The various SFF building strategies and deposition/fusion processes include photolithography, laser fusion, lamination, extrusion, and ink-jet printing. The figures included with the descriptions of these strategies schematically represent these SFF systems. A more detailed description of machine designs, component technologies used to implement each system, and CAD aspects will be presented in following chapters. Table 2.1 summarizes commercialized SFF systems.
Photolithography SFF systems build shapes using light to selectively solidify photocurable resins. There are two basic approaches: laser photolithography and photomasking. The laser photolithography approach depicted in Fig. 2.5, which is currently the most widely used SFF RP technology, was first commercialized by the U.S. company 3D Systems. Not only was 3D Systems the first company to successfully commercialize the stereolithography process, but the company must also be credited with both popularizing RP and establishing a marketplace for RP technologies.
Fig. 2.5. Laser photolithography.
Laser photolithography systems have since been developed and manufactured in both Europe and Japan (Fig. 2.5 and Table 2.2). With the exception of Kira's Solid Center, all RP machines manufactured in Japan are based on laser photolithography. While most laser photolithography systems use the building strategy represented in Fig. 2.5, there are significant differences in machine implementations, particularly in the recoating and beam delivery mechanisms and in the lasers. Chapter 9 describes these implementations in detail.
Laser photolithography creates acrylic or epoxy parts directly from a vat of liquid photocurable polymer by selectively solidifying the polymer with a scanning laser beam. Parts are built up on an elevator platform that incrementally lowers the part into the vat by the distance of the layer thickness. To build each layer, a laser beam is guided across the surface (by servo-controlled galvanometer mirrors, for example), drawing a cross-sectional pattern in the x-y plane to form a solid section. The platform is then lowered into the vat and the next layer is drawn and adhered to the previous layer. These steps are repeated, layer-by-layer, until the complete part is built up.
Since the photopolymers are relatively viscous, simply lowering the elevator the small distance of the layer thickness (i.e., ~.002 in. to ~.020 in.) down into the vat does not permit the liquid to uniformly recoat the upper surface of the part in a timely fashion. A recoating mechanism is therefore required to facilitate this process. For example, 3D System's stereolithography uses a "deep dipping" recoating, whereby the elevator is first lowered several millimeters so that the liquid entirely flows over the current upper surface of the part. The elevator is then raised to the desired height and a "doctor-blade" (wiper arm) traverses the surface to quickly level the excess viscous material.
With laser photolithography, features with gradually changing overhangs can be built up without support structures. Large overhanging features, however, require supports, since the initial thin layers that form them can warp or break off as the part moves down into the liquid. The supports are typically built up as thin wall sections that can easily be broken away from the part upon completion.
There are a few laser photolithography systems that build using slightly different approaches, such as those depicted in Fig. 2.6. In the machine manufactured by Denken Engineering of Japan shown in Fig. 2.6a, the part is built inverted, attached to a platform that rises up as each successive layer is drawn, and attached to the bottom-most face. The liquid resin layer is deposited on a specially prepared window that is transparent to the laser and to which cured polymer adheres poorly. The platform and prior-built structure are lowered into the resin, leaving between the part and the plate a liquid film having the correct thickness for the next layer. The new layer is drawn from beneath the plate. After the layer is drawn, the new structure is raised, separating the layer from the plate, and the process is repeated until all layers are fabricated. Mitsui Corporation in Japan, which the JTEC/WTEC team did not visit, also manufactures a process, "COLAMM," that scans from below.
Fig. 2.6. Other laser photolithography approaches.
The process represented in Fig. 2.6b is being developed by Professor Koji Ikuta of Nagoya University and optimized for producing microscale parts -- the laser spot size is 5 µm. (The panel did not visit with Prof. Ikuta). In this system a transparent plate is lowered into the vat to form a thin layer of liquid film over the part being built up. The growing part remains stationary in the vat, and the vat is moved relative to a fixed laser beam that passes through the plate, drawing the cross-section. The laser moves up with the plate to maintain precise focusing on the film layer.
