I also had to figure out a way to spin the spinny bolts. This turned out to be pretty straightforward – I modeled another rack-like thing onto the bottom of each jaw (on the parts I called the jaw carriages), and gave the bolts a shaft to spin on and a gear which meshed with those carriage racks. As a result, when the jaws moved, they would cause the bolts to spin as well.

At this point I had the handle, gearbox and the bottom portion of the jaws somewhat sorted out, and I exported a gaggle of STL files and started on figuring out whether any of it was printable. After some iterations in Fusion and some experimental prints, I had a prototype and it actually worked pretty well. The slider skates didn’t slide so well at first, but I sanded them and added some lubrication which got the jaw carriages moving nicely. (I later revised the slider skate design to fix those issues and lubrication is no longer needed.) The next hurdle was to see whether this setup would be able to handle the added weight of the jaws (and attached hammer parts) being present on their carriages, but first I had to design them.

I did a first pass at the jaw upper housings and even managed to include the little trapezoidal decorative parts on the sides, which I was not sure I could print cleanly. The jaw upper housings are hollow but have a nice strong wall count. They attach to the carriages using dovetail joints, so they slide on and then get affixed with screws. The tolerances on those dovetails were a bit tricky – I settled on a fitment that feels too tight at first, but this is due to parallel layer lines rubbing past each other. Once they are ‘broken in’ a little bit, they are snug but pretty easy to remove.

With the upper jaw housings all set, it was time to revisit the issue of whether the gear train could produce enough torque to move the jaws with all the weight present. Up until this point, I’d been using 3D printed shafts for the gear train. To be clear, the filament I am using for this project is PLA pro, and it’s formulated to stretch or bend rather than snapping. So, as I tested the gear train I noticed more and more looseness and play in the knobs before anything would begin to move. It turned out that with the full load applied, the shafts began twisting, rather than rotating, under the stress.

To address that problem, I made the decision to use a more rigid material for the shafts. I was not enthusiastic about using metal, because I didn’t have the tools to drill the required holes in the shafts. So I opted for G10 rods, which are lightweight, strong and easy to cut and shape. However, G10 is made using layers of glass cloth which are soaked in some sort of resin or epoxy and over time it turned out that these rods were prone to splitting under torque, especially when there are holes drilled through them near the ends. I bit the bullet, bought a small drill press and designed some drilling jigs to help me get the holes placed reliably. I used solid brass rods, and they worked out very well.

With the bulk of the design sorted out, it was time to handle smaller details and finishing touches. The cable which runs from the handle to the rear jaw was an interesting challenge – I wanted it to be flexible enough to move with the jaw, but generally keep its shape without needing constant adjustment. I also wanted it to be easily removable for shipping or transport, or in cases where the cable might interfere with someone’s other costume elements. I designed a ‘cable form’ which is printed in a coiled up state, causing it to prefer being that way even after removal from the print bed. There are holes on either end for attaching connectors using screws. I found an open source twist-lock design which was far too large for this use case, so I scaled it down and included it on the cable connectors. I sourced some cool looking cable sleeving and designed a part I called the ‘cable keeper’ which had the job of pulling the sleeving tight and sandwiching it in place between the cable form and the connectors, trapping the sleeving in place. When installed, the coiled cable puts a bit of tension on the lower twist-lock connector in such a way that it is pulled in the ‘tighter’ direction, which does a good job of making sure the cable won’t come loose on accident.

Side note: At some point, nearly every part in this assembly got revisited and redesigned at least once, for various reasons. The central gearbox part is the most revised one, and has had about 12 versions.

I decided to add a bit of decoration to the handle and gearbox in the form of badges showing my logo and shop name, sort of like they did with antique industrial machines back in the day. I used printed parts and encased them in UV resin to fancy them up a bit. Someday, I’d like to invest in a laser engraver which can handle metal, in which case I will start making these badges that way.

Part of the fun of playing with the Engineer’s Clamp Hammer is the tactile and sound feedback you get when turning the cog-knobs. This was something I had in mind most of the way though the design process, and I tried several methods of making it happen. I knew I wanted detents involved, and I first tried using magnets to cause the knobs to ‘home in’ on certain spots along their travel. The magnets, however, were too small and therefore not strong enough to provide the amount of feedback I was looking for. Next, I tried ball spring detents, which make use of a small ball bearing trapped in a steel tube with a spring behind them. I redesigned the knobs to make use of those, allowing them to drop into small depressions in the knobs for a satisfying click and a notchy feeling.

Another ‘user experience’ issue was that the jaws were sagging a small amount when they were fully open, due to their weight being shifted outward and the racks which move them being made of flexible PLA. It was not very noticeable, but it also came with a bit of rattling when moving the clamp hammer around. To combat this, I added a center brace which fills the gap between the jaws when they are closed. It has geometry designed into its underside which maintains downward pressure on the racks. This encourages the jaws to stay upright and stable when fully open. This is an adjustment/tuning part – each clamp hammer I have built needed a different amount of screw tightening here. Too tight, and it becomes hard to get the jaws to move. Too loose and they can sag a little when fully opened.

I knew what I wanted to happen, and I was pretty sure the functionality already existed in the mechanical world. I wanted to add a device which, under normal operation, did nothing. Then, when the torque applied to the gear train rose above a certain threshold, the device would slip, allowing the user to turn the knob as much as they wanted but saving the gear train from damage. When the torque lowered back down, the device would then resume doing nothing. I even had an idea that I could use some more ball spring detents to do this.

It turns out this is a common thing used to prevent gear train damage, and it’s called a torque limiter. They (mechanical engineers) do indeed use ball spring detents to build one type of them, as well. After reading a bit about them, I designed my own. The main issue this introduced was that I would need to chop the vertical gear shaft in half, which makes the resulting two shafts quite small. This was an acceptable tradeoff though, for the peace of mind of knowing this clamp hammer would stand up to all manner of clamp-trigger-happiness. Sturdiness requirement, restored.