Tuesday, November 8, 2016

Steak Knives Part 1 - Prototype

So I set out to make some steak knives. How hard could it be, right? Cut handle wood to the shape of the tang, epoxy and rivet the handle on, sand it a bit, and voila, no? Yeah. No.

After tracing the shape of the tang onto the handle material (or "knife scale"), I cut it out with a scroll saw, drilled holes for the rivets on a drill press, epoxied it to the knife, and then used a belt sander to contour the handle to the tang. It... uh... did not go super great. So, commence prototype v. 02. This time with less abject failure!

The primary failure of v. 0.1 was that the handles weren't shaped close enough to the tang in order to make a symmetrical handle that was the same size as the tang itself. So I contemplated how to improve the precision with which I cut the handles.

Behold: The router flush trim bit.

A trim bit has the bearing on the opposite (distal?) end of the shank
A flush trim bit has a bearing that is the same diameter as the cutter head. This allows you align a pattern to the bearing, and then the bit will cut to, but not into, the pattern. It allows you to "trim" a work piece "flush" ... maybe that's how they came up with the name?

Anyway, so I took a piece of wood, outlined the knife tang on it, cut it close on the scroll saw. Thenm in case the work piece kicked back into my hands, I wore heavy gloves and taped the ever-loving-sin out of the blade (I am aware that this was probably not the safest operation), and then double-stick-taped the knife to the wood. Then I set the bearing to ride on the tang, and trimmed the piece of wood so that I had a wooden pattern that was exactly the same size as the knife tang. No blood was shed in the making of this pattern. Then I switched bits to a pattern bit.

 Behold: The pattern bit.

A pattern bit has the bearing on the shank (proximal?) end of the bit

The main difference is whether the pattern you're routing flush to is on the top or on the bottom
A pattern bit is essentially a flush trim bit with the bearing on the other opposite side of the blade (Note: Apparently the proper way to describe these bits is that a trim bit has the bearing on the "bottom" and a pattern bit has the bearing on the "top." But, since I mostly use a router table, I think of them as reversed). Now, I could scroll saw out knife scales close to, but slightly larger than, the size/shape of the tang, double-stick-tape two together, then double-stick-tape that to my pattern, and rout it to exactly the size of the tang...

No the machine wasn't running when I took this picture. What do I look insane to you? Um... nevermind

Pictured: One destroyed knife scale courtesy of router bit tearout
... for the most part. Routers can tear grain out, which would ruin a knife scale (and is kinda dangerous because it has a tendency to send splinters of wood flying, or to catch the wood and kick it, and it should go without saying that you do not want your hands to come into contact with a spinning router bit). The easiest pieces to use a router on are large, long, pieces. Routers do not like end grain, and they especially do not like end grain that is curved and only about 3/4" long. I learned the hard way (see photographic evidence above) that the best practice here was to rout everything I could to the pattern, and leave an uneven end to finish off on the belt sander.

After shaping to the tang, I drilled holes for the rivets, applied a catalyzed epoxy, and assembled in a vice.

Crushed it.
The secondary failure of v. 0.1 was contouring the handle to feel comfortable in your hand. Initially I tried just rounding the edges off on the belt sander. It didn't work well. I ended up with very uneven, contours. I'm just not good enough at using that tool to create a consistent contour. So, I ended up using a hand tool that is designed to shape curved surfaces, the rasp (part 1 and part 2). This gave me far more control and led to a much nicer contour.

After shaping with rasps
After shaping the contours of the handle with the rasps, I sanded in stages up to 800. After using tack cloth to clear residual saw dust, I applied five coats of Tru-Oil, sanding between each coat with 1200 grit sandpaper).

All said, v. 0.2 was successful enough to take this operation out of beta-testing. Coming soon: v. 1.0.

No comments:

Post a Comment