U.S. Department of Energy Office of Science Office of Scientific and Technical Information

In the OSTI Collections: 3-D Printing and Other Additive Manufacturing Technologies

 

Understanding Electron Beam Melting

Is Additive Manufacturing Suitable?  Two Projects

Other Materials

Additive Manufacturing Technologies

References

Reports Available Through OSTI's SciTech Connect

Additional References

Until recently, manufacturing has largely been subtractive.  That is, one chops, saws, chisels, drills, grinds, etches, polishes, or otherwise subtracts pieces from a material to make parts, and then assembles the parts.  But in the last several years, new “additive” manufacturing techniques have been developed, in which each individual part is built into the desired shape by adding a layer at a time instead of subtracting from something bigger. 

 

There are several additive processes, but many of them resemble printing:  in these, each layer of a part is in effect printed on top of the layer below like ink on a piece of paper.  Strictly speaking, even “two-dimensional” ink on paper is actually three-dimensional, but the length and width of paper are much bigger than the one-to-few molecule thickness of printed ink.  “3-D printing”, on the other hand, results in components whose obvious thicknesses are more comparable to their lengths and widths. 

 

A report by scientists at the Manufacturing Demonstration Facility of Oak Ridge National Laboratory describes two advantages of additive manufacturing.  One is the lack of waste:   

For many subtractive machining processes … 90% or more of the original billet is machined away leaving only 10% of the material as the final product.  …there is a lot of waste in time, energy and materials to make parts by traditional subtractive machining methods. AM [additive manufacturing], in contrast, uses nearly all of the basic material to form the parts resulting in virtually no waste ….  (“The Use of Additive Manufacturing for Fabrication of Multi-Function Small Satellite Structures”[SciTech Connect], p. 3.) 

 

The second advantage is even greater: 

Because of the revolutionary capabilities of additive manufacturing in three-dimensional structures, new unheard of designs can be developed and fabricated in monolithic structures. No longer are the designers constrained by the rule ‘it must be designed so it can be fabricated’. The new generation of designers is just beginning to realize the expanded potential of AM coupled with computer-aided design (CAD) to revolutionize design and system integration tasks. … The real breakthrough is that design approaches that could not be used for subtractive manufacturing (such as embedded complex passages and reinforcing struts and embedded mesh) are now possible with AM. Designers of the past were constrained by the edict ‘design it so that it can be manufactured’ (by traditional machining equipment). Designers of the future will not have these constraints and will be bounded only by their imaginations and creativity.  (Ibid. [SciTech Connect], pp. 2-3.) 

 

While additive techniques are not yet available for every type of manufacturing, the techniques continue to be advanced.  Several recent reports available through OSTI’s SciTech Connect describe some of the progress. 

Understanding Electron Beam Melting

Figure 1.  Schematic diagram of the Arcam Electron Beam Melting system.  (P. 2 of 9, “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing”[SciTech Connect].) 

 

One additive technique for making dense metal parts, called Electron Beam Melting, was developed by the Swedish company Arcam AB and is illustrated in Figure 1, which is taken from the report “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing”[SciTech Connect] by researchers at Oak Ridge National Laboratory (ORNL) and the Oak Ridge Institute for Science and Education (ORISE).  As this report describes the technique,

 

The EBM process, as developed by ARCAM in Sweden, typically occurs in an unheated vacuum chamber, also known as the build chamber (see vacuum chamber in Fig. 1). At the top of the chamber is a computer controlled electron gun with sufficient current to melt the selected metal powder. The final part will be constructed layer-by-layer on the build platform that is located in the floor of the vacuum chamber. This platform is capable of being lowered in small increments, which will become the layer thickness during the build process. At the start of the process, this platform is lowered about 50 micrometers, and a layer of metal powder is raked across the floor of the chamber, filling the space above the build platform. An unfocused, low current electron-beam then uniformly pre-heats this layer of powder to a temperature slightly below the melt temperature of the powder. During the pre-heating phase the metal powder will weakly sinter together, holder the individual particles in place during the subsequent melting step. The pre-heat step is followed by the melting step wherein a high current focused electron-beam draws the pattern of the first layer of the part on the metal powder bed. This step only melts the powder directly under the electron-beam. The molten metal is held in place by the surrounding sintered powder and solidifies quickly. The build platform is then lowered allowing for the next layer of powder to be raked across. The process of lowering the platform, raking powder, pre-heating and melting continues until the part is completed. When the chamber is opened, the part is embedded within a sintered block of powder. This block of material is placed inside a grit blaster, which uses the same metal powder as the blasting media. The surrounding sintered powder is easily removed leaving the final part. The removed sintered powder is captured, filtered and recycled for future parts.  (P. 1.)

