defined ridges, while a convex blade will create less-apparent ridges where plane passes meet. But when using a convex blade for multiple passes, so long as the radius of the convex curve (or radius just at the blade's corners, if you prefer) is very large (a very shallow radius), and the shavings taken are very thin, most people in most circumstances are unable to see or feel the ridges.

A corollary of this is the way a smoother blade is cambered: too little camber and one creates tracks; too much camber (such as when only the centre third of the blade produces a shaving), then it will be difficult to avoid a sculpted surface. The aim is to take as near full width as possible.

Stan, do you have a method you can share?

Regards from Perth

Derek

[Stanley Covington]

A corollary of this is the way a smoother blade is cambered: too little camber and one creates tracks; too much camber (such as when only the centre third of the blade produces a shaving), then it will be difficult to avoid a sculpted surface. The aim is to take as near full width as possible.

Stan, do you have a method you can share?

Regards from Perth

Derek

Allow me to add one more point to your corollary. It is neither time efficient nor cost effective to spend time and money on a quality blade of any width if the entire width of the blade cannot be used for planing wood. Therefore, in the case of a finish plane, the radius or curvature of a convex blade's cutting edge needs to be very shallow indeed.

The way I was taught to make this curvature is to first begin with a sharp, straight blade on a truly flat medium-grit stone. None of this hollowed-out stone or profiled steel plate nonsense, at least not for a finish plane. Then the blade is rested half on and half off the medium grit stone's edge, and pushed forward while at the same time sliding the blade sideways until just a few millimeters of the blade are supported on the stone. Both right and left sides of the blade must be worked to the same degree of force and number of strokes. The point is that the center of the blade spends very little time on the stone. Only a few strokes are necessary. It takes practice to train the muscles.

This process is repeated on the finish stone.

Finally, center the blade on the finish stone and make multiple strokes with the entire width of the blade on the stone, while shifting pressure across the blade's cutting edge. This will help to smooth things out. Once the blade is shaped this way, it will not need to be repeated until it dulls enough to warrant time on the medium stone.

The blade should project from the dai's mouth full-width, and apparently straight (but actually slightly curved) until it curves up into the mouth. Once again, this is the case of a finish plane intended to cut translucent shavings, not arashiko or nagadai planes for dimensioning and smoothing.

Stan