Today I started gradually increasing the pretension in the torsion springs. I also started experimenting with different draw lengths with the 14 dactyl arms. The best results were with attained with draw that was 14cm less than maximum; all numbers are 2 shot averages.
| Bolt weight (g) | Velocity (m/s) | Energy (J) |
|---|---|---|
| 30.10 | 69.07 | 71.28 |
| 20.58 | 76.685 | 60.51 |
| 10.23 | 88.71 | 40.25 |
This is not bad at all, even though there is lots of room for improvement. To get some idea what these still preliminary numbers mean here are some comparisons:
- My wippe-operated ~300 pound medieval-style crossbow with 28.5cm powerstroke, a dynema string and a spring steel bow shoots the same 30 gram bolt at roughly equal velocity, 69.91m/s. A real medieval, forged prod with less optimal cross-section and and linen string would drop this number considerably, probably to 60-65m/s.
- 30 pound handbows - whether they're all-wood or modern recurves - give about 22 joules of energy with 22-28 gram arrows. So currently the cheiroballistra packs about the same punch as a 100 pound handbow would.
I feel that at least 50% improvement should be achievable as far as pullback is concerned. So far I have not really used arms to assist pullback, nor really attempted to force the slider back. This is primarily because I'm concerned about the durability of the arms. In today's tests the arms worked really well, but they did bend significantly during pullback. Even though the craftmanship of the arms is pretty perfect, they will break when overloaded, and I don't really know where that point is. Therefore I've been quite cautious in trying to squeeze more power out of the springs.
A few important lessons were learned from the shots I made, including those listed above. First of all, trying to go for the "full" draw does not make sense due to two related things:
- Near full draw the bowstring starts to pull the bar away from the cone, and the cone out of the torsion spring. This can easily cause the spring cords to slip over the butt-end of the cone.
- The performance gains start to drop after the angle of the botwstring goes past 90 degrees. In my preliminary tests the performance actually decreased at full draw, compared to draw 7cm below full. This phenomenom has to verified later.
A few other valuable things I learned today:
- The bar has to be attached to the cone somehow. While just having the bar tightly in the groove was ok-ish, eventually the bar started to slip towards the front of the cone. As a temporary remedy I drilled 1.5mm hole to end of the bar and through the cone and pushed a nail through both of them. A better solution is to make longer bars which can be bent around the butt-end of the cone.
- When the hoops in the cones are squeezed really tight, they will hold in place.
- When the wooden wedges are beaten really tight, they seem to stay in place even at high load.
At the end of today's shooting session I noticed that the field-frame bars had slipped away from the notches in the little ladder beams. Part of the reason is that the beams were bent to a suboptimal angle; this, in turn, was because I had not originally foreseen the need for the notches and thus had to bend the little ladder beam further apart than is actually necessary when the notches are used. I already started correcting the angle, but am only halfway through.