 

One example of an item made by Electron Beam Melting is shown in Figure 2. 

 

 

 

Figure 2.  An orthopedic hip implant (acetabular cup[Wikipedia]) in titanium, built with Electron Beam Melting. The implant has a tailored surface design for optimized bone ingrowth.  (P. 2 of 8, “Microstructural Properties of Gamma Titanium Aluminide Manufactured by Electron Beam Melting”[SciTech Connect].) 

 

While Electron Beam Melting is being used successfully, we don’t yet know the best ways to use it for every manufacturing problem.  Details of how the process is carried out affect the material’s microstructure, which significantly influences the material’s properties.  How some variations in these details affect one particular material is described in the report “Microstructural Properties of Gamma Titanium Aluminide Manufactured by Electron Beam Melting”[SciTech Connect] from scientists at ORNL and at Arcam AB.  The material studied, gamma titanium aluminide (g-TiAl), has a long-range ordered crystal structure which makes it strong as well as resistant to oxidation and corrosion at high temperatures.[Wikipedia]  This particular material is of interest because of its potential for aerospace applications. 

 

The authors additively manufactured some round test bars 100 millimeters high and 15 millimeters in diameter from powdered Ti-48Al-2Cr-2Nb alloy whose particle sizes were between 45 and 150 micrometers.  200-micrometer layers of powder were melted at 1050°C to produce the bars.  The powder layers were melted in consecutive circles from the outside in, which is different from the Arcam machine’s usual consecutive straight-line melting pattern (Figure 3).  Some of the bars were later hot isostatically pressed[Wikipedia] at 1200°C and 100 megapascals[Wikipedia] of pressure for 4 hours before testing, while other bars were tested as built.  The bars were examined for microstructure, grain size, chemical composition, tensile strength, and ductility. 

 

Figure 3.  In the experiments described in “Microstructural Properties of Gamma Titanium Aluminide Manufactured by Electron Beam Melting”[SciTech Connect], g-TiAl powder layers were melted in circles from the outside in (left), a different pattern from the melting machine’s standard straight-line pattern (right).  (P. 3 of 8, “Microstructural Properties of Gamma Titanium Aluminide Manufactured by Electron Beam Melting”[SciTech Connect])

 

The authors found some positive and some negative effects of the processing.  The bars contained more oxygen than the original powder, apparently because powder that wasn’t immediately melted into objects accumulated oxygen as it was recycled through the melting machine over a long time before finally ending up in one of the bars.  The microstructure (Figure 4) was found to be duplex, with gamma phase faced by fine flakes (or lamellae [Wikipedia])—too many for an optimal microstructure[Wikipedia], though according to the authors, further development of the process parameters should yield an optimal microstructure.  The electron-beam melted material’s grain size is “magnitudes smaller” and its tensile strength high compared to that of conventionally-manufactured TiAl, but the ductility was lower than expected, possibly due to the experimental material’s higher oxygen content. 

 

 

Figure 4. Scanning electron micrograph of the center of a cross-sectioned g-TiAl bar built by Electron Beam Melting and afterward hot isostatically pressed[Wikipedia]. (P. 5 of 8, “Microstructural Properties of Gamma Titanium Aluminide Manufactured by Electron Beam Melting”[SciTech Connect])

 

Experiments like these can show correlations between processing parameters and physical properties of the product.  Understanding how changes in the parameters lead to differences in the properties is the subject of another ORNL slide presentation, “Prediction of Material Thermal Properties and Beam-Particle Interaction at Meso-Scale during Electron Beam Additive Manufacturing”[SciTech Connect].  The slides present an approach to the control of direct-metal additive manufacturing based on a mathematical description of its physical processes. 

