Wednesday, February 14, 2018

Thoughts on designing a 240 WA gyuto

1. Introduction

Make yourself a nice tea and find a quiet place, this is a 6.500 word article :)

Please read this first ...

I was pondering to write an article on a design of a gyuto (or chef's knife) since quite a while, but was (and still do) feeling uneasy about it. Indeed - my personal practical experience is still very limited and I most certainly do not want to make the impression that I have 'mastered' the topic.

In my view, the Japanese approach to kitchen cutlery with its focus on the function rather than 'looks' has yielded the best performing kitchen knives known to men. So contrary to numerous kickstarter projects I am not trying to 'reinvent' the kitchen knife here. Instead I am just trying to make to learn from Japanese kitchen knives and transfer those parts of the design and geometry that are accessible to my skills, tools and techniques.

What I try to share here is in the first place based on my observations (and discussions with people who have much more knowledge on the topic than me), rather than some kind of 'teaching'. Japanese craftsmen have been improving their processes and designs over generations. In comparison I spend a few hours in my 6 m2 workshop every now and than.

Since I do not posses neither the knowledge, not the practical experience of a skilled Japanese smith or a sharpener, all I can offer is in a sense a somewhat superficial description of what I understand that constitutes a well performing kitchen. So when executed well it may yield a pretty well performing knife that is a pleasure to use and a cut above the knives you can find in a local shop.

What to expect

This article is an attempt to 'analyse' Japanese kitchen knives from a perspective of an amateur knifemaker from central Europe. Since a design of a kitchen knife is first and foremost determined by the utility function of the knife, there is not a whole lot space for 'innovations' as you rather quickly end up with a cool looking, but not a very useful knife. 

So the main point is probably the attempt to understand what makes Japanese kitchen knives to stand out, rather than to give you a tool to design something completely different.


Let's for the sake of simplicity talk about a particular type of a design, so that I can keep the number of variables down:

  • Gyuto with a 240 blade length
  • Blade made from single piece of steel via stock removal
  • WA (Japanese) style handle with octagonal cross section
  • Double bevel, symmetrical grind 

This topic can not be discussed without some hand drown sketches.  I even went as far as making a few short clips to show certain features of Japanese knives. please bear with me here, I am doing my best.

2. Japanese kitchen knives

Note: This is probably the most valuable part of this whole article as its contents were checked with a very experienced vendor and a professional sharpener of Japanese kitchen knives who helped me to correct my misconceptions and pointed out several details that would have otherwise escaped me.

There is little chance one is going to make a knife that could perform similarly to Japanese kitchen knife without using some and trying to understand how their design and execution relates to their cutting properties. I am using Japanese kitchen knives as a home cook since a few years and I am also lucky to have the chance to discuss the details about their grinds and geometry with people who have much more knowledge on the topic and have regular contact with Japanese craftsmen.

Grinds of Japanese kitchen knives

So if we go with the limitations as listed in the introduction, we are left with basically 2 different types of grind:
  • Convex grind
  • Wide Bevel grind (grind with a shinogi line)

  • Knives with both convex and wide bevel grinds have several features in common:
    • Very most of the knives have a distal taper - the thickness of the spine is larger at the handle than towards the tip. Obviously very thin knives will have less pronounced distal taper. Even beyond that - the distal taper can be quite different and there are differences across regions in Japan. In general though the distal taper is not linear from the handle to the tip. If the tang is forged thick (what has advantages when it comes to stiffness and prolongs the lifetime should the tang rust inside the handle), than the taper will be very pronounced towards the heel of the knife, than it will 'slow down' over the most of the length of the blade and will get again more pronounced towards the tip. If the taper is made linear over the length of the blade, than this results in a blade that is too thin towards the tip and thus also too flexible (what is usually not desirable in a gyuto).
    • The distal taper also influences the grind of the blade as it dictates the thickness of the blade at the given position along the blade.
    • Most knives also have taper of the thickness of the grind behind the edge, meaning that the grind is thicker at the heel and gets a bit thinner towards the tip. At 10 mm behind the edge this taper is usually between 0.1 - 0.3 mm, on a rare occasion more (usually only on generally thicker knives like Kato Workhorse). In general this will be less for thin knives and more for thicker knives.
    • Talking about thickness of the blade, than at 10 mm behind the edge a thickness of 1.0 mm would be the middle ground - not too thin and not too thick. Including the distal taper one could think of a blade that has thickness of say 1.1 mm close to the heel and some 0.8 - 0.9 close to the tip. This would only be a rule of thumb for where to start - there are considerably thicker and also considerably thinner knives out there.
  • The convex (hamaguri) grind:
    • The convexity of the grind is most pronounced first 10 - 20 mm behind the edge, the curvature of the grind is not constant but changes with the distance from the edge. 
    • Moving away from the edge the grind may remain convex up to the spine (like it seems to be the case on a Kato Workhorse), but often turns to flat (or gently concave) bevel, so the cross section of the grind would look like a very sharp 'V' where the last 10 - 20 mm at the tip are convex.
    • There is often a pronounced transition (shoulder) between the convex and the near-flat grinds - like as if two bevels were merged into one (what may well be the case). This 'shoulder' would usually be found at around 15 mm from the cutting edge.
    • Also, blades like the Gesshin Ginga, for example, are done in stages, so there are generally 3 distinct convex ranges- near the spine, middle of the blade, and near the edge (with increasing convexity moving towards the edge). 
    • Towards to the tip the bevels above the convex grind may not have any horizontal taper left. This is mainly because the thickness of the spine tapers towards the tip and there is simply 'not enough thickness' left for it.
    • In general it needs to be noted that the convexity of the grind is subtle (and I really mean SUBTLE) and should not be over done. The blade may even look flat at first sight (looking at the choil from behind), even though it is convex.
  • The wide bevel grind:
    • These knives have a sudden change in the angle of the grind which produces a distinct line that runs parallel to the cutting edge and is called shinogi. 
    • The width of the bevel can be anywhere between about 10 to 30 mm.
    • Nearly all knives with kurouchi (black finish after forging & HT that is left on finished blade) are of a wide bevel design.
    • The wide bevel (part of the grind between the cutting edge and shinogi) is very near-flat, but is intended to be slightly convex (hamaguri) - in particular close to the edge. It can be also slightly concave like e.g. on Kochi knives, but this is a result of grinding process (grind being made on a very large grinding wheels - around 100 cm in diameter), not the final intended geometry. Overt time - with proper sharpening and blade maintenance (thinning) the concave grind will be turned to either flat or convex wide bevel. Here I would add - if you grind knives on a belt grinder with a platen (the typical grinding wheels for belt grinders are too small to grind a wide bevel on a kitchen knife), than you would not produce the concave grind to begin with.
    • The bevels above the shinogi are most often flat (though on some knives like Takeda they are forged slightly concave), and similarly to what was said about the convex grind may show some vertical taper or be (usually only towards the tip) of nearly constant thickness.
    • Advantage of a wide bevel grind is that it is a little easier for the user to maintain the geometry over time - in particular if the bevel is not too wide. Once the thickness behind the edge needs thinning, than the whole bevel will be equally thinned up to the shinogi line and the shinogi line would move a tiny little bit up.

