Wednesday, April 23, 2014

Kanna help you, perhaps? (II)

Kanna: n. A Japanese hand plane - the greatest tool ever invented for slicing wood to the finest imaginable levels, capable producing glassy polished surfaces with nary a tool mark. Many nuances is set up and tuning, yet comprised of only a wooden block, a couple of blades, and a metal pin.

The Sino-Japanese character for plane, kanna, '', when broken into its two constituent elements of '金' (metal) and '包' (envelop, wrap), has the literal meaning of the metal which envelops a portion of lumber in planing. The important part one might take from this, perhaps, is that the core constituent element of the plane, what defines it and makes it what it is, is the metal blade, not the wooden block. While the set up of the wooden block, dai, is quite important, it is the blade which is the expensive bit and to which 90% of the energy going into making the tool has been directed. The block is something that can be readily replaced, almost disposable (not to say that there aren't very high quality and relatively expensive plane blocks out there), while the cutting and performance of the blade may over time become treasured to you, the retention of cutting steel jealously guarded even.

It is the kanna blade (kanna-mi) which we shall continue to work on as we proceed through the steps of setting up a new Japanese plane.

Step 4: Dealing with a Sori Situation

As mentioned in the last post, when the cutting steel, hagane, and backing iron, jigane, are forge welded together, it is the marrying of two materials with quite different characteristics. As the forge weld cools, stresses are induced to the body of the blade. As the back of the blade is hollowed by scraping and/or grinding, further stresses are released. The usual effect on the blade after all is said and done is that the blade cups along the length of the forge weld - termed sori - and it may also twist, and may cup crosswise as well. The blacksmith can attempt to compensate for this effect by making the pieces to be laminated curved shaped slightly opposite to the direction in which they will bend, or some other such stratagem which I can only guess at, however it is extremely difficult to obtain a dead flat blade given the variables and vagaries of the process. The worst outcome would be to have a blade which is bent towards the jigane side, as it is much harder to try and flatten a convex hard steel surface. As it is preferable to err on the side of a concave surface, in the interest of facilitating the later flattening of the blade back, that is precisely what blacksmiths do - err on the side of having the blade cupped to the hagane side:


If you try to flatten a blade with sori, 'as is', you will have the following situation upon a sharpening stone, which, for purposes of illustration, is dead flat:


The area shown in red in the above sketch is the hollow left by the curvature of the hagane upon the commencement of the flattening process.

One could proceed to flatten the blade, and let's assume you take the process along until the cutting steel has been made dead flat, like so:


This works, however the process has made the cutting steel thinner than ideal at both ends. Were the curvature in the blade more pronounced, the thinning would also be more pronounced at each end.

So what's the big deal with that? Well, this problem is really one that affects both the long term life of the tool and the ease with which it can be kept flat. Consider that each time you sharpen you will be dressing the back off to remove the burr formed from grinding and polishing the bevel. And, from time to time over the years of use the blade will need to be tapped out to re-establish the 'landing' of cutting steel immediately behind the cutting edge, and this process will require the cutting steel side, or ura, to be ground to some extent again, and again re-polished. With every re-working and polishing of the cutting steel side, material is lost from the steel, just as it is when grinding the bevel to establish a fresh edge. The steel is getting thinner and thinner with each sharpening session. Ultimately, by making the uphill end of the cutting steel thinner by the strategy of simply flattening a blade with sori, you risk the eventuality that one day you will end up with apparent length left in the cutting steel but the steel is actually too thinned by this point - or all gone - thus rendering the tool dead before its time. And after all that time spent getting to know it, the triumphs, the failures, good times and bad, you have to let a dear friend go prematurely.

Many tool users, including ones in Japan, could care less about such things, and flatten a blade with sori all the same. The characteristic of a blade in which the user has tried to flatten a curved back is that the ura perimeter (ura-suki) becomes a gourd shape:


A new blade on the left, the 'el gordo' outcome on the right.

Another side effect of removing so much cutting steel to obtain flatness on the ura is that what is formed with the gourd shape is an ura with a lot more cutting steel present, which means that flattening and polishing the back from here on out is now a more onerous task. The main reasons for the hollow on the ura side to begin with is to decrease sharpening effort and to make it easier to obtain perfect registration of the ura on a flattening stone. If there was no hollow, given the 70mm blade width and typical steel hardness of Rockwell 62~64, you would spend quite a long time  - hours if not days- trying to flatten the back of the blade. And trying to perfectly register a broad flat surface upon a stone means fighting against the tendency of the steel surface to want to ride atop the water and slurry below. It can be quite challenging to obtain a flat and even polish across a large area, especially where it is most critical - along the very edge of the blade.