 

The complete mathematical description involves several mathematical submodels, including submodels of material microstructure and of the part-building process.  Two submodels described at more length, which describe the transfer of heat among the material particles and the particles’ interaction with the electron beam, predict the beam-energy’s distribution and absorption over the powder bed and the powder bed’s effective thermal properties as a function of temperature and powder-consolidation stage.  These predictions provide input to the rest of the mathematical model to relate electron-beam and metal-powder parameters to various properties of the manufactured object, such as residual stress, shape distortion, and cracking. 

 

Knowing what properties a manufactured object will have if a given metal powder is exposed to a particular sort of electron beam is obviously useful.  To actually produce a given sort of object this way, it’s also important to know whether the process actually taking place really has the intended powder and beam parameters.  The aforementioned ORNL/ORISE report deals specifically with a challenge that Electron Beam Melting presents in ascertaining this, describes methods found to meet the challenge, and discusses benefits of the solutions. 

 

Since EBM is primarily a fast thermal process, it makes sense to monitor the process with a high-speed infrared camera. The IR camera is able to quantify the beam focus size, detect porosity and study the beam-powder interaction. ... Precise temperature measurements of the powder bed, solid parts and melt pool during the EBM process would be very useful in process control as well as model verification and development. However, before these measurements can be made, several challenges must be overcome to continuously image inside the chamber during the EBM process.

In order to implement Thermographic process-monitoring technologies, imaging must be performed through one of the two existing viewing ports in the vacuum chamber. Unfortunately, there are free metal ions released from some metal powders during the build process, and deposition of these ions will coat and opacify an unprotected view port window.

(P. 2, “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing”[SciTech Connect].)

 

The Arcam machines have a window whose inner surface is usually protected from this shower of metal ions by a shutter.  However, when the shutter is opened to adjust the electron beam during setup or monitor build progress, some coating by the ions does occur which gradually obscures the window, which makes temperature calibration by infrared camera “virtually impossible” and limits build observation time.  The window position, in the front access door, also makes spatial measurements “challenging” and prevents a sharp focus from front to back.  See Figure 5. 

 

Figure 5:  Infrared image acquired through the window in the front access door of an Arcam system.  “This image was taken with an unfiltered 50 mm lens on a FLIR SC-8200 midwave IR camera with an integration time of 1.0 ms. The front area of this image is in focus, but the rear is out of focus due to the limited depth of field of the lens. Also visible in the image are several ‘comet-like’ bright dots with bright tails. These dots are the E-beam showing up in various different locations during the 1ms exposure of the IR camera. The window became too opaque for imaging 3 minutes after this image was acquired due to metallization build-up on the inside. In order to over come the metallization of the window, a shutterless window protection system was developed.”  (P. 3 of 9, “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing”[SciTech Connect].) 

 

The investigators found two ways to overcome this problem:  spooling a Kapton[Wikipedia] film across the window’s vacuum side while the electron beam is on, and using a periscope over a view port at the top of the Arcam machine.  In the first method, the Kapton film becomes metallized instead of the window, but the exposed area is simply spooled out of the way so the camera always has a view through a fresh transparent portion.  As for the second method, although the periscope’s aluminum mirror is subject to metallization, the metallizing element is typically aluminum, so the mirror’s reflectivity isn’t too greatly reduced—12.5 hours of metallization reduces the mirror’s original reflectivity of 97.8% to 43.0%.  Maintaining an infrared view of the electron-beam melting makes it possible to spatially calibrate the width of the electron beam, easily detect and correct for overmelting in the metal that could otherwise lead to dimensional flaws in the overmelted regions, and similarly detect and correct for porosity during manufacture.  The investigators planned to develop a temperature calibration for the infrared images to allow the powder bed and melted surfaces’ temperatures to be measured, not only to use the measurements for feedback to the electron-beam control program, but to verify and refine mathematical models of electron-beam melting. 

Is additive manufacturing suitable?  Two projects

 

Studies of additive manufacturing techniques are not confined to understanding how they work or improving control over them; some studies are about when they would be useful.  As noted above, for some projects they’re not only useful, but the only practical means:  certain items would be impractical or impossible to make otherwise.  Two recent reports examine the suitability of additive methods for manufacturing two different types of device. 