3. Basic knife parameters

For this part I decided to use one knife as an example. It is actually a rather rare knife known in the community as 'Kato Workhorse' made by Yoshiaki Fujiwara (a.k.a. Kato) in size 240 mm. This particular one has a hand engraved Dragon on the left face and there were just a few made like that. This is a fairly thick & heavy knife for a Japanese gyuto with a WA handle at 250g and as such not to everyone's taste. Still - it is a great example of excellent design and execution.

Kato Workhorse Dragon 240 mm

Detail of the engraved dragon


In my current view the basic parameters that have the main influence on how the knife will behave while cutting are the following:

  • Weight
  • Position of the center of mass
  • Distribution of the weight
  • Profile of the cutting edge
  • Angle between the handle and the 'flat spot' 
  • Height of the blade at the heel
  • Grind / geometry
  • Grind close to the cutting edge

  • Hamaguri grind

  • Neck design
  • Handle type / design & materials

It should be noted that these are not completely independent from each other. If you make a significant adjustment of one of the parameters, most of the others will be affected too.

The weight

This is an obvious one - a lightweight knife will be more nimble and faster while a heavier knife will (with proper grind and weight distribution) be able to use its own weight to - as they sometimes say - 'divide instead of cut'. A 240 gyuto with a WA handle may weight as little as 130g or as much as 280g.

Important is that the geometry - in particular with lightweight knives - is tuned properly. While it is physically hardly possible to make a tick grind (like on Kato Workhorse) on a very lightweight knife, still - a lightweight knife that has grind on the thick side may require too much downward force when cutting. The same grind may be just optimal on a thicker knife. The other way round works though - a knife like for example Kochi (weight about 200g) has a very thin grind.

Position of the center of mass

With a small knife which weights very little (like 150 petty) the position of the center of mass is not that crucial (even though a heavy handle on a thin lightweight blade may still feel less comfortable), but with a gyuto it will make a large difference in how the knife will feel and handle. A knife that has centre of mass close to the handle (say on the heel or even the ferrule) may feel 'balanced' in a hammer grip, but potentially handle heavy with a pinch grip (which is the most common way to hold a gyuto). The further away the centre of mass is the more weight will there be on the tip in the use. This on one hand makes the knife a bit less 'nimble', but on the other hand makes cutting somewhat easier, as the weight of the knife means that one does not have to push down so much. This really is something one needs to experience. It is not about which is better or worse - it is mainly a question of a personal preference.

The distribution of weight

This should not be confused with a position of the centre of mass, even if in reality these two are related. Imagine a knife (A) which has a lot of weight located around the heel (tall blade at the heel with a thick neck), but tapers strongly towards the tip in both width and height, while having a simple and lightweight handle. Such a knife even if not super light will feel easy to move around. A knife (B) would have blade with little distal taper with a relatively tall blade towards the tip, but balanced with a more substantial, heavier handle. Both knives may have a centre of mass around the same position and may even have similar weight. The second knife however will feel less nimble or less easy to more around as any rotation of your wrist will have to move that much heavier tip around. Here again personal preference plays a large role.

The profile of the cutting edge

While there are some basic 'rules' that one should at least seriously consider when it comes to profile of the cutting edge, it needs to be recognised that many factors plays a role in what one may consider to be an optimal profile of the cutting edge - the most relevant ones would be cutting style of the user.

I personally find that about the best starting point for the profile of the cutting edge can be found on gyutos made by Konosuke, Ichimonji, Hide, Gengetsu, Itinomonn, Shigefusa, etc.

I will use the profile of the Kato Workhorse as an example (see the figure below).

So the 'rules' would be:
  • There should be a flat spot (B in the figure) at the heel that would allow 'accordion-free' cuts when coping vegetables. The length is hard to define exactly, but somewhere around 5-8 cm should work
  • The flat spot should not be perfectly flat, more like a very large radius, deviating from a straight line just a fraction of a millimetre over its length. This curvature makes the knife less prone to get 'stuck' in a cutting board plus it allows for that minimal 'rocking' motion while chopping, which is sometimes wanted, but in reality can never be completely eliminated as it is part of the dynamics of the cutting movement by one's hand/wrist
  • The flat spot should be integrated in the whole profile of the cutting edge without any 'bump' or sudden change in the curvature, otherwise it may disturb with push-cuts
  • The profile at the very heel (last few millimetres) should have a slight up-curve (C). This serves to minimize the risk of the heel getting stuck accidentally in the cutting board (and possibly get damaged in the process)
  • Once you move away from the flat spot towards the tip, the curvature of the profile should change continuously, but not linearly, so that the most curved part of the blade is close to the tip.
  • I often use a 'height of the tip' - a position of the tip of the knife relative to the line defined by the flat spot (D). The higher the tip is, the more 'belly' the blade has. Indeed this parameter does not describe how exactly is the belly of the knife shaped, but it gives a first idea how much belly the knife has.

Profile of the Kato Workhorse 240 gyuto

Zoomed-in version of the figure above.

Personalizing the profile of the cutting edge:

  • If one designs a knife for a certain customer, then one may try to tune the profile of the cutting edge to the customer's preferences and also external conditions. Here the relative height of the person and the height of the cutting surface play the most relevant role. A tall person working at a relatively low placed cutting surface may prefer blade with a more pronounced belly, while a smaller person working at higher placed cutting surface may prefer flatter blade profile. This is just a 'food for thought' and really needs to be discussed with the customer.
  • Then of course comes the preferred cutting style. Someone who makes a lot of push cutting or cutting where most work is done close to the heel while the tip more-less rests on a cutting board, a too flat blade profile may make it too easy to 'stab' the cutting board and damage the tip. On the other hand someone who make a lot of pull cutting may prefer flatter blade profile, and so on.

The angle between the handle and the 'flat spot' 

This sounds strange at first but is actually pretty important. On my first knife I made the tang (handle) and the flat spot practically parallel and you notice it the moment you start to use the knife. It just does not feel right on a gyuto.

In general there should be a slight angle between the flat spot and the tang. This has direct impact on the profile of the spine and thus also on the weight distribution and position of the centre of mass of the knife.

In the drawing of the profile of the Kato Workhorse one can see that the slope of the spine of the knife is about 5 mm (height) over 90 mm (length) relative to the flat spot. Since I use paper with 5mm  dotted grit I am using this way to describe the slope of the spine, rather than the angle in degrees or radians.

Note: The axis of the tang should be parallel with the spine at the heel. This is not well represented in the figure above.

Height of the blade at the heel

This one is seemingly simple - if you want a taller knife you just, well, make it taller. But it also means you are adding steel at thickest part of the blade - the spine. So if you would like to keep the weight in check and not changing the grind behind the edge, than you may need to think how to adjust the blade grind higher up the blade or even making the knife a little thinner.

In general most users I have talked to seem to prefer 240 gyutos to be 50+ mm tall, so in my book a 'normal' height of a 240 mm gyuto would be around 51 - 52 mm, give or take.

