The Chevy coil spring has been forged to hug the back of our old mild steel kamarion. Next it will be Tig welded in place. At this stage, cutting the kamarion in half allows the tangs to fit into the field frames without any undue stress from possible misalignment. An 8 inch long piece of 4140 at the the top the arch, will create a strong weld splice when we are finally ready to rejoin the two halves.

I remember that at the start of this project I had to decide what material to make the field frames and kamarion from. My final choice was to use mild steel, thinking it would be a viable substitute for the wrought iron assumed to be used in the original. By taking this lower grade approach, we now know enough to ask and answer certain questions about the artifacts that never would have arisen if we had used top grade materials from the start. Was the original Orsova ballista made to run with draw weights in the 5,000 lb class? As can be seen from the most recent rounds of testing, the stanchions and kamarion would need to be made from tempered carbon steel to hold up to that kind of draw weight. Short of that, mild steel or wrought iron for these parts is only reliable up to 3500 lbs or so, and that would imply a machine of much less potential power.
Whatever the metallurgical truth may be about the Orsova artifacts, the current upgrades on Firefly seek to prove that, if the materials and workmanship were held to a high standard, the basic Orsova geometry could have been sound for a steady diet of 5,000 pound pulls. Given our previous caveats, this likely means that the Roman machines were capable of generating 2000 or more foot pounds of muzzle energy, the same level Firefly was starting to exhibit a few weeks ago.
A number of people have sent in suggestions, that instead of obsessing on brute force, the performance may be improved by lightening and shortening the limbs by a couple of inches. These are good ideas and eventually we will try these tweaks to the design. However, at this stage, it is important not to settle for too low a draw weight in order to accommodate inferior hardware. It is really all a matter of balance; creating a mechanism where no one component is more likely to fail than another, and then running the machine within strict and prudent limits.
For the adventuresome, that translates into a 5,000 lb. playground. When we get there, tweaks will be inevitable.