WIP Unmachined, a D2 box cutter

Joined
Sep 29, 2008
Messages
10,394
I have two different ideas I'm working on. Up first is a kiridashi-inspired box cutter. I wanted a splinter picker tip, just enough handle to pinch or pencil grip, and something I could reasonably start out with. The second design is a full-blown knife with handles and a double sided grind. That will be harder, so I'm practicing with the single edged smaller knife first.

I'd welcome any questions, feedback, advice, or suggestions. My primary objective here is to learn. Secondary objective is to make something hard and sharp! This post has the design, the next will have heat treating and finishing, and the third will be some sort of conclusion/learnings.

Tools
I'm currently using or planning to use a drill press, bench vise, an assortment of hand files, and some miscellaneous fixtures. To be clear, I'm not shunning cost or machinery. I think it would yield a better knife. However, for the space I have on hand I absolutely do not want to be grinding metal. I also think this is something achievable with the tools I have at the tradeoff of time and elbow grease.

Knife Design
Here's what I'm hopefully looking to make. Note this is modeled in reverse for a lefty. This was designed in Fusion 360 and rough modeled on paper and cardboard before I settled on the shape. Fusion was particularly approachable for me and they have excellent tutorials for quickly learning the basics. I did some drawing initially on pen and paper, took a picture, and then started modeling lines around it. The rear hole is 5/32" and I plan to enlarge that to 3/16" with some carbide. Here's the model and the outline.

Screen Shot 2020-09-12 at 8.04.00 AM.jpg Screen Shot 2020-09-12 at 8.04.20 AM.png

And spoiler, this is how the (second) knife is shaping up.
APC_0179.jpeg

Roughing Blanks
I'm bound by some limited tooling. I first looked into raw stock, tracing a template, and then hitting it with a hacksaw (ugh) or a cutoff wheel (messy, hot). I found someone who will laser cut blanks in 1/8" D2 at around $35/ea. The input is a DXF (digital outline) and the output is a clean and accurate cut profile. Here are three blanks in various stages of work. The top is untouched, the middle has the edges cleaned up, and the bottom was my first stab at putting a primary bevel on. It looks like a mess because it is.

IMG_9279.jpeg

First mistake: I forgot to account for the HAZ removal. I didn't resize the outline before I sent the file. This means the profile is exactly correct, but I still have to take off about ~0.01" of heat affected edge. The handle hole would be great at 5/32" but now I have to clean it up and drill it a step size larger. On the left you can see a clean edge I made with some 80 grit silicon carbide sandpaper stuck to a piece of 1" thick flat stock. I'm not sure I actually removed all the HAZ here. On the next one I plan to scribe a line and check the bite with a file.

IMG_9280.jpeg

Workholding
With the edges cleaned up (mostly) I need to clamp this down so I can cut the primary bevel. My first take was done clamped to a 2x4. While this worked it caused the edge to warp slightly because of the pressure and soft backing behind the thin edge. I had a piece of ¼" aluminum flat stock so I profiled the blank and installed some 2-56 socket head screws to keep it in the same place. There's a rear hole that I intended to tension some wire behind, but a clamp works better. This whole fixture is rudimentary and I will be thinking about how to improve this if I cut more of this profile.

IMG_9282.jpeg

Cutting the Primary Bevel
I took a number of pictures throughout this second piece but unfortunately I changed the clamping situation up so it doesn't tell a good story. I started this one the same way as the first one: cutting a sharp 45° ish chamfer on the front until I hit about the ballpark of the final edge thickness. I was using a new file, this 16" double cut hogs metal much faster than the 10" one I was using before. I'm so much happier with this one! However, it means I left the edge a tad thicker to account for taking it down to size with a finer file.

Once I hit that edge I blued everything back up and started working on the top shoulder until I had removed all the blue. This let me file fairly indiscriminately by simply watching how far away from the blue I was. As I moved further up the bevel I started being more careful and approaching high spots specifically. For example, the heel is hard to draw file here since the teeth can grab that sharp edge and gouge it. I mixed techniques, either push filing (from the top of the bevel down towards the edge so everything was well supported) and draw filing (heel to tip) to make sure everything was flat.

About every 3-5 minutes I would clean everything off with a brush, re-blue the blade, and make some light scratch passes draw filing. This would tell me where (in one direction) the blade was uneven. I would then decide on how to approach flattening things out. Eventually I got to a point where everything was fairly flat on a scratch pass *and* I was hitting the primary bevel scribe line. This is a picture freshly finished with the 16" file. You can see some scallops on the edge near the heel where I messed up. I almost got rid of those completely in later finishing stages, but if I run a fingernail I can feel them. I think this will come out a bit when I put an edge on it, but it's still a flaw.
IMG_9296.jpeg

At this point I cleaned everything up, re-blued, checked the edge thickness, and started working with the smaller 10" double cut file. This removes a lot less material but is less likely to grab. Once everything was smoothed over I progressed down to a 2nd cut file and then to a single cut 8" file. I may grab a finer single cut file to pull this even tighter on the next one.
 
