[AR] Re: Electroforming Experiment
- From: Ed LeBouthillier <codemonky@xxxxxxxxxxxxx>
- To: arocket@xxxxxxxxxxxxx
- Date: Sat, 13 Jul 2019 10:16:29 -0700
On Sat, 2019-07-13 at 17:00 +0100, Peter Fairbrother wrote:
It is hard to get good channel walls that way. If you use a thin
layer the copper will overflow and thin, and eventually close, the
channels.
I've considered other approaches. My intent here is to come up with a
process which others could duplicate as easily as possible. If I can
come up with a process that doesn't require a plating shop plus a full
multiaxis machine shop plus 3D printing shop then it becomes more
replicable.
The alternative might be specialization where several people produce
chambers and tanks for other people commercially to pay for the
infrastructure.
For now, I'm experimenting with the DIY approach. It may not be
practical, though, I admit. I will "pivot" to other approaches as
failures indicate particular approaches as being unsuitable.
Perhaps the best way is to built up the thickness to wall+channel
height, then cut the channels in the copper.
I have considered it. I will probably try that approach as well.
I'm hoping that the setup uses the 3D printer as the basic
milling/lathing setup. If the tools are additions to something that
someone already has, then it becomes cheaper/easier to replicate.
Admittedly, I might be wrong or naive. I'll find out over time.
My thinking is that more and more people have 3D printers which could
be used to produce mandrels and specialized tooling. I'll be releasing
my tooling designs as they mature. For example, I've already released
one rotary table design [
https://www.thingiverse.com/thing:3012867 ;].
I'm currently working on another table that is simpler and more
applicable to this effort. I'll release its design as it gets to a more
usable point.
There are limits to the machining accuracy that can be attained with
this kind of equipment (but it can be done). 3D printers are often not
as stiff as normal machining hardware and so the processes will have to
be adapted to their capabilities and limitations, I'm thinking. Again,
I admit that I could be wrong.
For example, based on your recommendations, I'm thinking I may have to
design and open-source a power supply suitable for reverse voltage
plating (if I can't find one already available). Commercially available
power supplies are often focused on a different scale of activity.
Once the cooling channels are plated, then the closeout material
can be added (if needed).
Closeouts are needed. Pretty much full-stop.
Absolutely. I was talking about the filler material. If the plating
resist starts out thick enough, then it's not necessary to add
additional channel-filling material.
Fill the channels with eg wax, brush with graphite powder (which
should stick to the wax, not to the copper!), reverse electroform
until you can see the copper is being eaten away evenly, then
forward Ho!
Yeah, I plan on electrocleaning between out-of-bath-to-bath
transitions.
Thanks for the insight.
After building up the closeout layer you can then coat the chamber
with nickel for strength if needed. But do not skip the copper
closeout layer, the copper-nickel interface is not strong, and
differential expansion (both thermal and pressure) is acting on it -
you need the interface to be as large in surface area as possible.
Good insight. Thanks.
Use a high-power high-speed spindle instead of a pen? If it is high
powered and fast enough cutting forces will be light so you don't
need a highly rigid base or high power servos.
Use a (semi-) ball-end mill, the channels will not be perfect but
you save an axis.
Yes, this is what I was considering. I was concerned about the shape of
the channel if I used a ball-endmill. I'll end up trying several
approaches over time, I'm sure.
I was thinking that high-speed BLDC motors can easily be mounted on the
3D printer with a chuck to perform some basic milling/cleaning.
Best of luck, it all sounds very interesting.
Thanks. Your insight has been important. It certainly is fun.
Other related posts:
