Yesterday was another busy day on the pumpkin catapult front. We finished assembling the devices and tested them in Jason's back yard while he kept an eye on a smoldering pile of crop refuse and brush which he was burning in order to try and kill off a late-blight infestation in his tomato patch in preparation for the new season.

My direct-fire spring-pull ballista was kind of a damp squib. It fired, reliably, but no further than about twenty-five feet - about the same range as a high-end SuperSoaker. It wasn't really designed to be adjustable, so that was it, boom, 25 ft. So, the finished device works, but isn't much more than a bulky, wonky-looking toy. At least the spring didn't fail - we had padded it out with a series of rubber sheets around the pull-axle, which was enough to keep it from self-destructing. Speaking of which...

Jason's spring-torsion trebuchet started out brillantly, destroying the sweet onion we had been using as a pumpkin-substitute (we managed to directly smack into a branch of the big maple behind Jason's house, which pretty much put paid to that object as a throwable payload), and firing for another two-dozen pulls with a light obsidian-pumicy rock about the right size and shape. After changing the angle of attack on the throwing arm, and modifying the base to add a trigger and a shelf for the sling and projectile, we got it up to throwing small rocks about 80-90 ft across Jason's yard - just shy of his property line. Then the spring failed.

Happily, since the spring was in a protective sleeve (actually, the leg of a pair of blue-jean trousers) and it wasn't really under explosive outward pressure, it didn't hurt anybody. When we took it apart to see what happened, we found that one of the mounting pegs had also snapped, which explained one of the earlier "SPRONG!" noises the machine had made on an earlier cranking. So, we had been putting a lot of stresses on this device. The spring failed just above the welds holding it to the octogonal plate holding the mounting pegs; best guess is that the heat from the welding process weakened the aged spring (to remind y'all, this had been taken out of a trashed automotive shock-absorber) enough that it eventually failed under pressure. The mounting peg was a separate issue, and the design group concluded that having used drillcore for the mounting peg had been a mistake - it was too hard and rigid for the horizontal sheer force being placed upon it by the design. We should have used mild steel instead.

They went back to the drawing table, and quickly banged out a replacement design based on a completely different principle, which would allow the salvaging of as many parts of the wrecked device as possible. The new device would use the second salvaged shock-absorber spring in a compression device - playing to the actual design usage of the original manufacturing process. The spring was mounted in a box on the base, with a bottom-plunger on braided steel cord threaded through the spring, so as to pull directly upwards, distributed through the body of the device by the construction of the box. The steel cord will be wrapped around a pulley-cam on a moving axle, which will be welded to the salvaged throwing arm. This retains the torsion motion, while avoiding any weld-work on the surviving reserve spring. We spent yesterday afternoon and evening putting together the new base & welding together the pulley-cam & modifying the original three-quarters-inch bar, which had been a non-moving support element in the old device & which had to be reworked to turn it into a smoothly-turning axle in some brass bushings.