Transforming a conceptual design into a solid prototype used to take weeks and was primarily a job for highly skilled machinists. Now, rapid-prototyping and rapid-tooling techniques can duplicate these efforts in hours or even minutes. Broadly speaking, these techniques either add material to something to build a product or subtract it to shape a part using both laser and nonlaser-based techniques. The prototyping systems receive instructions directly from computer-aided-design (CAD) data.
The most common techniques are additive and build parts one layer of material at a time.1 They differ in the materials and methods of deposition, curing, solidification, and bonding. Additive techniques include stereolithography-in which a laser activates a photochemical polymer to solidify it layer by layer into the shape of the prototype; laminated object manufacturing-in which a laser cuts successive sheets of material that are then built up into a part; and selective laser sintering-which laser-melts particles together (see Fig. 1).
These techniques share common limitations. One is difficulty in creating highly smooth surfaces, which can be attributed to the lamination technique involved. It also may be necessary to build a support structure for parts with overhangs. Removal of the supports can leave rough sections on the prototype.
Fed by computer-aided-design data, selective laser sintering builds prototypes or tooling by melting powder, including plastic-coated metals, together, one cross section at a time. Using rapid-prototyping equipment, one European job shop laser-sintered steel tool inserts in about 5.5 h (right).
null
There are fewer subtractive rapid-prototying methods. The only laser machining technique accurate enough for prototype formation is laser chipmaking or cavitation, in which a laser produces small, accurately shaped chips, gradually carving a shape out of a workpiece. The principal nonlaser competition is computer numerically controlled (CNC) machining, which machines parts with a hard insert.
Both approaches have difficulties creating parts with internal cavities. The traditional CNC process can produce a finer, more accurate surface than additive processes, but at the expense of processing time. It can also easily machine a prototype or tool required to mold a part out of metal. Some additive rapid-prototyping methods, however, may be limited to working with nonmetals.
Building prototypes slice by slice
The first and still one of the most common prototyping techniques is stereolithography, which was pioneered by 3D Systems (Valencia, CA). The process feeds off CAD data that have been converted into stereolithography file format (.STL), which is basically a triangular tessellated surface description of the prototype geometry. The CAD software mathematically slices the object into cross sections and then defines a tool path for the laser to follow in each layer.
To build the prototype, an ultraviolet (UV) laser, either ion or excimer, follows the tool path across a thin layer of liquid epoxy resin (0.004 in. is typical), photocuring only the area traced. The worktable is then lowered, more uncured resin is swept over the top surface of the part by a blade, and the next slice or cross section of the part is cured by the laser. The process repeats until the prototype is complete. The part is then removed, cleaned (including removal of support structures), and further cured in an ultraviolet oven. Depending on the complexity of the part, the build process can take hours or days, instead of the weeks common with conventional prototyping processes.
Another laser-based rapid-prototyping technique is selective laser sintering from DTM Corp. (Austin, TX). This also builds solid three-dimensional (3-D) objects layer by layer, but from plastic, metal, or ceramic particles that are sintered or fused with a CO2 laser. Once the laser has scanned an entire cross section, another layer of powder is added on top, and the process repeats. Typical layer thickness is 0.1 mm, with a 30-50-µm grain size.
The powder can consist of thermoplastic polymers, zircon sand, or polymer-coated metal powder. The system sinters the coated metal powder, which can be baked out and backfilled with a lower-temperature metal. Basically, after the polymer coatings (about 5 m thick) are melted, the form of the part is complete, but it is very porous. To fill the part in, a water-soluble polymer binder can also be infiltrated by the capillary effect. The part is then dried and the metal melted into a finished part with a conventional oven. Other steps may also be necessary to produce a finished, dense part.
Unlike stereolithography, which requires users to build supports for overhangs simultaneously with the prototype, the DTM process has a self-supporting build envelope. Support removal is simplified because the "support" is basically powder that can be reused.
