This article reviews that an ultimate goal of 3D printing is direct manufacturing. A critical element to the commercial acceptance of 3D printing is the need to make high production rates. Theriform benefit derives from the way in which a tablet can be built layer by layer in any number of configurations. Concentric rings of pharmaceutical can be interspersed between rings of inert powder and thus create timed release in a lone oral tablet. The Theriform process requires a different approach to pill-making than most pharmaceutical manufacturers are used to. Soligen builds ceramic molds for metal casting with 3D printing, eliminating the pattern-making step. TDK is investigating the production of ceramic components for electronic devices. Z Corp. makes 3D printers for building rapid prototypes from inexpensive and benign materials. Using the Z Corp. printer, a designer can print out a model of a part with as much effort as it takes to generate a paper drawing.
What would it mean to an accident victim, suffering disfiguring injury to the skull, if an implant could be made that exactly duplicates the geometry of the patient’s eye orbit? Or what would it mean if pills could be made with active ingredients so precisely positioned within inert material that the medicine is dispensed automatically according to an exact timetable? Such seemingly unrelated objects are both made with an additive manufacturing process called Theriforming. The process, according to Christopher Gaylo, vice president of engineering for the Princeton, N.J., developer, Therics Inc., is true rapid manufacturing.
Therics is one of six companies that have licensed the three-dimensional printing technology that was invented in 1993 by MIT professor Emanuel Sachs. According to Sachs, every license is in a different area of manufacturing. While Therics concentrates on medical and pharmaceutical applications, another licensee applies 3-D printing to die-casting molds. Another markets the process for rapid prototyping.
Sachs said that in every case the basic 3-D printing is the same. Every licensee uses an inkjet print head, be it either drop on demand, which dispenses discrete amounts of binder, or continuous, which dispenses an uninterrupted binder stream.
According to Sachs, the binder is deposited onto a layer of powder. The powder, whether ceramic, metal, or polymer, then fuses into a thin section of the desired shape. Unbound powder remains around the fused stratum. With one layer down, another powder layer is dispensed over it. More binder is applied. In this fashion, an object emerges, layer by layer.
An ultimate goal of 3-D printing, Sachs said, is direct manufacturing. That’s just what Therics is doing, along with 3-D printing licensee Specific Surface Corp. of Franklin, Mass. A critical element to the commercial acceptance of this form of 3-D printing is the need to make high production rates.
Gaylo said the Theriform process does well in this regard when compared with the conventional manufacturing of multilayered tablets on triple compression presses.
Where the Theriform process has no match is in its power to create solid state drug delivery forms with various release profiles, Gaylo said. Some pharmaceuticals are so potent that mixing them with inert materials sometimes produces a random distribution of the active ingredients. At high dilution, he said, dry blending reaches a thermodynamic limit of uniformity. If the dose is small enough, the ratio between active and inert ingredients cannot be controlled from pill to pill. Theriforming, in contrast, can place a single microgram of active pharmaceutical dead center in the volume of a tablet. Every tablet thus made metes out the exact same dosage.
Another Theriform benefit derives from the way in which a tablet can be built layer by layer in any number of configurations. Concentric rings of pharmaceutical can be interspersed between rings of inert powder, Gaylo explained, and thus create timed release in a lone oral tablet.
Indeed, the Theriform process requires a different approach to pill-making than most pharmaceutical manufacturers are used to, Gaylo said. From the start, drug developers think in terms of the mechanical design of a tablet—its architecture, so to speak—rather than along the lines of a batch sheet or a process record. The CAD model becomes the principal design carrier, replacing any recipe by which drug developers might normally work. The tablet is, in a sense, a mechanical assembly.
Powder on a Cookie Sheet
A production run starts with a layer of powder spread over a “cookie sheet,” Gaylo explained. A carrier, or build platform, holds the sheet. Between successive layers of powder, the carrier retracts about 250 microns. From above, inkjet nozzles disperse binder droplets to the powder. By using more than one nozzle, binders of different constituents are directed down to the powder at 800 drops a second. A sensor counts droplets to maintain accuracy.
A roller spreads the powder from the feed bed over to the build bed, Gaylo said. To assure a uniform layer thickness, the roller rides on the outer edges of the feed and build beds. This “local registration,” he said, takes out any variation between layers, except that due to roller runout.
As for motion control along the x, y, and z axes, the 3-D printer uses a direct-reading linear encoder with 1 micron resolution, Gaylo said. That produces a positioning accuracy of about 10 microns.
No heat is used to bind the powder in the Theriform process. Instead, chemical energy fuses volumetric elements, or “voxels.” Heat can damage the pharmaceutical and biocompatible materials, Gaylo said.
The same procedure for making pharmaceutical products works to make medical devices as well, Gaylo said. Even the steps before and after manufacturing are similar. But where the shape of a multistratum tablet might come straight from a designer’s imagination, the shape for an implantable eye orbit, for example, would come off an MRI or CT scan. From there, one of several digitizing programs renders the data usable by a solid modeler, Gaylo said. Then, the solid model is sliced into sections that will correspond to layers of powder on the build platform. Next, a Therics CAD linking program writes machine instructions. From there, the Theriform machine produces an implant in the same manner that it makes a tablet.
