07 > molding & casting
Im experimenting with a thermoelectric
elements to produce a cooling facade system (using the
Peltier effect) in loop with a thermoelectric generator
(utilizing the Seeback effect) which is able generate a
current from concentrated solar thermal power (using a
Fresnel lens. These devices depend on a high temperature
differential but are simultaneously very delicate and
require heat sinks to dissipate heat to avoid failure. I
decided to use class this week to try casting a custom heat
sink for the facade system (I was hoping to cast aluminium
but this was just for demo purposes, so I continued with the
low temperature Bismuth alloy to familiarize myself with
metal casting. Since >4 TEGs are required to provide
enough voltage for a single TEC, I incorporated this into
the shared heat sink design.
My initial plan was to make a branching
heat sink to be shared by the internal faces both looking to
dissipate heat away from the thermoelectric devices, however
trying to make a mold for this was proving to be too much of a
deterrent with lost wax casting being the most feasible
method, but even then the mold would only have a single use. I
decided to cast the cold heatsink instead - creating a
repeatable tiled pattern with smooth surface deformation to
increase the surface area of the cooled surface to maximize
the radiant energy transfer to the interior.
To create the positive for the mold - I
designed the mold in Rhino and grasshopper and first tried
milling out of the wax - unsuccessful - modela did not pick up
the surface details though not entirely sure why so moved on
and tried to 3d print the positive on the makerbot/ultimaker
and all machines were down / failing due to poor extrusion
results. Finally I just sent the print to the zcorp. And cast
it the OMOO Max 60 high temperature resistant silicone mold
measuring out the part A & B at a ratio of 100:3 and
mixing for 3 minutes. Poured into one corner and manually
vibrated to remove any air bubbles. Left it for 24 hours to
cure and 6 hours post demolding.
To cast, we used low temperature Bismuth
alloy, found here on
rotometals that melts at 281F so can be heated in a
tabletop oven. Once poured into the mold it takes less than 5
mins to solidfy and remove. Unfortunately its one of the
lowest thermally conductive metals so wont actually have great
performance as a heat sink
This term I managed to land a spot in the
coveted glass blowing classes at the glasslab in the basement
of building 10. Molds are often used in this craft to
manipulate the molten material into desired forms and can
provide alot of control for novices like me. I was interested
in trying out some molds for rolling cups with an angled base.
Materials for glass molding have to be highly heat resistant
which makes a metal such as aluminum ideal. Looking for
something more easily machinable, wooden tools are prone to
wear quicker but their life can be extended by soaking them in
water to minimize burning. So i secured a chunk of cherry
wood, and started with the half mold due to time contraints.
Following Chris Dewarts suggestion - milled an impression in
the center and then manually carved it. It was tested as a
spinning mold with some good results. Further improvements:
there wasnt enough constraint at the bottom so gravity reduced
the definition of the angle (the two part mold should solve
this / structural barrier to allow it to retain shape) and
stems area wasnt quite wide enough for moyle, so will carve
additional space and soak in water to minimize burning
through. The second half will be milled so can compare the
results and will hopefully produce a more refine angle at the
base when since it can be fully contained. Many thanks to
Peter Houk and Marty Demaine at the glass lab for their
support in the process. Further experiments to come soon!
