The last four days have been intense as I worked to get a printable design to Motive so they could get started. As it turns out they had the plastic I wanted (red!) already in the machine and they had time to do it this weekend, hence the expediency.
As shown below, I went through four iterations in 24 hours, the first three drawn from scratch:
The first design was because I envisioned their FDM machine using powder, but it is an additive process which means it builds free standing structures. Design #1 would have required a lot of dissolvable supporting plastic inside and would have increased the cost greatly. Hence, designs #2 and onward are designed with as little support as possible and everything higher up on the model is supported by something below. The third iteration fixed a few problems and shows that instead of creating a large cap, I am going to cut out a piece of the pure copper sheet metal I have and bolt that to the large end. The narrow end cap has that strange square structure jutting up because that is where the tuning mechanism is to be mounted, shown below:
The final cost was $840 (GST included) which is very good considering alternatives and the technical help (we exchanged over thirty emails in a 78 hour period). The part is now being printed as I type this and baring any unforeseen events, it will take a full three days for it to finish.
While I was drawing up the models, I also had a concurrent conversation going on with an electroplating place on the West Coast called Pacific Plating. I have been lucky enough to deal with the head guy who apparently likes challenges, and he is the reason for revision #4 of the model above. If you look closely it has extra protrusions, the top two for hooks to hang the part in the electroplating bath and two at the bottom to put weights on it keep it from floating. Pacific Plating specializes in plating plastic and he suggested that I send them samples of the ABS covered with the copper conductive paint on the surfaces I want coated. They can then test each sample for a different time interval, and I can decide which is best, or if the process works at all. The model is designed to handle upwards of 0.45 mm of electroplated copper around the narrow end, which is a significant amount to electroplate.
The copper does not have to be thick because most of the energy is centered in the middle of the end pieces (assuming the TE0,1,n mode), which is where thick pure copper plates will reside. One benefit of copper plated plastic will be weight, which given this cavity is to be hung from a pendulum, will make any force that more apparent.
Before we push a kilowatt of power into the 3D printed cavity, I am going to built a second test cavity using the wood molds I have mentioned previously. The idea is to strap pieces of copper over the mold, brazing them, then repeating for the other side. Then the two sides can be bolted together after the probes have been mounted.
With the handmade mold, we can make sure there is no arcing between the probes and other unforeseen events before attempting it with the more expensive but more accurate printed cavity. I plan to use the cracked mold I already made to test the idea, and then I will have to put the second piece of wood on the lathe. The good news is I have lots of copper.
My To Do list now consists of:
- Purchase the copper conductive paint
- Get the ABS plastic samples painted and sent to Pacific Plating
- Finish the vacuum chamber once the drill bits for the acrylic get in.
- Get the copper only test cavity built.