In general - the height changes the way the knife handles rather significantly. While a taller blade (say 55 mm) gives you more room, it may quickly start to feel less 'precise' - in particular to a less experienced user than a blade 48 - 50 mm tall.

Note: as a general rule of thumb - for knives with a WA handle design one needs somewhere between 35 - 40 mm minimum height at the heel to allow enough clearance for you knuckles when cutting on cutting board, though the angle between the tang and the flat makes a difference too. 

The grind / geometry of the blade

This is in my view the most important ingredient of a well performing kitchen knife. No, a Full Flat Grind (FFG) is literally not going to cut it if you want to make a knife without obvious deficiencies in performance. I have described two main types of grind in the section 2, let's have a look at some examples.

I can not help but notice that in particular knifemakers that are coming to kitchen knives from general outdoor knives tend to make FFG knives. While with an outdoor or hunting knife such a grind make make sense, with a kitchen knife it is not the optimal blade geometry.

So what is wrong with a FFG than?

  • Because of a very low angle of the primary grind which goes all the way down from spine to the cutting edge, the knife requires often larger edge bevel than optimal, as otherwise the edge would be way too thin and weak and would get damaged easily (rolled or chipped).
  • Should the user attempt to maintain the blade geometry over time, he or she would have to regularly grind (sharpen) the whole surface of the blade, because otherwise the blade would quickly become too thick behind the edge and the knife will give more and more resistance of cutting hard vegetables (or even onions).
  • Ingredients love to stick to the flat bevels what disturbs during cutting. Large contact area also means larger drag effect when slicing - this can be very noticeable with raw meat.

The following figure shows schematically a FFG. This grind on a kitchen knife is very thin - with a blade 3 mm thick at the spine and 50 mm tall would result in a bevel angle of only 1.72 degree per side or an inclusive angle of 3.44 degrees (Fig. A in green). Such an edge would be very thin and easy to damage. Normal edge thickness (thickness of the blade just behind the edge bevel - shown in red in the photo, the edge bevel would be produced with sharpening the blade at some 15 degrees per side) would be around  0.15 mm. On a blade with a said geometry this thickness is 2.5 mm from the 'geometrical edge' (Fig. B).

The Fig. B shows the problem the user will have to deal with over time. (1) shows edge on a new knife. As the knife will be sharpened over time, the edge will move up (2) and necessarily will become thicker and the knife will cut with more and more resistance. If the user wishes to maintain the knife geometry, than the only solution is to sharpen (thin) the whole blade (see the dotted line) up to the spine.

Not to scale: cross section of a FFG and zoom of the tip.

The Fig. C shows a cross section of a blade with a convex grind. Here I used each side of a blade to make a different example, I did not mean an asymmetrically ground blade.

The left side of the blade is meant to show a grind where there is a 'shoulder' in the grind (one would find this usually around 15 mm from the edge). Here the (1) in red would stand for a convexly ground bevel and then continuously going to (2) in green which might be convex, flat or even gently concave. In particular the concave grind higher up the blade (2) is sometimes found on asymmetrically ground Japanese kitchen knives (I personally worked on a Hide gyuto - see my Project #11).

The right side of the blade (3) is intended to show a blade with a continuous convex grind from the edge up to the spine. It is important to say that the curvature of the grind of course changes with the distance from the edge - it is NOT a single radius grind.

Not to scale: A convex ground blade
An example of a continuous convex grind is Kato Workhorse by Yoshikaki Fujiwara. This is a hefty gyuto with a 240 mm cutting edge length weights ca. 250g. This also means that it is thicker and the convex shape of the grind is more pronounced. I took a short video trying to show its grind.

Now we get to the wide bevel grind. In Fig. D one can see a sketch of a wide bevel grind. In general the main recognition feature is the shinogi line - in the photo below it would be where the green and red lines meet (at the dotted line). But wide bevel grind is much more than just a combination of 2 flat grinds. The (1) in red can be on a new knife convex, flat or even concave. As mentioned in the section 2, the concave grind is more of a result of the production process than the intended final geometry of the wide bevel. Over time this should be turned by the user to a hamaguri bevel - which would be best described as a very gentle convex grind (with most of the convexing happening first few millimetres behind the cutting edge).

The Fig. E shows a closer view of the wide bevel grind. As the knife is being sharpened (either by applying a micro-bevel at the edge or by convexing the cutting edge - none of these is shown in the figure) the knife will also need regular maintenance and to keep the geometry of the wide bevel the whole bevel will need sharpening. This is in a sense similar to what was mentioned at the FFG, but here the area that needs to be sharpened (thinned) is considerably smaller. One one wants to achieve is - as the cutting edge move upwards, so should also the shinogi line. This is represented by the red lines and small arrows.

Not to scale: wide bevel grind example
Here is a short video showing a nakiri with a wide bevel grind. I did some work on the wide bevel which as of now is practically 100% flat (which in spite of being fairly thin causes some issues when cutting that are similar to a FFG), but I will be giving it a hamaguri grind soon.

But how to get the grind (in particular close to the edge) right? You need to test it while making it. Once you got the blade ground close to a zero edge thickness just put a quick edge on it with a coarse stone and cut come potatoes, carrots and onions to get an idea where you stand. If the performance is not as intended try to figure out where is the problem - is the blade too thick behind the edge (will not want to even start the cut), or is the blade too thick higher up the edge (will wedge), or is the geometry of the grind too flat (sticking & resistance/drag) or is the convex too steep behind the edge (edge will feel thick and cause wedging).

Monosteel & san-mai 

There are many Japanese double bevel knives that are made from single piece of steel. Most of these knives are either rather thin blades (sometimes called 'lasers'), or honyaki knives (differentially hardened knives with visible hamon line - pinnacle of Japanese knifemaking). However most double bevel Japanese kitchen knives have 3-layered construction like san-mai or warikomi since this has several advantages for the maker and for the user.

Note: I am going to loosely use the term ´'san-mai' as a synonym for multi layered blade construction like san-mai or warikomi, where a hard steel in the center of the blade is clad on both sides with iron, soft steel or damascus.

From the maker's point of view there is indeed the need to master the skill of weld-forging 3 pieces of steel together (otgherwise delamination would destroy the knife later in the process) and of course to forge a blade out of it. However once the bladesmith has mastered that skill forging of a san-mai blade can actually be even faster than from a single piece of steel, because most of the blade will be formed by a soft iron or steel that is easier to forge. On top of that this blade construction allows the blade to be straightened also after heat treatment without risking damage to the blade. After HT the grinding of the blade is also easier than with a monosteel knife, as most of the material that will be removed in the process is again not hardened steel.

From the user perspective san-mai construction makes the knife less prone to larger damage or breaking in spite of the fact that the cutting core has 60+ HRC. Once the knife needs thinning as a part of geometry maintenance - this can be done more easily as again, most the material that will be removed in the process is the soft cladding.