Implants are made from coacervated tricalcium phosphate or demineralized bone. According to Gaylo, the process could use other materials that would provide trellises through which new tissue could grow. Material such as resorbable polymer gives the body a frame around which to build new tissue. Eventually, such material is completely absorbed by the body, he said.
Another advantage of Theriforming carries over from the pharmaceutical side. Just as a batch of tablets is made using different binders, both active and inert, implants can be manufactured using different polymers so as to encourage cell migration into specific areas of the part.
Therics’ commercialization plan calls for the introduction of standard sizes of biocompatible bone replacements in 2002. From there, the company will expand the line into custom implantables, concentrating on products for bone and cartilage, Gaylo said. Making artificial organs may be the ultimate goal, he added.
On the pharmaceutical side, Therics is working with a number of major pharmaceutical makers to bring these novel drug delivery systems to market.
Already making a commercial product, Specific Surface uses 3-D printing to produce industrial filters. Such filters offer a number of advantages over filters produced by conventional extruding or molding. Said the company’s CEO, Mark Parish, “We put a lot of work into making a reliable manufacturing device because we can’t live with any flaws in filters, such as cracks or misprints.” Like the Theriform process, Specific Surface’s 3-D printing technology, which it calls Ceraprinting, makes full use of the technology’s capacity for controlling the microscale placement of materials.
Specific Surface uses 128 inkjets, printing continuously, to deposit its binders to ceramic and metal powders. Unlike Therics’ binders, Specific Surface’s materials rely on heat to fuse. A secondary sintering process at several thousand degrees Centigrade follows the printing step to impart final strength to a filter, Parish said. He made a distinction between the two binders his company uses. Reactive binders combine with ceramics during heating to become part of the filter, he said. Nonreactive binders simply evaporate or burn during heating.
Like Therics, Specific Surface is in the business of rapid manufacturing: It builds its 3-D printers to provide filter products. Parish said the company could increase production rates on its machines by dedicating each to a specific product. By taking away the flexibility of the system, eliminating its capacity to make products in a range of shapes, Specific Surface could reduce the time to build a layer by one or two orders of magnitude, he said.
For now, however, Specific Surface gains productivity by eliminating manufacturing steps downstream. For example, a diesel filter typically consists of columns of channels. Channels are plugged on alternating ends so that exhaust soot is forced through the filter media.
Filters that are extruded or molded must be plugged in a secondary operation. Instead, Specific Surface adds these plugs during the build, thereby eliminating the labor-intensive activity of installing them later. For now, Specific Surface is making diesel filters for trucks and buses. It is building filters for scooters. The high-volume business of automotive filters is likely to come, Parish said.
Another strength of Specific Surface filters is in the ability of the printing process to control the arrangement of particles. A more uniform pore size is possible than with molded or extruded filters. Explained Parish, stress rearranges particles in a ceramic extrusion or mold because the powder must first be mixed with liquid, then moved around. Better control of the particle arrangement means that filters made by 3-D printing can produce more uniform voids, leading to higher flow rates and efficiencies, and lower pressure drops.
Specific Surface also makes filters having tapered channels. The pressure drop in such a channel is even along its length, Parish said. A tapered channel “fills from the back, forward, preventing clogging,” he said.
A straight channel behaves more like a lint screen in a dryer, he said. “Lint can’t pass through it,” he said. “It forms a barrier layer that catches other lint. If those were channels instead of a screen, it would clog up right in the front of the filter. But if you could imagine cone-shaped channels that are big at one end, the lint can travel in pretty deep before it begins clogging up,” he said. Parish said he knows of no other process, except possibly building it by hand, that can make a tapered channel.
The company has found a way to make candle filters as a stack of modules rather than as single, monolithic castings or extrusions. The typical ceramic filter, 2 to 3 meters long, costs several thousand dollars. Mishandling can damage the filters. Or stray material can flow in and crack their brittle surfaces.
Specific Surface’s modular candle filter uses a tongue-and-groove joint to interlock the segments. It is easy to replace any of the segments, which are held together with a stainless steel rod running down the middle of the stack. Another design advantage of the segmented candle filter is that the rod places the assembly in compression, a preferred state for ceramics, Parish said.
Neither Specific Surfaces nor Therics is alone in developing MIT’s 3-D printing. Parish said that the six licensees, which also include Extrude Hone Corp. of Irwin, Pa.; Soligen Inc. of Northridge, Calif.; TDK Corp. of Chiba-ken, Japan, and Z Corp. of Burlington, Mass., meet once or twice a year to compare notes.
Extrude Hone uses 3-D printing to build tooling from steel and bronze. After printing a part, the Extrude Hone process places it into a sintering furnace, then into a second furnace where molten bronze is drawn in, producing a near-net part.
Soligen builds ceramic molds for metal casting with 3-D printing, eliminating the pattern-making step. TDK is investigating the production of ceramic components for electronic devices.
Z Corp. makes 3-D printers for budding rapid prototypes from inexpensive and benign materials. Using the Z Corp. printer, a designer can print out a model of a part with as much effort as it takes to generate a paper drawing.