My university lab has a long 1/16”-thickness 1”-inner diameter braided carbon fiber tube that I don't see us using anytime soon, so I permanently borrowed it for this project. I ordered the cheapest carbon steel Katana Shirasaya I could find. Carbon steels are used in actual swords while most cheap knockoffs are stainless steel. SS blades are purely decorative since they can shatter upon hitting something. Shirasaya means a plain uncoated lacquered-wood scabbard and handle. There is no guard between the blade and the handle making a shirasaya unsafe for cutting. Since I only needed the blade for this project, there was no reason to order a more expensive fully dressed sword with the same blade. After some research, I found out that my sword’s manufacturer Musashi Swords makes some of the best functional blades out of all the bargain katanas. Although now I wish I bought a curved blade instead of a straight one.
I created a simple CAD model in Solidworks to calculate center of gravity and moment of inertia about the CG. Reading up on the mechanics of sword performance, I stumbled upon the center of percussion. This is the point on a shaft-like object where if a force is applied, there is certain rotational node down the shaft where there is no change in translational velocity upon impact. Designers of baseball bats and tennis racquets take this into account. Essentially, this means that if you strike an object at the center of percussion (blade) with your hand located at the rotational node (handle grip), then your hand feels no vibration or discomfort. The hand is neither pushed back or forward; it continues with the same velocity after the impact as it did before.
The CAD model allowed me to play around with the counterweight mass to result in an optimal CG and rotational node location assuming the blade impacts 12" from the tip. The CG is located at the pink arrows and the rotational node location (created plane) is right between where I would place my hands.
Here is the blade, the carbon fiber shaft, and the scabbard:
I cut spacers from 1” aluminum rod to create a tight fit between the blade tang and the CF shaft. Steel pins hold the spacers to the blade and to the shaft. I cut out a ¼” slice from a 2.5” aluminum rod for the guard. I found a big ass steel bolt that had a flat 1”-diameter section which fit perfectly into the end of the shaft, and cut it to the correct weight. It was suggested to reinforce the joint part of the shaft, so I found a 1/16” thick aluminum tube that fit perfectly over the shaft as a collar. This will help absorb some of the bending force in that area from the shaft. The guard, collar, and habaki(copper fitting between blade and guard) were sanded and polished.
I then sanded the shaft with 1000grit, epoxied the counterweight to one end, and coated the shaft with epoxy for a glossy finish.
Pinning the spacers onto the blade:
Sliding the blade assembly into the shaft/collar assembly:
Pins are inserted for holding the spacers to the shaft. These are essentially non-load-bearing, since the blade is held almost completely with the friction fit between the spacers and shaft. I figured that placing all the torsional and longitudinal loads on small holes drilled through the weave matrix was a bad idea. Last step is to slide the collar up to hide the pins. Bring on the zombies….
Overall length: 80.7”
Blade length: 27”
Carbon shaft length: 51”
Overall mass: 3.20lb
Blade mass: 1.50lb
Counterweight mass: 0.60lb
Chosen center of percussion point from tip: 12”
CG location from tip: 39.5”
Rotational node location from tip: 61.3”
Moment of inertia about CG: 1918lb-in^2