However I assume that very most readers of this article are making knives via stock removal method from a single piece of steel and I think it is reasonable to consider the implications for the user. The main point here is to think of the knife maintenance and thinning over time - in particular if your choice of steel is some of the modern stainless steel with high resistance to wear and abrasion. In the first place the knife should be made thin behind edge - the user should not have the need to thin the knife out of the box. The next important point is that the knife should be reasonably easy to maintain (thin).

This is just something one may want to keep in mind when considering how to grind a blade, but not necessarily something that should be central to how the blade will be ground, since performance of the blade should be the primary objective.

Grind close to the cutting edge 

Especially at the beginning one may feel unsure about how thin the blade is to be ground and how much material needs to be 'spared' for subsequent surface finish (scratch removal) and one may easily end up with a blade that is too thick first few millimetres behind the edge.

Example: Think of an edge that on one hand has been brought to the intended thickness of say 1.0 mm @ 10 mm behind the edge, but is still 0.7 mm thick @ 5 mm behind the edge and 0.3-0.4 mm thick just right behind the sharpening (micro) bevel (or about 1 mm behind the edge). If you run the blade between your fingers from the spine to the cutting edge you will feel a sudden 'jump' very close to the edge. In use such a blade will give a lot of resistance when cutting anything dense. Do not get fooled by the fact that the edge will shave or push cut paper - that is just a testament to your sharpening skills, but it tells nothing about the edge geometry.

Once the knife will need thinning, the grind higher up the edge should ideally not require too much metal removal. So for example knife with a thick spine or very wide bevel would require more work to thin regularly than a thinner knife and/or a knife where the wide bevel is not quite as wide. The best example here that I can think of are knives by Robin Dalmann who makes a rather narrow wide bevel and then a gentle concave grind above and as a result maintenance of the blade only requires to sharpen a relatively small area behind the edge.

Hamaguri grind

I have mentioned above the hamaguri edge, but did not explain how to create it. Imagine that you have made a wide bevel grind that is 15 mm wide and you are in the stage where you brought the edge thickness close to the zero (say 0.2 - 0.3 mm) and at the moment the whole wide bevel is still perfectly flat as it came off the grinder.

What you want to do is to grind the first 3 - 5 mm behind the edge with a slightly larger angle, so that your wide bevel and then blend these two grinds together. You will end up with a wide bevel grind that is practically flat ~5 - 15 mm behind the edge and a gentle convex the first ~5 mm.

Neck design

Well, 'neck' is not exactly a proper term, but I will just stick to it :) I am using it to describe the part of the tang that is exposed when the handle is mounted. Or in other words it is the narrow part of the blade between the handle and the heel.

One possibly minor point is the design of the neck of the blade. On a hidden tang knife there are in general 2 ways to do this. the height of the tang is (near) the same as the height as the handle. That is the case on the Kato Workhorse in the photo above as well as the custom gyuto by Bryan below. This means that before the handle is mounted the tang has 'shoulders' which define where the handle will be mounted. There is one more option here - often on Japanese knives the handle is not mounted all the way to the 'shoulders' of the tang and a short (2-3 mm) part of the tang is exposed. This is then called 'machi'. 

The other option is that the tang has no 'shoulders'. The main practical difference is that the handle sits a few millimetres higher relative to the spine of the knife and thus making the effective blade height a little larger. This is not really relevant on a 50 mm tall gyuto, but would be on a small knife with less blade height that should still be tall enough to clear knuckles on a cutting board.

Left: custom 195 gyuto by Bryan Raquin
Right: 165 nakiri by Masakage 

Taking a closer look

Handle type & design

WA handle is a fairly easy to make (I show the details of my approach in several other project related articles), but when designing one it should be seen as a part of the knife, rather than a separate component. For me as a maker the large advantage of a WA (basically a hidden tang) handle is that I can make it separately until it is 100% finished before mounting in on the knife. This allows me to test it in hand (given that blade is fully ground and thus is shape and weight is final) and should I make a fatal mistake while making the handle I do not have to figure out how to remove if from the blade without damaging the finish of the blade.

The relevant features of a WA handle are:
  • The handle is taller than wider
  • The handle is both taller and wider at the end than at the ferrule. In other words the handle tapers in both vertical and horizontal direction along the length of the handle. This improves the grip considerably over a handle without any taper.
  • One may want to apply additional taper under larger angle on the sides of the ferrule - this will make the front of the handle narrower and allow it to sit more comfortably in your hand when using a pinch grip, while the rest of the handle retains enough thickness.
  • Length of the handle can be used to adjust the weight of the handle and thus the position of the centre of mass, but both too short and too long handle is going to look strange on a gyuto.
Check out the article on the Project #25 for an example.

When it comes to materials and design I would recommend to start with a simple handle made of 2 materials for ferrule and body of the handle. Add a spacer if you feel the 'need', but avoid metal spacers at the beginning. You can use natural or stabilized woods, buffalo horn, micarta, etc. In general natural woods tend to be lighter than stabilized ones. I would recommend not to get 'scared' by using non stabilized woods. There is great variety of them and apart from very soft woods are mostly suitable for handle. It may make sense to use either a more dense and oily wood for the ferrule (which will have in general more contact with the food and its juices) - african blackwood is a great example. I sometimes use stabilized wood for the ferrule and natural for the rest of the handle. Water buffalo horn (comes in rolls and is usually black) is also a great material for the ferrule.

When choosing the handle materials do not only think about the aesthetics, but in the first place about how heavy you want the handle to be and where the centre of mass of the knife should end up. Indeed - this is not easy to judge upfront and will require some experience. 

Note on metal spacers: These can look great on a handle and provide very strong optical separation, but once the whole handle is glued together it should be power-sanded very carefully. The metal heats up very quickly and if you are not careful it may and will cause the epoxy to over heat and the handle will fall apart. I have a first hand experience on more than 1 handle. Also - the thicker the sparer the higher the risk. I have read a tip where one knifemaker while power-sanding the handle dips it regularly to ice water (obviously only applicable with stabilized or very oily woods) to keep the epoxy safe. Also - working in parallel on 2-3 handle and rotating through them would allow one to cool down while the other is being worked on.

4. Putting it all together

A very experienced knife user and vendor often uses a term than a knife or its design should 'make sense'. With that he means that the basic parameters (as loosely discussed above) plus heat treatment should follow the same philosophy (or simply purpose) so the knife can be used to its full potential. What exactly that means can be best understood from first hand experience with Japanese kitchen knives.

Well made knife that suits the style of its owner largely 'disappears' in use - it will simply do its job without getting in the way and without requiring 'workarounds' to be used efficiently. One often hears the term 'prolongation of one's hand' - that is basically what I am trying to achieve (even if it is a bit of an over-statement)

Once you made a knife - how are you going to know if it is any good? The best you can do is try to find a few pro chefs who themselves use Japanese kitchen knives at work and ask them to test your knives. You will most likely get a very to-the-point feedback that will help you to improve. Of course - every user has different 'sweet spot' and it will be the same here. Not even a 'perfect' knife will please everyone, mainly because a perfect knives do no exist.

