Hatcher Knives :: Scaeva Work-In-Progress

Scaeva Work-In-Progress

For this build I'm borrowing some design ideas from the various iterations of the Roman Gladius and the Iberian swords which preceded it, as well as some elements from earlier bronze age swords & daggers.

As drawn, the OAL is 23" with a 17.25" long blade. Steel will be 5160. I have yet to decide on the hilt materials, but I'm shooting for under 2lbs/900g total weight, which will somewhat constrain material choices.

Got her profiled. Next step is to break the 90, straighten, normalize, thermal cycle, harden and temper. Planning to grind this one post-HT. Still undecided on hilt/handle/pommel materials.

Using the arbor press to get her somewhat straight before normalizing. In all likelihood this will be repeated several times between thermal cycles and while grinding.

First normalization cycle:

The short sword was normalized and thermal cycled three times, then through-hardened. She warped a bit during hardening. I got the worst of it straightened during the second tempering cycle. Here she is shimmed and clamped and ready for the third tempering cycle. Once straight, I'll clean her up and draw back the tang and center of the blade with a torch.

I did my best to heat both sides evenly, but drawing back the tang/middle of the blade with the propane torch induced a bit more warpage. A few more cycles of shimming/clamping/tempering and things were straight again.

Next step was getting some non-linear distal taper ground in (.225" at tang, .215" at widest, .180" at tip), and blended/cleaned up on the disc grinder.

Time to break the 90:

90 is broken and she's ready for bevel grinding. That tip has already run through two water buckets and I suspect more will perish before she's finish ground.

Beveling has commenced. May be grinding this one over several sessions. She's right at 2 pounds at this point.

It took many hours of beard scratching, but I may have settled on hilt/handle materials. Stabilized afzelia, Honduran rosewood, and copper.

Another grinding session or two later . . .

Right now she's at 25oz with an edge thickness around 40-50 thousandths. Back to the grinder tomorrow. Planning to take the edge down to 20 thou or so, then convex to sharp and do some test cutting. She's already coming alive in-hand.

Refining with stones and paper.

Feathering in the edge on the rotary platen.

Followed by a bit more hand sanding:

Shaping the shoulders and tang to receive the hilt components.

It occured to me that in order to have a clean guard fit-up, no part of the tang can be proud of the blade. So back to the grinder I go to carry the bevels down the tang.

Followed by yet more hand sanding . . .

Copper, stabilized afzelia, and Honduran rosewood hilte components rough cut.

Time to drill/file/fit the copper portion of the guard to the sword, then inlay it into the handle.

File and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit, file and check fit . . .

Almost there . . .

Time to bring the copper guard to its final shape and carve a relief pocket into the afzelia to accept it.

I'm no good at carving yet, but it sure is a lot of fun. Big thanks to David DelaGardelle for setting me on the right path and to Jake Powning for his awesome carving tutorials.

Filing the slot through which the sword's tang will pass.

Honduran rosewood drilled, filed, and fit to tang.

Radius > parallel.

Afzelia pommel drilled, filed, and fit to both the tang as well as the Honduran rosewood grip.

Starting to remove excess material before shaping.

Slowly taking shape . . .

Shaping finger grooves into the Honduran rosewood using the 2" wheel.

Rough-shaped at 60-grit. Time to refine with a 220-grit belt, then finish by hand.

After hand-shaping:

After a bit more shaping . . .

I carved a pocket into the afzelia pommel and inlayed a piece of copper that was drilled and fitted to the end of the tang. This will serve as a peen block during final assembly.