The gourd shaped ura is a possibly a sign of a tool user who doesn't care much for his tools (or perhaps doesn't care about the blacksmith who may have put all his hard-won knowledge and skill into making the tool), or doesn't quite know what they are doing, or who does know and doesn't care all the same - or some combination of the above. I've done this to a plane blade myself before I learnt that it was a poor way to flatten the ura. A plane blade with a gourd-shaped ura will work fine, but its ultimate lifespan has been likely shortened, and you've made dressing the black flat more work than it needs to be from here on out. So, best avoided.

Another approach to dealing with an ura having sori would be to only work the last 1/2 or 1/3 of the hagane upon the stone, like this:


This solution can work well IF the amount of sori on the blade is minimal, and IF you take a lot of care to focus all the pressure on the cutting edge of the blade, virtually 'floating' the mid section of the blade along on the stone arris. Those who are good at freehand sharpening could look at this method.

If the blade had a pronounced amount of sori however, this approach will likely not work so well, and you will simply replace the curved ura with one that now has a faceted flattened front half. The ura would likely look something like this:


It's at least functional - let's call it the 'mini-gourd'. The polish won't be even of course between the front half and the back half of the ura, and if you were to inadvertently move the flattening stroke on the stone further inboard, the registration off of the flat created is lost and you will create a secondary bevel right at the blade edge. This will be obvious also from the various polishing streaks you will see on the uphill side of the hagane.

Worse though, is the outcome in regards to fitting the blade to the dai: the back of the tool is still not flat, and thus as the tool is fitted further and further into the dai the likelihood of the blade digging into the support ramp down low is significant. The support ramps are supposed to be totally straight, and by cutting a curve into them, you spoil the fit. The arrow shows where the damage occurs in the following picture:


I'll remove the blade and zoom in so we can see the damage at the bottom of the indicating arrow:


You really want to take care not to cut the ramp like that.

In recent email conversations I have had with the proprietor of Japan Tool (see link to the right of the page), I learned some interesting tid-bits in regard to various blacksmiths and their approach to sori. I had thought that a blade with a lot of sori was a sign that the smith was less skillful, and a blade with just a smidgeon of sori indicated good work, however it is more complicated than that.

After the blacksmith has done their work, the blade is sent to a suiken-ya (水研屋), and one of their tasks is to perform a basic sharpening of the blade, and this necessarily involves dealing with the sori issue. Some smiths are very fastidious in their work, taking extra care to stamp their makers mark cleanly, shape the blade carefully, form a shapely hollow, etc.. There are other blacksmiths who consider this fastidiousness as, well, undesirable. It's not manly. It's over-fussed. Too pretty. They prefer a bit more of a rough and ready approach, and are less interested in making something pretty than in making something that works well as a workaday tool an average craftsman can afford. I hope you see what I am getting at  -they don't produce a tool with all the wrinkles worked out, partly because of cost considerations and partly because of philosophical reasons. The time of the suiken-ya costs money, so in order to keep the costs down on the tool, the labor the suiken-ya might otherwise have been asked to do is simply off-loaded to you, the consumer. And if the manly, rough-and-ready plane buyer couldn't give a fig if the ura looks like a gourd and is happy to flatten the frickin' thing as is, then all is good.

It's an interesting philosophical point of view, and while I understand it, frankly I would rather have paid the suiken-ya to do a thorough prep on the tool. They are used to working with metal - plane blades specifically - all the time, and are doubtless going to do a better job than I will.That said, perhaps the performance of the tool over time will win me over, and I'll forget about all the struggles we went through together. To borrow one comment from the Japan Tool owner, I may end up marrying this girl, though I found her ugly and obnoxious at first....

Another option now for dealing with sori - this is the method I use. The idea is to find a way to suspend or support the blade so that it can ride upon the stone with the rear portion of the sori close to parallel with the stone surface:


You could try free-handing this if you like. I prefer to find a means of supporting the blade, at least for as long as it takes to obtain a registration flat at the bottom end of the hagane.

