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.

Project #21 - Damascus paring knife

Note: This is a long one, get a nice cup of tee before you dig in :)

Another custom order - this time making the knife a little more fancy was part of the request. Also - the person (or should I better say the couple) ordering the knife were inspired by the parers made by Del Ealy (who's work you definitely check out) and handles by Oatley Knives. And while I was not attempting to copy any of the two makers (nor would I be able to), some traces of their designs can be seen in this work and I wanted to state that clearly.

The idea

The knife was supposed to have the following features:

• Stainless damascus blade of ca 100 mm length, 
• Colourful, D-shaped handle with tapering

Materials

Once the selection process concluded, the following materials were selected:

• Swedish Damasteel 
• Stabilised blond ebony ferrule (X-cut)
• Stabilised & dyed Maple Burl (both from Stabwoodlab)
• 'Some' sort of spacer (later I decided to go with a thin white G10)

Stabilized blond ebony and stabilized & dyed maple burl.

Damasteel stock. This little piece of steel was nearly 90€

The blade

Design of the blade shape went in several steps - below are just last 3 designs. #12 was the selected one. Not shown in the sketch below is the hidden tang.

Since this blade was considerably different to what I have done before + I wanted to be sure that the proportions (in particular the width of the handle) will sit well in hand, I also did the same blade from Niolox steel (I will eventually finish that one too and gift it to my mom :) ) before I proceeded with the rather expensive Damasteel. The main difference being that the Damasteel stock was 3.2 mm thick, while the Niolox only 2 mm. I would have actually preferred the Damasteel to be thinner as for a knife this small, the 3.2 mm was simply overkill.

In the process of handle-making (since D-shaped handle was also something new for me) I have made 2 testing handles from beech wood. I used then one of these as working handle while grinding the blade.

 After cutting out the Niolox blade and testing how it feels in hand (with that testing handle) I then proceeded with the Damasteel.

Stock Niolox and the blade design.

Filing the shoulders with the help of a filing guide.

Above: Damasteel stock, Below: Niolox blank

Damasteel blank ready for bevel grinding.

Once the blanks were ready there was one more test to be done - etching of the logo. I did not know how the Damasteel will react to etching. Since the whole surface of the knife would be ground during the bevel grinding I could test-etch the logo without causing any harm to the knife.

I did etch tests - one just a surface (black) etch with only 10 x 3 seconds AC (12 V) and one including a 10 x 3 seconds deep etch first.

Making the surface etch.

As one could expect, the surface etch (left one in the photos below) was cleaner and sharper, but would be easier to scratch weaken with subsequent light sanding with steel wool. The deeper etched logo was not quite as sharp, but was much more scratch resistant.

Right after etching:
Left: Black etch only
Right: Deep etch + Black etch.

The same two logos as above, but after some steel wool polishing.
The deeper etched logo looks much better.

As usually I would clean up the edges on the blanks up to about 240 grit, then painted them blue with a permanent maker and scribing a centre line on both spine and edge.

Getting ready to scribe the centre lines on the blanks

 Once both blanks were prepared I would always first do some test-grinding on the Niolox and then proceed with the Damasteel.

First grinding of the Niolox blade. You can see the blued edge and centre line.

First grind looking rather consistent. I must be slowly improving :)

Once the bevels were set I have used the filing guide while grinding higher up the blade face. The reason were to eliminate the possible slipping and grinding-off the shoulders of the blade. It was indeed a bit awkward, so I did not use it after HT for further grinding.

Using a filing guide while grinding the bevels.

Before the blades would be sent off for HT I would grind them up to #120 grit. Since I was a bit paranoid about the Damasteel blade getting bent during the HT, I have left the edge about 1 mm thick.

Blades back from the HT. The 'black blade' is from 80CrV2 carbon steel
and was quenched in oil, so it is covered in scale.

After HT I would continue grinding with #36 grit belt. Once the edge thickness was at nearly 0.5 mm I switched to #60 belt and slightly changed the angle, made sure that all #36 scratches are gone, and went to #120 belt and repeated the process.

About to start grinding after HT.

In principle I could have continued with finer belts or hand sanding, but I have decided to finish the bevel setting and scratch removal on stones. Since Damasteel proved rather easy to grind and sand, I have moved directly to AI#1000 (that is probably the fastest 1000 grit stone there is). Doing so gave me a lot of control over where the blade would be ground and thinned and I would also make sure that there are not hidden low spots left. I would like to stress that at this stage I was not after an even finish, just a complete scratch removal from the previous step and fine tune the blade geometry.

Once I was tone with the #1000 stone I moved to Gesshin #4000 as a final stage of the work on stones. The Gesshin would leave me with seriously patchy finish (not a stone's fault, but it is not easy to get a smooth finish on a curved surface on a hard sharpening stone), but that did not matter since I the surface finishing was going to be made with high grit sandpaper.