5. Summary

This brought me to the end of this lengthy article. I hope you found it interesting or maybe even helpful. I would very much appreciate your feedback, so please do not hesitate to leave a comment, question or a critique.

Saturday, January 13, 2018

Project #25 - 150 mm tall petty from 1.2442

This was one of the fastest projects I have finished till today. A friend of mine who got a Moritaka AS 150 mm petty just a few months ago managed to bend it at work (while cutting !) and was in a need of a stronger & stiffer petty.

Since we have some experiments with 1.2442 steel in mind for the near future, I offered him that I could make him a petty from this steel as I have not used it yet (though I bought it about a year ago).

From the design point of view the idea was simple - ca 35 mm tall, 150 mm on edge, WA handle.

The blade

As usually, I have started with the blade design. Here I have decided to use the profile (of the cutting edge only) of a Yoshikane SLD 150 petty, which has proven to be very practical. I have then adjusted the height as the Yoshi is only about 30 mm tall. 

This was the first time I have worked with the 1.2442 steel (not in production since quite a while), which bears some similarities (composition wise) with Japanese blue #1 steel. When cutting the steel with a hack saw one would notice that there are alloying elements present, but it was still easier to work with than for example Niolox, though not as easy as O1.

All of the pre-HT steps were 'standard' ones, so here just shortly:
  • cut out the blank from 2.7 mm thick stock with a hack saw (yes, I leave the tip flat until after HT)
  • finish the profile with a #40 belt
  • clean up to #220 belt
  • paint the edges blue with a permanent marker
  • scribe the center line on both edge and spine (including tang)
  • grind the bevels and distal taper - I went down to about 1 mm on edge to minimize the risk of warping during the HT - I have not worked with this steel before.
  • Finish the profile of the heel (much harder to do after HT)
  • Sent the blade off to HT to Achim Wirtz to be hardened to HRC63

Blade after HT.

After HT I would first thin the the grind down to last 10% (leaving about 0.3 mm edge thickness) with the #40 belt.

Grinding after HT with #40 belt
Distal taper - the tang will be worked on a little later.

Close to final geometry

At this stage I would go briefly to 300 water stone from JNS to check and clean up the bevel grind to make sure that there are no defects like over-grinds and low spots. I was not trying to get the final bevel grind - really just checking the consistency. It all looked pretty good, just at usually the heel was a little thinner last 2-3 millimetres.  I was not trying to get 'perfect' bevels at this stage or even shinogi line.

After that I would proceed with shaping the tip and tang of the blade.

Shaping the profile of the tip (after working the bevel on 300 stone)

Once all the coarse grinding was done (blade geometry, tip shape, tang shape) the next step was scratch removal. During this step I would not worry too much about that bevel I did on stones as I was not trying to get as little as possible material removed. I would proceed with #60 and #120 ceramic belts and then proceeded to 3M Trizact 'gator' A100 belt. The A100 would be around #200 grit and frankly I would have been preferable to have the coarser A160, but it appears not available in 1x30" size. 

Grinding with #60 belt under different angle to remove all #40 scratches.
After the A100 I went  to A65 - always slightly changing directions. At this point I realized, that I forgot to round the choil and spine (ideal point to do that would have been after #120 belt), so I first coarsely rounded the spine with a belt and finished to #1000 grit by hand. The choil was rounded entirely by hand and then also polished to #1000. Once these parts were done I continued with #65 and #45 Trizact belts.

Rounding the choil with #240 grit. The light accentuated the scratches,
they were far less scary :)

At this stage of finish I would normally switch to hand sanding going to whatever grit would look nice (I am normally by semi-matte finish). But I was under a time pressure as I wanted to present the knife to my friend (who did not expect to see it anytime soon). So I took a shortcut. I decided to make  an experimental finish. I had a second session on the stones - this time a progression of 300, 1000 and 4000 and ground the bevel down to is final edge thickness (not more than 0.1 mm before sharpening).

I followed by finishing the blade with A30 belt making sure all scratches are gone. I tried to avoid the bevel as I did not intend to remove any larger amount of material there, just final finishing before sharpening once the knife was completed.

In the process of last scratch removal with A30 belt.

The last step on the belt grinder was made with a fine (blue) Scotchbrite belt. There I just wanted to get the finish a little smoother (the belt is nearly too sharp for that). I knowingly also toughed the the bevels so that I had an even finish on the 'shinogi' line.

Finish after fine Scotchbrite belt. Again - the direct light makes the scratches
to stand out more than they would appear under softer light.
The same Scotchbrite belt finish under different angle

The blade would not be finished without a logo ;) On this blade I only did deep etching with DC without using AC afterward to blacken the logo - as was wished by my friend. I have to admit I like this less obtrusive and will probably stick to it with most of my future knives.

About to etch the logo.
As always until now I have used the Electrolyte 94 as electrolyte, 12V setting on the etcher and etched for 6x3 seconds (e.i. lifting the etching head every 3 seconds so that the gas that is produced in the process can escape and does not cause uneven etching

The handle

While waiting for the blade to come back from the HT, I would start to work on the handle. The design was again a 'standard' one using buffalo horn for the ferrule and stabilized & dyed Australian wood (curly spalted Boxelder) with a beech dowel inside. Since this handle was a rather subtle one, I decided that apart from cutting the ferrule and handle block to rough shape with a band saw, I will do all the shaping & sanding by hand on a stone sanding block.

Handle materials

The basic process of the handle making:
  • cut the & square up a piece of buffalo horn and handle block
  • drill the ferrule and handle block
  • check that the handle block axis (defined by the drilled hole)aligns with the ferrule & glue them (using my own gluing press that allows using a piece of dowel material for initial alignment)
  • Once the glue cured clean & square up the future handle

Handle components cuts and squared-up

Gluing the handle using a jig. The dowel is
removed once the handle is clamped down.

Normally I would glue in the devel before I would glue the handle parts together, but this time it took longer to get the blade back from the HT and I did not want to loose too much time, so I glued-in the dowel at a later stage (when the blade was back and I could test the fit).

Handle squared-up after gluing and ready for shaping

Normally I do the rough shaping (= final height and width including distal taper) with a coarse belt, but I decided to do this completely by hand. I would do this on a #80 grind sandpaper taped flat on a stone (old table-top)

Handle after coarse shaping
Once the desired height and width (with about 1 mm in every direction left for further sanding) was achieved, I continued through 180, 320 and 600 grits.

Coarse shaped handle refined to about #600 grit.

In the next step I have sanded the handle to octagonal shape. Again - this is something I usually do with the little disc sander that is part of my belt sander, but I simply wanted to do it by hand. This is an interesting experience and allows to get the feel of how important is even pressure distribution.

Once the handle was shaped and finished to #600 grit the blade finally arrived from the HT and so I could make the fitting dowel for it.

Handle shaped and finished up to about #600

I have to admit that since I got the little Record Power BS250 band saw, the cutting of dowels got so much easier. Once the slit was cut it gets widened on a belt sander (#80 or #120 belt seems to work the best) until I get a nice, easy with with the tang and the handle

Dowel cut, now the slit will be widened with a belt sander.
As you may see in the photo below - I tend to make the dowel some 5 mm longer that the drilled opening in the handle - the reason is, that the front side of the dowel might have suffered uneven grinding with the belt sander, or it might get sanded by a rasp after gluing. So I rather make it a little longer and then sand to length after gluing.