The third major laser-based prototyping process is laminated object modeling from Helisys (Torrance, CA). This technique builds prototypes with composite sheets that are randomly oriented glass fibers mixed with ceramic materials and covered with thermoplastic binders. Sheets are placed on the worktable one at a time and cut with a CO2 or other infrared laser. After one sheet is cut, another is then laminated on top of it. The two basically stick together when the binder and an epoxy underlay fuse together. The laser cuts the sheets in such a way that the excess material can be removed as cubes.
Competition stacks up
Two nonlaser-based additive systems compete with these laser-based approaches for market share. These are inkjet modeling and fused deposition modeling (see Fig. 2), which recently overtook stereolithography in annual unit shipments worldwide.
Just as inkjet printers still compete with laser printers for two-dimensional reproductions, so does Sanders Prototype Systems (Wilson, NH) use inkjets to compete with laser-based prototyping techniques in building 3-D prototypes. In the process, an inkjet deposits thermoplastic and wax materials layer by layer to form the prototype.
Prototypes built by the inkjet process are quite accurate. Free-standing walls can be as thin as 0.004 in., and the vertical axis is determined to within 1/8000 in. Perhaps because of this accuracy, other companies have entered the inkjet modeling business, including 3-D Systems and Z Corp. (Somerville, MA).
Fused deposition modeling developed by Stratasys Inc. (Eden Prairie, MN) uses a thermoplastic filament as feed stock when building models. Fed via rollers from a spool into a heat liquifier, the filament acts as a piston to extrude molten material out of a nozzle onto a worktable, once again building the prototype cross section by cross section.
Fused deposition modeling, a nonlaser-based rapid-prototyping method, builds models using a polymer filament. This is fed through a computer-guided liquifier, turned into liquid polymer, and then deposited on a worktable to form prototypes.
null
A subtractive competitor, albeit still based on a layering principle, just emerging on the market is laser cavitation or ablation (see Fig. 3).2 The process, developed in Germany, removes ferrous metal such as steel one chip at a time with a high-power laser operating in pulse mode with 750 W to 1 kW power and pulse-repitition frequency up to 100 kHz. With each pulse, a small section of metal is heated and melted. This immediately oxidizes, causing a sudden expansion and doming up of the material. When the laser shifts to the next spot, the chip area starts to cool down, with far more rapid cooling at the bottom. This section of the chip shrinks quickly and separates from the work material. The chip then bends upward away from the part surface.
In the Lasercav process, a pulsed high-power laser heats a steel workpiece (left), and oxidation expands the heated region to some 100-300 ?m in depth to produce a metal chip (middle). Subsequent cooling causes the chip to separate and bend outward (right).
null
Called Lasercav by its developers, this process removes material layer by layer with a typical chip thickness of 100 µm. Finishes as low as 5 µm are possible, as is textured patterning of the surface. In contrast to additive processes that produce primarily plastic models, laser cavitation can produce tools such as forging dies without intermediate steps in less than a day, depending on the tool. One company producing Lasercav machines is LCTec GmbH (Pfronten, Germany).
Moving into medical prototyping
The main applications for rapid-prototyping equipment still include creating prototypes and tools for injection molding. Nevertheless, interest is growing in niche markets, such as medical applications. One example is the Phidias project funded by the European Union, which has demonstrated the ability to make 3-D medical models of human organs and bones with stereolithography, working with data provided by medical imaging technologies such as magnetic resonance imaging.3 With such models, surgeons can rehearse their operations on life-size molds of a patient`s skull, liver, and so on.
In the semitransparent models, structures can be outlined with different colored plastic to distinguish bone from flesh or a tumor from healthy tissue. One company involved in Phidias, Materialise (Lueven, Belgium), hopes to move on to the next step and use stereolithography to manufacture prostheses or implants directly from scanned images with input from surgeons.
REFERENCES
- S. C. Danforth and A. Safari, Proc. IEEE International Symposium on Ferroelectrics, 183 (Aug. 1996).
- D. Smock, Plastics World, 45 (Aug.1996).
- S. Mraz, Machine Design, 36 (Feb. 6, 1997).