Concept hybrid additive subtractive full strength precision metal printer
This printer is a concept that has come out of my library research, highly messy notes on
which can be found on pages like the metal deposition head and multimaterial printer concept
pages (see my edit history). It seems to be probably the easiest to develop methods of printing metal IT5 or so accuracy full strength low stress metal parts, although there are other ways that might of course turn out better, can't really know... but it looks like it would turn out to be useful in any case.Greenatolsecondtry 16:20, 9 September 2011 (UTC)
Provide the capability to make parts to approximately IT5 precision, although even if it is
lower it will work as a sort of high precision casting production machine.
It cannot machine all surfaces because there are some circumstances, like the underside of the very top underside of a dome, that you cannot reach with the milling tool from above
during the print process.
Overview of how it works
This uses superheated (heated above melting point some ways) dripped onto the metal or
support material below. This is called microcasting, and the advantage is that the object
you are printing does not have to be electricaly grounded. This provides full strength,
void free metal with low cost heating on a per watt basis (no lasers or expensive power
supplies). The metal may then optionally be stress treated immediately after deposition.
See the stress relief section for details.
The build process is done under an argon atmosphere. Since the machine can print a device
to fractionate air, the argon supply should not be considered a problem.
After the metal is deposited in layer 1, another few layers are deposited and then the metal
edges of layer 1 are machined by a small mill head. There is a separation of a few print
layers or so between the area where metal is being deposited and anti-stress treated, and
the metal that is being machined, to reduce or eliminate distortion caused by any forces and
thermal stress in the machining area after the surface has been shaped.
After the metal has been machined a support material, in this case a water soluble ceramic,
silica glass, with a fine powder material to adjust the coeeficient of thermal expansion, is
deposited onto any volume inside the current print plane. This is done by depositing a hot
saturated solution/slurry of it, and allowing the water to evaporate. The build volume is
heated to 90 degrees or so and there is a fan at the top of the build chamber the help with
this, as well as a radiator for the water to condense on (in turn cooled by some means).
The support material is recyclable easily by redissolving. The powder can be removed easily
by filtration if desired, so it can be replaced with a newly reformulated powder, to match
the CTE of a new metal alloy.
Build process order/direction
It is a 5 axis machine. 3 linear axis on the deposition and milling head, and 2 rotating
axis, with the build plate capable of tilting 30 degrees around 2 axis.
During the print process, the build direction can be tilted by about the 30 relative to the
horizontal (relative to the build plate) by tilting the plate. The main purpose of this is
to overcome the difficulty of printing, and then milling, the underside of horizontal or
nearly horizontal ledges and surfaces. To print such a ledge, you just tilt the print plane
so, relatively, it is not horizontal anymore.
There will still be some geometries you cannot print this way of course, but you can print a
lot. Secondly, when there is a surface you cannot print in this way, you can still print a
less precise surface by laying down support material, machining it to shape, and then
depositing the metal on top of it.
One of the main problems with metal deposition that involves melting is that there is stress
left behind in the deposit. There are a large number of options to try to deal with this,
see the metal deposition page for more details and options.
However in this concept the following strategy is employed: First, the drop by drop deposition method is taken advantage of. Drops are first deposited
1 drop-width (the width of the drop after it spreads out on the surface) appart. Then they
are joined together. This can help a bit but not relieve all stress.
Secondly, the drop by drop method is very low heat input, and the input can be controlled
quite well by controlling the temperature of the liquid metal, which also helps.
Lastly, when the object is doen it will naturally be embedded in ceramic as result of the
build process. It is then annealed/heat treated to remove almost all stress. For
reference, tool steel takes around 650 deg C for about 3 hours. It may also be hardened
during this process by adjusting the cooling/heating rate, as long as any volume changes are
not so much as to crack the ceramic.
During the printing process, cooling channels can be left in the ceramic to some degree.
Simulation and compensation for any distortion may be done as well and could help a lot.
Self reproduction arises naturally as a result of the capacity to produce adequately
accurately shaped metal parts. The final design may use hydrostatic bearings as they
require lower precision than rolling element bearings. It will be able to print leadscrews
etc. but not to high accuracy. However they do not have to be accurate, as long as the
motion produced is precise, and the stiffness is adequate etc. Retaining accuracy in between
generations needs to be done by recalibration or the use of accurate encoders.
This process can produce I may try to build something like this if I can get access to a hackerspace. Most likely it
would be 3 axis at first, to get the details worked out. The main question is what
precision can be achieved.
The process might also be scaled up relatively easily to the frame of a bicycle/car etc. At
the very least it will presumably provide the equivalent of high precision castings. That
will open up the option to finish the object with EDM or other methods. EDM is force free
so it opens up the possibility to reach way in there and do point by point removal, without