Tip: The grit rating of sharpening stones and sandpaper is different. #1000 stone will give you finish somewhat comparable to #400 sandpaper. Trying to go from #1000 stone to #1000 or even higher sandpaper will not really work.

Gesshin 4000 - nearly milky white before you start.

4000 stone finish

Cayin N3 player, Music Two custom monitors from The Custom Art
and a load of freshly converted FLAC albums made the hand sanding
nearly a positive experience :)

Here you can see the rather uneven 4000 stone finish.

Before I continued with the finish of the blade I have rounded the spine (with a A65 ) and choil area and finished it up to #2000 grit paper. To finish the choil I have used a wooden stick (12 mm in diameter, the one I use to make dowels) as a support for the sandpaper.

After the stones the blade had final geometry and close to 0 edge thickness (sub 0.1 mm, some burr was already raised) I moved to hand sanding. I have started with #1200 sandpaper, but then moved one step back to #800 - and after a few minutes, once the blade started to clean up - I have discovered a few faint scratches from the #120 belt. I was not pleased.

To get a smooth finish on the convex grind of the blade I glued a piece of mouse-pad on a piece of wood.

There it is - a faint scratch from the #120 belt.

it turned out that I was lucky and I actually managed to remove those scratches with the #800 sandpaper. This would have not been possible with a Niolox blade.

Once I was done with the #800 sandpaper I moved to #1200, #2000 and finally #2500. The #2500 sandpaper gave me a semi-mirror finish - definitely smooth enough for etching.

One side finished up to #2500

This is what I call semi-mirror finish.

Once the first side was done I re-taped the blade and started to work on the other side. The tape provided near-perfect protection of the finished side.

Putting and orange Tesa masking tape on the blade to protect the
already finished surface.

Etching

Once both sides were finished it was time to etch the blade to reveal the damascus pattern - which was barely visible on the polished blade. After reading around I proceeded as follows:

• Clean the blade with acetone
• Clamp the blade in a lockable pliers (O forgot to drill a hole at the end of the tang)
• Insert the blade in a 37% solution of HCl for 5 minutes
• Once finished neutralize the reaction with a saturated solution of kitchen soda
• Was the blade with soap and water

Of course I did this on a balcony with full protection gear - respirator, full face shield and gloves that are made for handling chemicals. Better safe then sorry.

After etching one would normally sand the blade, but since the result looked really nice I skipped that. The 5 minute etch gave the blade a rather deep etch - next time I would use 2 or 3 minutes.

Etching of the logo was a simple and fast procedure - after cleaning the blade I would first etch the logo a little bit into depth with 6x3 seconds of 12V AC and then blacken it with 6x3 seconds of DC. Done :)

Important: One thing to watch out for when etching a logo with stencils as below, is to make sure, that the edges of the logo and taped well and that no etchant and no current will find its way around the edges, as that will leave black streaks or line around the logo.

Stencil with the smaller version of my logo taped in place

Freshly etched logo.

At this stage the blade was finished and the last step to make was to sharpen the knife - this was done after the handle was glued on.

The handle

Once the blanks were shaped I could start in parallel to work on the handles. As already mentioned, I have first made 2 prototype handles - the first was also drilled, the second only shaped. The main reason was that I was not sure how well will I be able to get the palm edge of the D shape ground and how the finish (sanding) will go.

Drilled beach block, ca 10 cm long

Coarse shaped block with slight distal taper both in height and width.

Rough shaped handle.

Checking how the handle sits in hand.

Once I had a rough idea how to proceed with the handle making I started as usually - squaring up all the handle components. The main reason for the squaring up is to be able to align the block while drilling and to get the best possible alignment of the handle block and the ferrule.

I usually start on #80 grit sandpaper taped on a piece of stone (originally part of a small serving table). Once I get the block roughly squared up, I continue with #180 sand paper. I do not go any higher - not even with the surfaces that will be glued - to get the best possible connection between the glued surfaces. If I am adding another material that has glossy surface (metal, G10, micarta), than I will roughen it up with #180 sandpaper before gluing.

Last visual check of the handle materials. I have poured some ethanol
over the blocks to allow the colours to stand out.

Machinist square is a must for me here.

I love the different colors of the sanded wood.

Once the squaring up is finished, the next step is drilling. I check how deep the tang is and add about 2 more cm for the dowel. This is a small knife, so I use a dowel with 10 mm diameter. Since this knife was going to have a 'hidden dowel handle', when drilling the ferrule I drilled first the centre hole with just a 2 mm drill, then drilled nearly the full depth with 10 mm drill (I left about 5-6 mm material). Then I drilled another two 2 mm holes that would align where the opening for the tang would later be.

Drilling the dowel opening on the handle block. I always enjoy
drilling the dyed woods :)

The handle materials. The ferrule has only the first 2 mm hole drilled.

Filing the tang opening on the ferrule (first connecting the three
2 mm holes with a needle file)

Making a dowel - first cutting the beech with a hack saw ...

