MetalicaRap Future Developments
CNC 5mm end mill addition
Nasa in discussion papers suggested adding one tool to the machine, normally adding a tool requires a high strength cast iron support which is impracticable for home use, but if you machine a surface at each layer then the force is so low this is achievable without cast iron support, this approach has been shown with Matsuura LUMEX Avance-25 machine.. The disadvantage of this approach is that using normal non vacuum specified motors may overheat as with the build platform motors, (if it requires vacuum tolerant motors these are currently expensive.) Greaseless bearings would need to be used So as to avoid out guessing these would normally be of a magnetic variety but expensive. A cheaper possible option is to wash bearing oil out with acetone and then using all diffusion pump oil to re-grease the traditional bearings as this oil is designed not outgass (possibly with the use of a vacuum to ensure the oil goes to all corners).
Live build part heuristic simulation Rather than just remove blob errors on every 10th layer, develop a live model of the build progress of the part and update shape with 3D map of error blobs from scanning electron microscope, then modify following layer G code to take in to account existing blobs during the next layer metal addition , and thereby only intervene subtractively when blobs effects outside of the part.
A later innovative development could be the MetalicaRapWin with a beam window. A window between gun and build platform enables the use of high brightness small spot size LaBa6 filaments that last 1,000 hrs as opposed to 70 hrs for tungsten. Allows the use of barrier argon gas at atmospheric pressure surrounding build platform and associated mechanics so no pump down time after accessing build chamber. Will also offer large part manufacturing in inexpensive argon "tents" as only the gun requires a vacuum chamber. Disadvantages include; some ballooning of beam as it passes through approx 5mm's of argon atmosphere between build platform and window, lowing resolution of SEM vision system (x-ray sensing option may help or build platform chamber pump down for vision system and argon for printing), high tech stationary window will involve high tech manufacturing, Low tech aluminum slot window will reduce print speed and increase mechanical complexity as build platform to gun physical scanning motion will be required. Repeated door opening will be overcome through these windows and thus ion pumps or electron beam sublimation pumps will be less stressed.
Low tech option: Scanning aluminum slot beam window (14cm length x 100µ width)
A narrow slot window which is physically moved across the build area. Window will be cooled through thermal conduction to water channels surrounding the window. Requires minimum 100kV beam to penetrate a 20µ thick AL window, to keep beam losses below 21%. Beam loss is inversely proportional to acceleration voltage.
High tech option: Stationary window
A stationary high tech window that will be cooled through convection and radiation alone. "Transparency" of window enables the possibility of a less penetrating beam of 60kV. 400µ sheet with 50µ micro dead end holes creating an ebeam window.
Optional beam windows; High tech stationary window or Low tech scanning Aluminum foil slot shaped beam window.
- As the window version will run at 10-7 Torr LaB6 filament is possible (expensive but 1000 hour filament life and 40 times brighter),
- One Self printed Electron beam titanium sublimation pump Or distributed Ion pump with some tantalum, slotted cathode cells for argon gas collection or oil diffusion pump or turbo pump. (prototype uses turbo pump)
- Second pump is a self printed High vacuum Titanium sputter-Ion pump (with some tantalum, slotted cathode cells for argon gas collection ) or our invention Electron beam titanium sublimation pump. Prototype before self print, bought in oil diffusion pump (messy with expensive oil 100euro/litre) or turbo vane pump(4K euro extra) .
- Second pump types sputter ion pump or oil diffusion pump will need to be closed off from the chamber during chamber access and roughing pump cycle, to save cost avoiding large gate valves reduced pumping rates will be accepted through the use of smaller radius gate or ball valves with smaller 4 inch or below type pumps . ( later self print gate valves will be investigated)(An oil diffusion pump can not pump at atmospheric pressures and can ruin the oil trying to do so).
- (For ion pump; Use electron guns focus coils to provide magnetic field to sputter ion pump  50L/s max, short duration between maintenance 30Hours , self cathode refreshment by electron beam surface melting exposure in MetalicaRap's beam)
Our high vacuum pump will be a sputter ion pump as has a cost reduction and ease of use advantage for us especially with the window option . As the high vacuum turbo pump cost has been a block to costs coming down, but after redesign an Ion pump seems a good solution, approximately 300 tubes (anodes at 8KV) 15mm diameter 26mm long made of stainless, with 8mm diameter titanium plates ( cathodes at 0V) fitting in either end of anode tubes , leaving a 3.5mm gap for the gas to enter, these are situated around the outside of the lens coils providing magnetic fields in their own stanless steel cans and a 8KV supply hooked up, we have create a electron gun and ion pump combination pump. The number of tubes control the pumping rate it lasts 400 hrs at 1 x <math>10^-</math> <math>^4</math>torr but 40000 hrs at 1 x <math>10^-</math> <math>^6</math>torr . slotted cathode cells for argon gas collection 
14 inch pipe (NPS 14 min. SCH 20 ) with one 100mm thick Aluminum plate with interior "carved out" for hopper box sides and top, a further 304 plate for bottom of hopper box with o-ring seal, 304L Top and bottom pipe end caps if not domed min. 18mm thickness typical 25mm thick with copper CF flange and hinged window access.
The refill hopper will be situated in the side of the 14 inch OD diameter gun tube and will obscure a little part of the build chamber from the beam(a necessary compromise). (Technical background: See 5.2 see lecture 4.02/11/04,
- Viewing window/ Door 8inch borrsilicate glass, 3/8 thick, standard 10 inch CF plate and Oring.
Solar cell production demands conductive transparent top surface layers Indium tin oxides traditionally uses electron beam physical vapor deposition EBPVD . Gun details required.
OD 16.5 inch Glass tube/vacuum wall constructed in a two stage process, first printed within metalicarap by having a dedicated rotating center attached sand deposition head and dedicated build chamber/mould, is electron beam heated , making an under size 14inch OD tube so fits within existing glass tube chamber , secondary stage known as slumping (carried out without metalicarap) to extrude the tube out to final OD 16.5 inch using a former internally heated by gas burners ( the former printed in parts within metalicarap), which the tube is lowered over expanding and thereby achieving its large diameter final form.
Polishing Plasma through the addition of argon and a defocused beam, polishing out roughness from a typical powder printed finish Rq of 25 nm to a polished surface Rq of 4 nm is possible , there are other low tech polishing methods that may be easier eg. Tumble finishing. EBeam Polishing 
Ion Pump to remove waste vaporized metal.
Focused ion beam / Ion beam etching gun to take dimensions to sub µ level(1µ over 20mmIT0)(Future development issues, SEM measuring absolute distances over a 300mm depth of field in FIB mode .)