I think you would be time and money ahead to pick up a small grinder. Honestly I don’t think everyone needs a 2x72. That has been the gold standard for a long time but nothing says you have to have one. I personally like the idea of 1” wide belts but that’s for my uses. I personally think a small simple 2x48 or 1x42 that bolts directly to a motor would be the bees knees for knives like this. Doing all that by hand Is a ton of work. Hell when I started making knives I would use my dads board sander and flip it over and clamp it in a vise and grind knives in that. I was only 12 or 13 at the time and making blades out of scrap but man I had fun. Are you putting any scales on these blades?
 
I think you would be time and money ahead to pick up a small grinder. Honestly I don’t think everyone needs a 2x72. That has been the gold standard for a long time but nothing says you have to have one. I personally like the idea of 1” wide belts but that’s for my uses. I personally think a small simple 2x48 or 1x42 that bolts directly to a motor would be the bees knees for knives like this. Doing all that by hand Is a ton of work. Hell when I started making knives I would use my dads board sander and flip it over and clamp it in a vise and grind knives in that. I was only 12 or 13 at the time and making blades out of scrap but man I had fun. Are you putting any scales on these blades?

You're not wrong. Unfortunately I absolutely cannot fit a belt grinder in the mixed use shop area without serious dust management investment. By serious I mean, no dust at all and everything get's a wipe down when done. Plus Jarod, if I picked up a belt grinder, I would be compelled to do a hollow grind and make it nice. :D I love, absolutely love, a nicely done hollow grind. I'm working on just getting a single bevel flat and even at the moment. When I think I have it down (spoiler, I won't) I'll invest in some tools with oomf. The knife making is the fun way to expanding my thinking as I develop more skills cutting and fitting physical things together.

Right now I'm targeting ~2 hours to finish "grinding" each knife. I think I can do that for ~10 or so knives before revisiting what equipment I'm using. I'm sure my mind will change when I'm trying to finish things up after heat treat. I decided on Saturday to try a rounded spine on the bottom portion of the handle. I quite like how it turned out, but now I need to make another fixture.

I'm not planning to put scales on these box cutters. I am planning scales for another design. The flat option would have a simple chamfer around the edge and I can probably do that on a router. The 3D option would require milling because the surface contours are detailed.
 
Last edited:
Alright! I’m back at it and carved out a few hours this morning. I had two goals. First, refine the bevel on the “better” blank. Second, try some new techniques to clean up the laser edge on my last blank.

The bevel cleanup is maybe less interesting. I took my very cheap (costs less than just the covers Starrett makes) surface plate and put some 400 grit sandpaper on it. The 400 grit was handy and I was just trying to lay down a reference scratch pattern, not remove material. Here's a quick diagram of where I was at after this. There's a distinct low spot around the heel, but the bevel is /mostly/ flat otherwise. This sounds decent and I'll take some more material off to flatten it out more.
Napkin.jpg
My progression is very similar to what I started with except I added my "flat platen sander" to the mix. Blue -> file -> blue -> scratch -> blue -> file -> etc. Clamping didn't change, I have the small fixture and a 2x4 underneath it. I'm still thinking about this problem since I believe holding this more rigidly with better access would let me (1) apply more force and rough profile faster; (2) inspect and adjust with more accuracy.

Clamped with some blue sharpie.
IMG_9487.jpeg

Started back with the "2nd cut" file. This is a 10".
IMG_9488.jpeg

In progress adding some more blue to make sure I'm not hitting the low spot.
IMG_9489.jpeg

This is out-of-order, but I mixed in a Swiss Pattern #1 and #2 files for more control over removal rate. I'm happy with the finish draw filing on the 2nd cut, but these push cut with a cleaner finish.
IMG_9497.jpeg
 
Here's my "flat platen sander." It's a 1" x 12" piece of flat stock. It's not as flat as the surface plate, especially since I've dinged it up, but it's flat enough. I have some 80 grit sandpaper with an adhesive back. This comes on a 2.5" wide roll. I might grab a piece of 2.5" x 12" flat stock so I don't have to cut it and I get a wider sweep.

IMG_9490.jpeg

I'm using this to do lightweight flatness checking while the knife is still all clamped up and the surface plate carefully tucked away in a cardboard box. This is a quick scratch pass, it's looking not bad. Still low in the heel but getting more even. I'm taking off material where I want to be taking off material.

IMG_9491.jpeg

Further along and making sure I can get most of the file marks out and I'm not gouging it too bad. I'm not doing a good job here, there's some deep scratches here.
IMG_9492.jpeg
IMG_9493.jpeg

More flatness checking intermixed with filing. The gouging gets worse as the bevel gets closer to the edge, but I get that cleaned up a little later.

