My GRT (Graduate Resident Tutor), Kyle, on my floor, 5W (aka QED), works at CSAIL, has waterjet access, and knows OMAX.
OMAX is the software that converts an Autodesk or a SolidWorks file into a coordinate language the waterjet can use to cut out the part. The first time I went waterjetting with Kyle, we had to format all the parts in OMAX (a couple hours of work at the computer in the waterjet room). Later that night, I got the software to be able to finish all the formatting for future water jet trips.
I used OMAX Layout to set the quality of the cut (how slow the water jet cuts along the edges), set the cutting path (this is extremely important, because if the water jet cuts on the wrong side or in the wrong order, parts will be messed up).
This feature helps you see what the water jet thinks the parts are. If the highlighted parts don't look right, then the lead-in and lead-outs are probably incorrect. (Check the angles of the leads and make sure they are obtuse to the direction you want the water jet to go/cut).
The OMAX water jets compensate for the cutting thickness of the water stream (with abraisive). Therefore, if the machine has a cutting diameter of 0.03" and you input that the cutting diameter is 0.03", the machine will cut the part 0.015" bigger on each side to compensate.
Because I want some parts to interlock, I didn't want the water jet to cut my parts to their exact dimension (the top of the part would be cut at its intended dimensions, but the bottom would be slightly over-sized due to the diffraction of the water jet stream, the taper).
So, I "tricked" the machine by inputting a cutting radius 0.014" (cutting diameter 0.028"), a smaller cutting diameter (its actual cutting diameter is 0.03"). This would cause the water jet to center its nozzle closer to the dimensions of my part, which would then cause the water jet to cut the parts 0.014" smaller around all the edges. Success! Sort of. I didn't input the right thickness to trick the water jet enough. So I consulted DGonz, who has experience cutting t-nutted parts on the CSAIL water jet, and he said 0.0115" works and requires no sanding afterwards.
Kyle taught me the process of turning on and operating the CSAIL water jet. There are a bunch of valves and buttons to push in certain order otherwise things aren't happy. (This post is more about documenting that I water jetted parts rather than a step-by-step instruction of how to use the CSAIL water jet).
Look at it go!
My first water jetting experience was a success thanks to Kyle! We were able to cut out the inside plates and back plate before the water jet started having problems cutting all the way through a 1/4 aluminum plate (evidence: water was shooting up and spraying everywhere rather than escaping beneath the part).
The water jet is out of commission for now until they fix the nozzle. (It had worn down after a lot of use. It wasn't explicitly broken by us). It will take a couple days to ship, test, and replace it. Until then, I have enough parts to start machining and assembling them into an orbit wheel shape.
I first drilled and tapped the inside plates using the mill and a 8-32 tapping bit.
I tapped down 3/4" because the first 1/4" of tap doesn't contribute to the threading. Therefore, I would end up with a good 1/2" of threads for my 3/4" flat head 8-32 black-alloy socket screws I ordered from McMaster. (The screws will go through a 1/4" aluminum plate before threading into the tapped holes, thus only 1/2" needs to be tapped).
I countersunk the holes with the mill, because they look so much nicer when done on a mill. A cool trick Ben showed me for knowing how deep to countersink: If you can fit the head of the screw in the countersunk hole then it will sit nice and flush when screwed in.
I filed off the edges of the countersunk holes with a small file, and then screwed the inside plate assembly to the back plate.
Because the water jet has a small taper, I made the wider edge face towards the hole to prevent having big gaps on the outside.
I finished machining the parts I water jetted. Now I can start working on the posts that span between the front and back plate. These will also hold the VXB bearings which will support the wheel and ring gear.
This pic shows all the screws that go through the back plate. I used 1 1/4" 8-32 black-alloy flat head socket screws from McMaster for going through the battery clips.
I used some of Ben's scrap steel rod stock from his differential which he built for his "death trike". I made six 1.5" rods and tapped them for 8-32 flat head socket screws. It's amazing how much a small drop of cutting fluid helps with tapping!
I made the steel rods stick out past the inside plates by a couple thousandths of an inch to make sure the front plate is mainly supported by the steel rods. The inside plates are just to cover up the ring gear (and potential aluminum flakes) inside.
I got a deal on the bearings from VXB (30 bearings for only twenty dollars, plus a cheaper set of digital calipers). They recently updated their site so it's much easier to navigate. The 8 mm inner diameter fit well over the steel rods.
Next is the battery pack. I've got some ideas on how to connect them into a 5S2P pack, but I need to talk with some experts (Charles and Shane).