In contrast to "drawing out" each cross-section with laser photolithography, it is possible to image an entire cross-section in a single operation using photomasks. This approach was originally developed and commercialized by Cubital (Israel/Germany).2 The Cubital system, called Solid Ground Curing (SGC), is depicted in Fig. 2.7. In SCG, each cross-section is imaged onto an erasable mask plate produced by charging the plate via an ionographic process and then developing the image with an electrostatic toner (e.g., like the Xerography process). The mask is then positioned over a uniform layer of liquid photopolymer, and an intense pulse of UV light is passed through it to selectively cure the material. Uncured photopolymer is removed from the layer with a vacuum system and replaced with a low-melting-point, water-soluble wax that serves as the sacrificial support. After the wax has cooled, the layer is milled to produce a flat surface. The pattern on the exposed mask is erased by wiping off the toner, and the entire process is repeated. After the part has been completed, the wax is removed by melting. The various processes used to implement SCG are performed at different stations.
Several systems use lasers to selectively fuse powdered material to build up shapes. The "selective laser sintering" approach depicted in Fig. 2.8 was originally developed at the University of Texas at Austin and then commercialized by DTM Corporation (U.S.). EOS, Inc. (Germany), has also developed and marketed its own laser sintering machines. The Fraunhofer Institute for Production Technology (IPT) has also produced an experimental laser-sintering-type unit designed for direct metal sintering.
In these systems, a layer of powdered material is spread out and leveled over the top surface of the growing structure. A CO2 laser then selectively scans the layer to fuse those areas defined by the geometry of the cross-section; the laser energy also fuses layers together. The powders are joined by a variety of fusion mechanisms, including melting, surface bonding, sintering aids, and polymer coatings. The unfused material remains in place as the support structure. After each layer is deposited, an elevator platform lowers the part by the thickness of the layer, and the next layer of powder is deposited. When the shape is completely built up, the part is separated from the loose supporting powder. Several types of materials are in use, including plastics, waxes, and low-melting-temperature metal alloys, as well as polymer coated metals and ceramics for making "green" preforms. Direct fusing of metals and ceramics (i.e., uncoated powders) is also being investigated.
While both the DTM and EOS machines are based on the same underlying methodology, there are significant differences in machine implementations, including their respective material delivery approaches. The DTM machine delivers powder from a cylinder adjacent to a second cylinder in which the part is grown; a roller is used to spread and level the powder. In the EOS system, powder spreading is through a slit nozzle whose leading and trailing edges are contoured as the head is vibrated from side to side.
There are currently two commercialized SFF lamination systems. Laminated Object Manufacturing (LOM) is a lamination method that was developed and commercialized by Helisys Corporation (U.S.). LOM builds shapes with layers of paper or plastic (Fig. 2.9a). The laminates, which have a thermally activated adhesive, are glued to the previous layer with a heated roller. A laser cuts the outline of the part cross-section for each layer. The laser then scribes the remaining material in each layer into a cross-hatch pattern of small squares (see insert in Fig. 2.9), and as the process repeats, the cross-hatches build up into tiles of support structure. The cross-hatching facilitates removal of this tiled structure when the part is completed. LOM builds up large parts relatively rapidly because only contours are scanned. LOM is also being investigated by Helisys and the University of Dayton for building up ceramic and reinforced composite shapes using layers of "green" tape castings (i.e., sheets of bound powder); the final part must subsequently be sintered.
The only nonphotolithographic SFF system being produced in Japan is a lamination system manufactured by Kira Corporation. While Kira's basic building approach is the same as that used for the Helisys LOM machine, the Kira Solid Center (SC) machine (Fig. 2.9b) is implemented in a significantly different manner. The SC machine uses standard printing paper that is fed into the machine using a conventional laser printer. The printer uses an adhesive-based toner to print the outline of the cross-section as well as a cross-hatched bonding pattern on each piece of paper. A hot plate then laminates the paper to the previous layers. The cross-sectional outline is then cut with a carbide knife that is mounted on a swivel base. Additional segments of "parting-plane" sections are also cut to facilitate removal of the support material.
Internal cavities are hard to form with the lamination systems described above, since it is difficult to remove the sacrificial material from the internal regions. To address this issue, Case Western Reserve University and CAMLEM, Inc. (U.S.), are developing a lamination system using green tape castings with a separate supporting material fugitive tape. Each section is individually cut with a laser and then stacked in place. The fugitive tape is then burned out during the final firing process.