IMG_9495.jpeg IMG_9494.jpeg

I'm thinking it's close to done, so I blue everything up again and put some fresh paper on for a light scratch pass.IMG_9498.jpegIMG_9499.jpeg

Not bad. The primary bevel exit isn't parallel with the edge (my bad) and it's moved further up the blade.. I need to make a firmer scribe line (maybe mask it off) and stick to that next time. I think this is a shallower primary bevel angle than I intended.

I still need to check things on the surface plate, but it's more flat than when I started.
 
Look, a wild 80 grit "belt" finish has appeared. Still low on the edge of the heel, some gouges near the tip and edge. I expect to clean this up on the surface plate.
IMG_9500.jpeg
 
HAZ cleanup. The first profile I did this with some 80 grit sandpaper. This wasn't great. Specifically grit got everywhere and I wasn't sure if I was keeping true to the profile. I decided to try something new. I already had a few stones for various grinding/polishing laying around from McMaster, but I was searching around specifically for hardened edge cleanup. Apparently there are "EDM stones" which seem to be "specific grits and wear rates for removing metal." I ordered a sampling set from Falcon tool and got to work with the "ROUGHOUT 80" (very rough, use dry) stone. This seemed okay. It's abrasive, didn't wear very quickly, and took material off with a decent finish. I'm going to keep experimenting here since I have a few different firmnesses and sizes to work with.
IMG_9501.jpeg

The edge has cleaned up nicely. I think I can tell when the HAZ has been removed since there's a texture difference and the stone feels different. Instead of feeling abrasive it feels more like I'm digging into a dough or putty. It feels mushier.
IMG_9502.jpeg IMG_9506.jpeg

This is a 150 grit round stone (not from Falcon, McMaster 4347A46) which I used on the handle curves. I varied between using the tip or the rounded portion. It's not as fast cutting as the ROUGHOUT but it's what I had on hand.

IMG_9503.jpeg IMG_9504.jpeg

This is not finished, just a progress shot. I’ll get clean pictures once I’m finished.
IMG_9505.jpeg

I'm trying to pay close attention to the profile. I want this edge to be square and not rounded or lopsided. The butt of the handle also has a creased edge in it which I want to preserve. This makes two distinct curves, the underside from the heel of the blade to the crease and top from the tip of the blade all the way around the lanyard hole.
 
Last edited:
This is it for today. For comparison left to right -> First profile (rounded handle), second profile, third profile.
IMG_9510.jpeg
IMG_9509.jpeg
 
Thanks! I’m learning and I don’t want to rush to heat treat. I can imagine fixing a grind or edge when hard wouldn’t be as “fun.”

Its easy with a belt grinder lol ;)
 
I'll probably get there sooner than I expect but slower than ya'll would recommend :D
Move at your own pace, don’t let anyone rush you faster then your wanting to progress. This is your ball game and you control the field.
 
Very quick update time. I want to measure the profile of this bevel because I want it to be flat and even, and if not at least consistent. I want to eliminate high and low spots since the cutting edge will be "wavy" if it's not even. If it's not flat the cutting edge will taper, which isn't the worst thing but something I want to avoid if at all possible.

I haven't thought of something worth trying yet. I think I might build some sort of fixture which locates on the rear hole and aligns with a fixed spot on the handle. Then I can mount this on some sort of adjustable base so I can sweep parallel to the edge.

However, while I was distracting myself thinking about all this I noticed (with eyeballs) that maybe this isn't straight. I wiped everything down and put it on the import surface plate with the import test indicator. The tip is ~0.01" low (or high, if measured from the rear, modulo all mistakes on my end). Since I blocked this all up under a plate specifically to better support the edge I'm a little disappointed—but not surprised. I knew this block wasn't correct (see the chamfer filed into it) and the tip overhung by a hair or so (see this in post #8).

:sneak edit: The tip is bent down (if viewed from the other side), not inconsistently ground. I strongly suspect it's not a perfect flat and even grind as well, but I can't get a number on that until I think of a good way to measure that.

IMG_9645.jpeg IMG_9646.jpeg
 
Last edited:
I spent about an hour fixing this fixture. I used a larger piece of steel flat stock for rigidity and 3/16" dowel pins. I squared the edges up and then incrementally laid out each hole. The rear pin went first, then I could rotate the blank around that to lay out an intersection for the second pin against the blade. To get the right distance I scribed a line around using a 3/16" drill bit, then swung the profile around with a scribe. This gave me an "x marks the spot" exactly where I wanted.

Getting the right hole location for the small clamp was less precise and involved some eyeballing. The clamp can take up 0.03" so I didn't have to be absolutely precise for it to work.

Everything came together very well, and then I realized I did the whole thing backwards for the wrong side of the knife. Oh well. I can make myself a lefty version now. I flipped everything and did it again. Never hurts to practice.

With a more rigid and fully supported edge I should be able to eliminate the bending the first one experienced.

IMG_1041.jpeg
 
Last edited:
Back
Top