Before supporting the blade though, it is important to determine exactly how much curvature you are dealing with in the first place. I do this by placing the hagane portion of the blade down upon a reference flat, and seeing how large a feeler gauge can be slipped in to the side:


This blade has a very minimal amount of sori - 0.0015". One of my other new plane blades however had nearly 4 times that, at 0.0055".

The idea with this method is to find some tape that you can place on the uphill side of the blade, and  the tape you select will have about the same thickness as there is sori.

How thick is tape? Well, it can be easily measured.

Painter's tape:


Aluminum duct sealing tape (which I used on the blade with the worst amount of sori):


Clear heavy duty packaging tape: 


I don't think those round numbers for tape thickness obtained above are accidental.

Once you've selected the appropriate tape, place it on the upper half of the hagane:


With the tape in place, you can then start to work the blade on the stone, the entire hagane zone in play over the stone:


You can start with a finer diamond plate, like an Atoma 1200, if you are confident that the plate is flat. Or, as shown in the above photo, start with a 1000 grit stone. When using a stone to flatten the back, you must be extremely fastidious about keeping the stone flat. I would suggest taking about 10 strokes, then re-flattening the stone again. Concentrate your finger pressure only on the bevel end of the blade.

The function of the tape here is to act like a little temporary sled to support the uphill side of the hagane. You are still looking to pretty much float the blade over the stone on the taped end, applying almost no pressure there beyond the weight of the blade itself. The tape of course doesn't last long, even when taking care not to wear it, so you need to look at the condition of the tape each time you go to re-flatten the stone.

If the tape starts to wear through, replace the tape. It's only tape. It only needs to do its job long enough for you to adequately flatten the sori at the end of the blade. Once you have established a flat on the end of the cutting steel, it will begin to self-jig. Then you can spend most of your time working the hagane on the last 1/3 portion.

Again, about 10 strokes on one side:


Then 10 strokes on the other side:


I also use the middle strip on the stone to work the bevel. Then re-flatten the stone. My grip on the stone you see above is in compensation for holding the camera in my other hand - normally my right fingers would be pressing on the bevel.

Once the back is flattened, you can work the bevel, staying on the same 1000 grit stone (or drop down to the diamond plate if the bevel is pronouncedly convex and there is a lot of steel to remove). It is important again to concentrate on keeping the coarse stone dead flat, as a flat stone will produce a flat bevel. A check to see if you are obtaining a flat bevel can be done when you are in the middle grit stone(s) - the blade should stick to the stone on the bevel for a period of time:


If it barely sticks, either the bevel or the stone is not flat. If it sticks for about 5 seconds, that's decent. 10 seconds is excellent with a larger blade. With a smaller (shorter and lighter) blade, like a 48mm, I have had one stick for over 30 seconds.  It's not a contest, but it is satisfying to know you have obtained a decently flat bevel - even if you may later plan to make the blade edge have a slight smile (curve) to it.

You keep working the same pattern with the next stones in the sequence. I move from 3000 grit ceramic (shown above) then on to a natural stone.

Here are a couple of blades after the above ura flattening process has been completed:


You can perform a check at the end similar to the start: place the hagane side down on the stone and see if you can get a feeler gauge in there. There should be no space. Both of these blades had only a modest amount of sori to begin with, so the amount of work was on the reduced side. The blade on the left is about where I want it, with a land on the back side of the edge about 1.5mm wide. The land is not perfectly straight, but acceptable. The blade on the right has a 1mm land, and I would prefer it be a little wider. I could keep flattening the entire back, starting down on the coarse stone, however the portion of land on the sides of the blade- threadlike in the above examples, would then begin to fatten. And, for as long as possible, I want to keep those lands - those legs - on the skinny side.

The blade which had the most pronounced twist and greatest degree of sori ended up like this after the above flattening process:


I have had more of a battle with this one, as you can see. Here the land behind the edge is more like 4mm wide, and the left side, which had ura-a-ge performed (see the previous post) to remove the twist, has ended up with a somewhat irregular hollow. Not the prettiest, and there are a few scratches yet to be dealt with. I feel like I won the battle.

So, with the one blade needing a bit of a wider land behind the cutting edge, rather than trying to take the entire back down some more, I will instead perform ura-dashi on the land behind the edge. That will be the subject of the next post.

Thanks for tuning in, and comments always welcome.