Checking the fir.

Once the dowel fits nicely I glue it inside the handle with epoxy. I tend to create small 'pockets' (dips) on the side of the dowel with a round file so that if there is a bit too much epoxy it can go there instead inside the slit itself. You really only want to use just a little epoxy at this stage, because if too much of it leaks inside the handle you will spend a lot of time with needle files and rasps trying to get it out.

As always I use a piece of scrap steel shaped like a very thin edge to push the dowel open while the glue cures, so there will be no gaps between the dowel and the ferrule.

Gluing the dowel

Once the handle was finished this far I tested a little more to see how it feels in hand (with the blade inserted) and got the impression that while I like the overall size and shape, it would feel more nimble if it were a little narrower in the front. I have have covered most of the handle with tape (remember, it was already sanded to #600) and used #180 sandpaper to give a bit of horizontal distal taper to the ferrule.

About to start sanding the distal taper.
Thanks to the fact that the handle already had a #600 finish, it was very easy to observe where the material was being removed. I started by tilting the end of the handle upwards very slightly and applying pressure only on the very front of the handle. I would stop after every few passes and check the progress. Once I have established a new bevel I would continue until it reached the end of the ferrule. I would constantly check and if necessary compensate any unwanted tilts.

After first few passes - looking good so far.

The horizontal distal taper sanded in.

Once the taper was sanded the whole handle was finished up to #1000 grit and the front and rear edges were chamfered. I also used steel wool when finishing the handle, but it was not a particularly good idea with this wood which has a somewhat coarse and not homogeneous structure and the steel wool would tear little pieces out. 

Front side of the handle sanded to #1000 and the opening lightly filed
so that the shoulders of the blade would fit nicely.

Because of the slight tear-out I have the handle once coat of Birchwood Sealer & Filler, sanded it lightly afterward and followed with 2 or 3 coats of Tru-Oil for a good measure. As always, I would apply the Tru-Oil and while & polish the handle with a paper towel immediately after the application.

Tru-Oil curing after application.

At this stage both the handle and the blade were finished (the blade will still get slight refinishing of the bevels on stones) and thus ready for gluing.

Well, let's do the last step.
As with the last few handles I would heat up the G/flex epoxy to make it easier to get it inside the narrow tang slot, then heat up the tang of the blade and insert it carefully. Once the epoxy started to come out I wold let the blade to sink in the handle slowly while wiping the excess epoxy. Since I use G/flex which has 45 minutes pot time I did not have to hurry and even had time to take some photos :) Once the blade was in place I would heat the top of handle gently with the hot air gun to allow most of the bubbles to escape.

I have used small pieces of paper towel wound round ice stick and dipped in Toluene to clean up the blade and ferrule from epoxy rests.

The handle is in place, the last bit of epoxy is still to be cleaned off.
Gluing the handle this way also means, than more often then not the epoxy will shrink a little while cooling down (some air bubbles always remain trapped inside) and thus a shallow void between the dowel and the blade would appear. This is is filled with epoxy (Blade Bond Ultra that has pot time 15 minutes) the next day in a similar manner - the mixed epoxy is heated up and just very little of it is applied to the contact area between the blade and the dowel. Heating the applied epoxy with the hot air gut would allow it to fill the narrow void and the bubbles to escape. This time however I would wait until the glue starts to set it (get thicker) and only then I would use clean the rests off (paper towel, ice stick and Toluene). If I were to do it too soon, I would risk 'sucking' out the epoxy from the freshly filled void. The reason to use different epoxy is that I do not want to wait for an hour for the epoxy to start to set it :)

Gap filled with the epoxy, the excess was already removed.
After last clean-up of the bevels on the stones the knife was officially finished.

Finished knife

Lessons learned

Not so much went (thankfully) wrong this time
  • 1.2442 steel was nice to work with (in comparison to Niolox every steel feels nice to work with though) 
  • The "narrow wide bevel" grind appears to work very well (feedback from my friend)
  • The gentle horizontal distal taper improved the feel in hand considerably
  • I am very happy how the belt finish turned out, but I have a lot of room for improvement (the finish and the efficiency) there. I will need to experiment with different belts to get the most out of it. I plan to make this simpler but very functional finish in the future more often (not only on Niolox)
  • I may be doing more knives like this one in the future. In general I will probably start to make small batches of knives once I get a knife that I (and in particular my test users) like.
  • Again - I was in a hurry to finish this knife and so I did not make proper photos or a video. I need to improve on that.

Sunday, December 10, 2017

Project #22 - Thinning & refinishing & rehandling a Takeda AS 170 Santoku

This is a tool-related project. I was asked whether I could thin, refinish and re-handle a 170 mm Takeda Aogami Super santoku, which, as many of Takeda knives made in the recent years, have the grind a bit on the thick side (and hit-and-miss forging work as it later turned out) and profit greatly from thinning (pushing up the shinogi line).

This kind of blade work was my first and so it took me a lot of time and I did not manage to avoid making some mistakes as you can read below.

The knife in original condition.
Detail of the original grind

Thinning jig

The main part of this task was to move the shinogi line higher up the blade face and thus make the primary (flat) grind more acute and then put some decent finish on the blade.

To do this kind of job requires either a very stable & experienced hand (none of my hands fits that description), or a jig of some sort that would allow to make the most of the material removal in a well controlled manner. I was considering to do this free hand, but Jon Broida pointed out that I could actually (after some work) use my bevel grinding jig which does something very similar, but with a file (check out my article on this topic)

To make re-adjust the jig for this kind of work two main parts need to  be made:
  1. A way to attach (in a secure manner) the Atoma 140 diamond plate to the guiding rod of the grinding jig
  2. A holder or base to be able to securely fix the knife that it would hold the blade in a well defined position AND keep the blade from being scratched or damaged accidentally. In particular the cutting edge would need to be protected from accidental 'bzumping into' as that could leave some serious chips in the high HRC and thin blade.
To make the attachment of the Atoma I used a piece of wood and a double sided Tesa tape (so that I would be able to attach and remove the Atoma without making some complicated clamps.

Atoma holder for the grinding jig - top view ...

... and the bottom view

Attaching the blade for thinning was a bit less straight forward. I was considering different complicated solutions (including special clamps or Nd Magnets), but I finally went with the simplest option and I have used a double sided tape to attach the blade to a pre-shaped piece of reinforced multi-ply board.

The frame of the thinning jig, and an additional board with
the tape already attached.

Since the kurouchi on the Takeda is not particularly strong some of it came off with the tape later, but still - this was probably the best solution aside from making a custom knife holder with neodymium magnets (which I may consider for the future). To mitigate this problem somewhat I applied a orange masking Tesa tape to the blade first before attaching it to the double sided tape (which was much more 'sticky'). The orange Tesa holds pretty well and as  experience shows it can be removed without leaving any residue.