Extruding freeform shapes was first developed and commercialized by Stratasys, Inc. (U.S.). This approach, called Fused Deposition Modeling (FDM), deposits a continuous filament of a thermoplastic polymer or wax through a resistively heated nozzle (Fig. 2.10, top left). The material is delivered as a wire into the extrusion head and heated to slightly above its flow point so that it solidifies relatively quickly after it exits the nozzle. It is possible to form short overhanging features without the need for explicit support; in general, however, explicit supports are needed. These are drawn out as thin wall sections that can easily be removed upon completion. Various U.S. investigators, including Rutgers University, Allied-Signal, Lone Peak Engineering, and Advanced Ceramics Research, are also exploring the use of FDM with thermoplastic wires and rods loaded with ceramic powders to build "green" preforms.
Multiphase Jet Solidification (MJS) is another extrusion-based process (Fig. 2.10, bottom left). MJS is being jointly developed by the Fraunhofer Institutes for Applied Materials Research (IFAM, Bremen) and Manufacturing Engineering and Automation (IPA, Stuttgart). IPA is working on software development and IFAM on material aspects. MJS extrudes metal or ceramic slurries using metal injection molding technology. The slurry, which is about a 50/50 mixture of wax and metal or ceramic powder, is contained in a heated vessel and pumped through an attached nozzle with a screw-activated plunger. The JTEC/WTEC team's hosts at IPA mentioned that commercialization may be through Fockele and Schwarze. The Fraunhofer Institutes have a German patent for MJS and are applying for a U.S. patent. The basic methodology of depositing 3D shapes by extrusion, in layers, is considered to be public domain. The Fraunhofer patent pertains to the feedstock (slurry composition), the heated material supply (up to 200ºC to liquefy the binder and obtain the desired viscosity), and the extrusion nozzle.
Several SFF processes have taken advantage of ink-jet printing technology to print layers of structures. The first process that successfully demonstrated "printing" of shapes was the Three-dimensional Printing (3DP) process, depicted in Fig. 2.11a, which was developed at MIT as a method to form "green" preforms for powdered metallurgy applications. While different powdered materials can be used, 3DP is currently commercialized by Soligen Corporation (U.S.) under the name Direct Shell Production Casting (DSPC) for creating ceramic shells and cores for casting applications. In 3DP, the part is built up in a bin that is fitted with a piston to incrementally lower the part into the bin. Powder (such as alumina) is dispensed from a hopper above the bin, and a roller is used to spread and level the powder. An ink-jet printing head scans the powder surface and selectively injects a binder (such as colloidal silica) into the powder. The binder joins the powder together into those areas defined by the geometry of the cross-section. The unbound powder becomes the support material. When the shape is completely built up, the "green" structure is fired, and then the part is removed from the unbound powder. 3DP of metal powders, such as stainless steel bound with a polymeric binder, is also being explored; subsequent infiltration of the matrix is then required for densification.
Other processes use ink jets to directly deposit low-melting target materials. Ballistic Particle Manufacturing (BPM), which was developed and commercialized by BPM Technology, Inc. (U.S.), uses a piezoelectric jetting system to deposit microscopic particles of molten thermoplastic (Fig. 2.11b). Like FDM and SLA, support structures are required for "unconnected" features. The supports are deposited in a perforated pattern to facilitate removal. The BPM jet head, however, is mounted on a 5-axis positioning mechanism so that overhanging features can be deposited without support, as the figure shows.
The Model Maker system (Fig. 2.11c) of Sander's, Inc. (U.S.), dispenses both a low-melting-temperature thermoplastic and a separate wax support material. In addition, it incorporates a slab cutter to plane each layer to the precise thickness.
In other commercial developments, 3D Systems, Inc. (U.S.), has just introduced a new ink-jet prototyping system, "Multi-Jet Modeling," (Fig. 2.11d), which uses a printing head with 96 individual jets that deposits a low-melting-temperature thermoplastic. Using the same material the support structure is deposited as thin, needle-like structures.
In other research efforts, the Technical University of Munich (not visited by the panel) is developing a "modified" three-dimensional printing process that injects streams of UV-curable binder resins under a UV light. Also, investigators at MIT and the University of California, Irvine, are developing jetting systems to deposit metal alloys.