This week I made some real progress. This picture is also a bit over-zealous. I completed an entire metal component and it’s beautiful.
Well at least I think it is, I’m extremely excited about it. I created a three piece joint component. It’s broken up into 3 piece’s for a couple reasons. First and for most, the joints need to have enough throw to actually fit the strandbeest design. Secondly, the joints need to be strong enough to support the whole beast. I could achieve the same degrees of travel by making a deeper notch into each piece, but I am afraid that that will leave a longer thin piece of aluminum (rather than the current stubby piece) and therefore that would mean less force for the part to shear off (force x lever arm = torque). Lastly, I also think it looks cooler than just a single point of rotation. As for the axles of the joint, I just took the ends of some long rivets and banged each end down with a hammer. I don’t think I’ll settle on this as a means of connecting the joint pieces, as I’ll want to be able to easily assemble and disassemble the legs, and also fine tune the amount of friction each joint has. For now though, chopped up rivets it is.
This joint system will hopefully make a home very soon with the rest of the Grazer leg, as I’ve just made a large metal order: 24 feet of 3/8′ inner diameter box aluminum (for the length-giving pieces) and 6 feet of 3/8″ outer diameter square aluminum rod (pictured, for the joint components). This gives me just enough aluminum for one leg, my goal for the presentation on the October 9th.
Also I think I finally figured out how the CNC machine keeps messing up, and created a workaround to solve the issue. Every Tormach — probably every CNC machine has some variation on this — has what’s called a quick change tool system, to easily swap tools in and out. It works by synchronizing the tool number in the CAD program (the thing that spits out the code for the machine to cut the part), with the tool number in the CNC controller tool table (the computer program that runs the machine). Each tool in the tool table, assuming they’re even synchronized with the tool number that the CAD software used, has a built in Z-axis offset (height) that automatically subtracts from the previous zero value. This theoretically makes it easy to swap tools without having to find a new zero. Or, if the operator doesn’t entirely understand this, it can make for some really confusing “errors.” The machine may run perfectly when the tool numbers are all synced up with one specific tool that matches (the 1/2″ x 1/2″ ball endmill) the tool table in the CNC controller. But with all the other tools, the machine would either dive or rise unexpectedly. This was such a difficult problem to spot since I never thought that it would have to do with what tool I put into the machine, and that problem actually arose in the syncing of the CAD program and the CNC controller, not the seemingly sketchy controller itself. The work around: don’t use offsets. I loose a little time in finding zero again, but also loose the amount of mangled pieces. I hope to refine my understanding of the system next week so I can actually use it correctly.
Since I have made these new discoveries, and a new step in production, I’m suddenly overwhelmed with excitement to continue pushing the project forward. Before, I had an idea of what it will look like, and I was a bit disinterested, but now I can really start to see it, and I’m excited to continue the work to completion. Onward!