Of course I removed the original handle as it would be getting in the way. This is a great advantage of the traditional handle mounting in Japan. Once the new handle will be installed, the original can be used for a different knife (it is a fine handle, after all).

Orange Tesa for a blade protection.

I did not use any edge-damage-protection that would prevent making the stroke with Atoma too long, jumping off the blade and inevitably bumping into the cutting edge with the front edge of the plate. Simple solution would have been to put a clamp on the guiding rod so that it would not be possible to jump off the blade, but that would have prevented the rather convenient cleaning of the Atoma with a water spray (for which the whole Atome/rod part needs to be taken off the rest of the jig. But since there were plastic guides on the front and back side of the Atoma holder, the very occasional contact with the cutting edge did not leave any damage, but I do plan to upgrade this in the future.

Refinishing the Takeda

Once the jig was finished it was time to put it to some good use. Before starting with the thinning I would paint the edge with blue marker to be able to see the progress. I wanted to avoid grinding past the cutting edge and since the knife was sharp with basically zero edge thickness, I had very little room for error.

Blade is attached to the jig, edge painted. The fun can start.

I used water spray to keep the blade wet - the excess water would drop to a bucket I have put underneath.

I did not calculate the grinding angle - at angles this small that would not be precise enough as it is hard to place the blade at a well defined angle relative to the jig (the tape would give a little, etc.), so I just first checked what would be the angle of the original grind and then go down a little. 

I would use back and forth movements while grinding, but I would always move along the edge during each stroke to minimize the risk of uneaven results.

After a few minutes of careful grinding.
Since I used Atoma 140 diamond plate for the work (the fastest metal removing solution without going to some powered tool) it was necessary not to apply too much pressure, as that is something that can shorten the usable life of the plate considerably. Great help in this regards was the fact, that the Atoma attached in the holder together with the metal guiding rod had its own weight, so I was just moving it up and down without applying any additional pressure. 

When getting close to finishing one side I would leave a very narrow stripe of blue (= not ground) of the blade just behind the edge. The reason for that that some more material will be removed later when finishing the blade on stones and I did not want to waste the blade height for that. It would also leave me a little room for corrections.

Grinding with Atoma finished with a narrow strip of blue paint
still present.

Using a grinding jig like this gave me a very good consistency of the grinding angle along the blade. The side effect was, that I could not try to compensate for imperfections in the forging work. In particular the area close to the tip on the right side was a little thicker what pushed the shinogi further up the blade.

Shinogi like at the tip went up as the tip was a little
thicker than the rest of the blade.

As I mentioned I did some trial-and-error tuning of the grinding angle when working on the first (right) side of the blade. Once I switched to the left side I kept the very same angle to keep the blade grind as symmetrical as possible (within the precision of the whole jig). 

Once the right side was finished I would measure the width of the bevel and painted a line of the left side - to have an estimate where the shinogi line should be once the left side will be finished. Should the actual grinding angle not be the same as on the right side, I would see that I would either approach the designed shinogi too soon (then the angle would be too small), or I would reach the cutting edge before the new shinogi would (than the grinding angle would be too large) and so I could make adjustments if necessary (though that was not the case).

Left side ready for the grinding, shinogi line marked on the blade.
I am happy to say that the left side was forged more evenly and the resulting bevel had a rather constant width.

Left side finished.
Look at those scratches, there is some more work to be done.

New handle

At first the handle was supposed to be just a simple dark-ish wood with some light color ferrule, but then the taste changed and the request came to add some color. So we finally agreed om the following construction:
  • reindeer horn ferrule (I got a nice piece from Robin Dalmann)
  • blue dyed maple spacer
  • Turkish walnut body of the handle (lovely wood - nice to work with and rather lightweight)
  • blue dyed maple end cap.
  • white fibre paper spacers

Materials (minus the butt cap)
The construction of the handle was going to be a 'standard' one with the addition of the piece of dyed maple on the butt of the handle. So I dowel would go through most of the length of the handle.

As usually I would start with squaring up the handle block so that I would be able to drill it in a well defined way (and have the top of the handle as square to the drilled hole as possible. Then cut ca. 30mm long piece of the horn and drilled it too. Since the dyed maple spacers were rather thin, I have clamped them to a piece of wood for the drilling, to minimize the risk of them breaking. That worked well.

Drilling the handle block. I know these are boring shots, but somehow
they became part of my articles :)
Once all materials were prepared I did a dry run - to make sure they all fit together. It is important to make sure that once assembled the parts fit together well without any gaps. The gaps can be created by either not having contact surfaces perfectly flat, but also if they are not square to the common axis (defined by the drilled hole). If you look closely to the finished handle you will see that I indeed did get a tiny gap between the ferrule and the first spacer. It is filled with epoxy, so it is not a danger to the handle, but it should not be there.

Tip: Rarely you will get a perfect gapless fit of the ferule and body of the handle - in particular with horn (which is round and thus not easy to clamp well when drilling). However if you are not far off, then rotating the ferule relative to the handle block (while having a dowel inserted) may allow you to find a position in which these two parts match well. If I do that I then mark their relative position and I use it when gluing the handle.

Dry fit of the handle before gluing.

This handle was a bit special as it had also a butt cap. Since I did not come up with any simple and fast solution how to secure it better, I just glued it with an epoxy without any dowel. It should be fine as it is rather short, but ideally there would be some material connecting the body of the handle and the butt cap. I have used the G/Flex epoxy as it proved very strong and glues wood very well.

About to glue the butt cap to the handle block.

Once the first glue cured I could glue the whole handle minus the dowel which I glue in a separate stage later.

Gluing the handle
 Since recently I have started to use my slightly adjusted handle gluing press (which I originally built for gluing of through-tang knives like finish puukko knives with birch bark handles) for gluing this style of WA handles. It allows me to clamp the glued-up handle with a temporary dowel inserted. Once the handle is clamped down I then carefully remove the dowel. This has 2 advantages - it allows me to make sure that all components are aligned and - it allows me to remove most of the glue that would leak in the drilled hole (and thus make it later hard to fit in the dowel). Important is to use slow curing epoxy so that one has enough time to clamp the handle, make sure that every is at the right place and remove the dowel before the glue starts to cure.

Gluing the handle with the temporary dowel still in place.
Once the the glue is cured (the next day) I would first grind off excess glue and spacer material that sticks outside the dimensions of the handle block.

Glued handle before cleaning up.

Glued handle cleaned up and ready for the next step.
At the time of writing this article I have already added a small band saw that helps greatly with handle making to remove bulk material or to cut large blocks to smaller ones. But when I was making this handle I still used a belt grinder to size the block down close to its final dimensions. After that I would flatten all 4 sides by hand with a #80 sandpaper attached to a stone sanding block.

At that point I would scribe centre lines on all sides, mark the exact dimensions for the handle (including the distal taper) and remove the removing material on a belt grinder and flatten again.

Handle close to final size.
 Once that has been accomplished then it was time for the most 'sensitive' part - grinding the octagonal shape - something what I usually do with the small (125 mm in diameter) disc sander.