... then widening the slot with a #80 belt until the tang would fit.
This would be checked with all the components assembled.

Nothing is glued here yet - just checking that all parts fit together before
I would start to glue the handle.

Once all the handle components were ready, I would first glue the dowel inside the handle block with a 5 minute epoxy (I did not want to wait to the next day to continue). I would use a piece of scrap steel (shaped like a very long and narrow triangle) to keep the slot in the dowel 'opened' while the epoxy cured. In the next step I would use G/flex epoxy to glue the remaining components into one piece.

Note: Of course - one could glue all the parts together - on the blade tang - in one step. But I still do not dare to shape the handle with the blade in place (unless there is no other way), plus in this case the blade would be getting in the way of shaping the handle. So I do it the round about way.

Once the glue cured I would cut off the excess wood (the block was wider than necessary for the handle, but I prefer to keep it that way up to gluing), grind it to shape and hand sand it on a sanding block up to #180 grit. I would also check whether the blade can be inserted. Often small bits of glue leak in the dowel slot and need to be removed with a flat needle file or rasp.

Cleaning and then squaring up the handle block after gluing. 

Before I would proceed with coarse shaping I would insert the blade and check for alignment. if the blade would not be aligned with the expected axis of the handle If the blade would not be in line with the expected axis of the handle (I know the middle point on the bottom side of the handle as I marked it with a tap before I started to drill) I would re-draw the axes to make sure that the further shaping happens around the axis of the blade.

Checking the blade alignment relative to the handle.

Cleaning up the block on #80 and #180 sandpaper.

At this stage the handle was ready for the coarse shaping - that means I would draw lines on all sides of the handle that would reflect (slightly oversized) future taper of the handle. In this case the taper was not supposed to be too pronounced. I would do the coarse shaping with a #40 belt and then clean up the blade again by hand with #80 and #180 sandpaper.

With the handle ground to tapering shape and with all unnecessary excess material removed I would start with final shaping. For that I use (for now) the 125 mm disc sander that is part of my 1x30" baby grinder. I have used a zirconium #150 grit disc. One could get the job done faster with #80 grit, but it also becomes easier for the disc sander to 'grab' the handle out of my hands and send it flying across my shop. Anyhow - I always use a full face shield when working with the disc sander, even though it is a tiny one.

Note: This sort of handle shaping could be done with more control on a universal (= capable of vertical and horizontal position) large grinder, but since I do not have that option for the moment I am  trying to get most from what I have.

Handle is ready for final shaping. D-shape please!

I have to admit that I was very nervous about the final shaping - it is so easy to make a mistake when you are nearly finished and grind off a piece of that facet that defines a D-shaped handle. But I really have to say that I managed to stay cool and ground the handle to shape without any serious problem.

Fresh off the disc sander, the taper at the front side of the ferrule is still missing.

Info: I did one process mistake. This handle was requested with a single facet on the right side - tapering the handle width towards the blade. Without thinking much about it, I ground this facet right after I was done with the handle shaping. This made the subsequent hand sanding (up to #1000 grit) much more complicated, as I had to take care of the additional edges.

I would have made more sense the sand the handle up to final grit and then grind-in the facet and refinish only the facet. It would have made it easier to keep the edges cleaner and sharper.

Once the handle was finished I have applied several very thin (just applying and immediately wiping dry) coats of Tung oil and then lightly sanded with a 0000 steel wool. Finished :)

Finished handle. Oiled, but not sanded.

Finished handle. Oiled, but not sanded.

Finished handle. Oiled, but not sanded.

Finished handle. Oiled, but not sanded.

Finished handle. Oiled, but not sanded.

The very last step to do was the glue the blade inside the handle and sharpen the knife.

Gluing the handle.

Finished knife

Far from perfect, but still very rewarding:

Lessons learned

• Accepting a custom project that lies far beyond your current set of skills will take very long to finish.
• Test any new process step before you apply it to the product, or make sure that you can redo it with reasonable amount of effort and expenses.
• Custom projects are fun, but very time consuming. They are a challenge and the customer should be made aware of that and adjust his/her expectations on timing accordingly.
• The more outside your skill zone the project lies, the less 'perfect' the result is going to be.
• The more open & communicating you are with your customer, the more understanding will he/she have when unforeseen problems cause yet another delay.
• If you plan to make a new shape of handle, write the whole process down and see whether all steps make sense and are doable with your set of skills and tools
• I have 'discovered' finishing the grind of the knife on stones and will continue using this technique. As of now I have another 2 blades that have seen this treatment and I am pleased with the result.

I would like to finish this long article with a big Thank You to my customer who not only showed active interest in the whole process (we exchange a few more than 100 emails), but also granted me all the time I needed to finish this project. It all started back in March 2017 and was finished in July 2017. My original estimate was "it will take at least 3 months" - it seems I was right :)