About to start putting octagonal shape on the handle. 
This part of the process make a lot of room for a mistake that may not be possible to correct, because once you ground too much (or under a wrong angle), than you may need to start over.

Of course - slight deviations from 'ideal' angles are OK (none of the normal Japanese handles are perfectly symmetrical), but I am trying to get some reasonable level of symmetry and consistency.

Handle with octagonal shape ground with #150 grit disc sander.
I made this handle slightly larger than the original and since it was a handle for a very lightweight blade, I have decided to grind in some additional taper from sides, to make the ferrule thinner when holding the knife in a pinch grip. I did that just by hand on the sanding block as that gave me most control. Once that was done I would proceed with finishing the handle on a sanding block with #180, #320, #600, and #1000 grit sandpaper and also chamfering the edges of the handle.

At this stage I asked the customer whether he liked the handle and once I got the green light I have made a dowel from beech wood, cut the slot with a band saw, then widen it with a #120 grit belt so that the blade would fit well and glued it inside the handle with an epoxy. Then it was time for the final surface finish of the handle.

Since the Turkish Walnut is not an oily wood I have applied a few thin coats of grain sealer (always allowing the coat to cure fully), then polished the handle and finished it with a few very thin layers of Tru-Oil using the 'Dalmann technique' (sorry Robin :) ) - that is applying a coat of Tru-Oil and immediately polishing the handle with paper Towel. This would mean than each coat would be very thin and the handle would not need any additional sanding or polishing.

About to apply the grain sealer.

Finished handle.

Finished handle - you can see the distal taper of the ferule.

Finished handle.
Before mounting (gluing) I would do the bulk of the finishing of the blade on stones. Remember - at this stage the blade has still that 'lovely' finish from Atoma 140 ... which is pretty awful and coarse.

Refinishing the blade

I have used the following stones in the process:
  • Bester 220
  • Gesshin 400
  • Gesshin 2000
  • Gesshin 4000
  • Ohira Uchigumori
  • Finger Stones from Uchigumori
I started on Bester 220. For its grit this stone does not wear that fast and it removes steel quickly. The main point here was to remove all the scratches left by the Atoma. Still - it is a muddy stone what on one hand makes it easier to leave a more even finish, on the other hand the thick mud may get inside low spots and 'make them disappear' so that you discover them at later stage ... and must go all way back to remove them :) 

After Bester 220 I moved to Gesshin 400. This is a very fast stone that leaves very even finish, but it also wears very fast and makes a lot of mud. It leaves a very nice finish,

Finish from Bester 220.
Finish from Gesshin 400.
Gesshin 2000 is mediun hard stone (though still creates some mud). It is very fast and removed the scratches from the 2000 stone rather quickly. The Gesshin 4000 is even harder. 

Feeling under pressure I quickly moved the the Uchigumori and at that stage it was starting to become obvious that I have overlooked (more like ignored) a few low spots. from the scratch pattern (nearly perpendicular to the cutting edge) it was obvious that they are from the Atoma. One can also see in the above photo that the low spots go down to the cutting edge and actually a slight 'dips' developed as I was working on the finish. Oh great.

Two low spots well visible - still showing the scratch pattern from the Atoma 140.
It meant that I needed to go all way down and redo all the finishing. I decided to go to Gesshin 400 for this job, but first I would put the knife under 90 degree to the stone surface (like I would want to cut the stone) - fix the blade profile so the dips would disappear (the knife lost some 0.5 mm of the blade height in the process). To make a low spot disappear one needs to grind all the remaining areas until the low spots are not low spots anymore and then keep grinding the whole bevel until the scratches from the low spots disappear. and then refinish the blade. And that is exactly what I did and it took me additional 2 hours of careful work.

Once I got the bevels fixed and finished up to Gesshin 4000 I decided to mount the handle and make the very last steps of finishing and sharpen the knife afterwards.

Knife ready for the handle to be mounted.
Just before I would start to glue the handle I would give it a coat of board butter, to minimize the impact of wiping some parts of it with Toluene later when cleaning it from the epoxy.

Since recently I have started to mount (glue) the handle in 2 steps. First I do the gluing itself - I mix the glue, heat it up with a hot air gun so it becomes less viscous and pour it inside the handle. Than I warm up the tang and slowly insert the tang - trying to allow the air to escape. After that I make sure I remove all the excess glue from the blade and the handle (I use paper towel and Toluene for that) and then I allow the glue to cure.

During curing (mostly because temperature related volume change) the epoxy would slightly contract and thus empty voids would appear next to the blade at the top of the ferule.

The next day I would therefore mix a different - somewhat faster curing (15 minutes working time) epoxy to fill the voids. I would again use the hot air gun to make sure the epoxy fills the voids nicely and that air bubbles do not remain trapped in the epoxy. And I would wait until the epoxy would start to get a little thicker as it starts to cure. At this stage I would use small pieces of paper towel wrapped around the wooden ice cream sticks dipped in Toluene and wipe the excess epoxy away. The reason for letting the glue get a little thicker is to avoid 'pulling' it out of the voids I am just trying to fill. Once the glue gets thicker this does not happen anymore.

Gluing the handle.

Allowing the glue to cure. Yes, I need
 a better solution to keep the blade upright.

Filling-in the voids.

Once the handle was mounted I would give it again a coat of board butter for the good measure and then finish the bevel on the Uchigumori stone and finally even out the result with Uchigumori fingerstones. And then the knife was finally finished.

Finished knife

(photos taken before final sharpening of the cutting edge)

Here a tiny gap can be seen between the ferrule and the spacer.

Lessons learned:

Lots of stuff :)
  • Be very careful when grinding close to the heel or to the tip - if you do (even only just partially) grinding movements that are perpendicular to the cutting edge (instead of sweepping movements under angle), than you have a good chance to grind in low spots. If these are shallow enough, you will not notice them until later when finishing the bevel on stones. You do not want that to happen.
  • Do not grind the bevel down to 'zero grind' (or 0 thickness at the edge). Once you will go to sharpening stones for finish you will need some more material to be removed before the scratches are gone PLUS since it is not really possible to keep perfectly the same angle and thus will start to loose material at the very edge what will create uneven profile (which you will have to correct what will cost you blade height and work) PLUS you will experiene pieces of the cutting edge chipping out (in particular on coarse stones) - these will get caught in the mud and cause deep scratches in the cladding generating further work.
  • IF you do get to the point that you discover low spots during bevel finishing (they will become visible latest on 400 grit stone) than you have 2 options:
    1. If you do not have dips in the cutting edge yet (at those low spots), than keep working the bevel, but avoid material removal at the low spots. Continue at the coarse stone until the rest of the blade will match the low spots.
    2. If you (like what happened to me above) already have 'dips' in the edge profile appearing, than take the knife at 90 degrees relative to the stone, grind the edge until the profile is restored (thus loosing some blade height) and then continue like in the point (1)
  • Do not hurry when finishing blade (or handle) as even small mistakes can set you back and you will be forced to re-do many steps and loose much more time.