 

One type is described by the title of the report first cited in this article’s introduction:  “The Use of Additive Manufacturing for Fabrication of Multi-Function Small Satellite Structures”[SciTech Connect]

 

Flying multiple spacecraft in formation has long been a dream of space scientists. The ability to fly many spacecraft in controlled formations is directly linked to the ability to maintain spacecraft spacing and the ability to synthesize new missions over multiple platforms simultaneously. Development of the technologies needed for satellite constellations has been hampered in the past by the cost of individual spacecraft and the cost to launch. ... What is needed for the development of spacecraft constellations and the associated technologies is a satellite class that is affordable to fabricate, is relatively inexpensive to launch, and has the capacity to test new technologies without an overwhelming cost penalty for failure. ... Small satellites are a much needed class of test satellites because they can afford to fail. It is only through failure and the inherent learning that accompanies well instrumented testing that new technologies will be introduced to the space community. This can directly support constellation and formation spacecraft applications such as remote earth observation, satellite repair monitoring and up-close satellite inspection.  (P. 1)

 

According to the report, the required sensor and propulsion capabilities might be integrated into a relatively simple multifunction structure that could be 80% smaller and 90% less massive than other types of satellites, and take less time to assemble.  But the size reduction alone has not made small satellites cheap enough to try out on a large scale. 

 

Multifunction structures have been pursued by government and industry research organizations in the past but positive results have been limited by manufacturing concerns and the inability to test a sufficient number of multifunction structures in space. … Overall, we are left with a promising approach for spacecraft optimization but with limited space-proven results and insufficient research activity in new multifunction spacecraft structure approaches.  (P. 2)

 

The report then points out that additive manufacturing may make experimenting with small satellites feasible.  “Often a technology concept meets a hurdle when there is not enough technology to enable efficient implementation of the concepts. In the case of multifunction structures, AM may well be the enabling technology that will allow the spacecraft industry to introduce operational multifunction spacecraft structures, beginning with small satellites.”  The new effort described in the report is intended to highlight the additive manufacturing R&D resources available at ORNL’s Manufacturing Demonstration Facility for industry collaborative development, through the design and fabrication of a notional multifunction structure that, while not intended for flight, complies with the CubeSat[Wikipedia] standards and includes integral propulsion and cooling systems in the structure itself.  At the time of the report (August 2013), a demonstration version of the design (Figure 6) was to be fabricated in titanium using Arcam’s electron-beam melting process. 

 

Figure 6.  Left:  a notional CubeSat[Wikipedia] rectangular pressure tank for cold gas or monopropellant.  Bottom:  three-unit CubeSat structure (left, transparent; right, solid).  Each CubeSat unit occupies one liter (10 cm X 10 cm X 10 cm).  (Pp. 3, 4 of 7, "The Use of Additive Manufacturing for Fabrication of Multi-Function Small Satellite Structures"[SciTech Connect].)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The report, “Distributive Distillation Enabled by Microchannel Process Technology”[SciTech Connect] on work by Velocys, Inc., Edison Welding Institute (now EWI), and the Mid Atlantic Technology, Research, and Innovation Center (MATRIC), describes a project for which the first additive manufacturing technique considered was found to be impractical at the current state of the art.  The project’s purpose was ascertain the industrial viability and benefits of microchannel distillation, an innovative liquid-liquid separation method that “uses a number of small distillation units in new configurations to greatly enhance process efficiency … by 25% or more compared to conventional single tower operations”, leading to a go/no-go decision on the process’ commercialization (p. 7). 

 

One of the project’s objectives, to make the manufacture of a commercial-scale microchannel distillation unit cost-effective, required evaluating different manufacturing techniques.  In the first technique considered, Ultrasonic Additive Manufacturing, thin strips of metal are bonded into a structure by high-frequency vibration.  The method is proven for softer materials like copper, but distillation units are typically made of stainless steel.  Edison Welding Institute did multiple experiments to identify suitable ultrasonic parameters for steel but found several outstanding challenges, chiefly:

 

  1. excessive transfer of stainless steel to the welding tool’s ultrasonic horn, resulting in “very short tool life” and
  2. varying bond quality, which was insufficient for a pressurized vessel;
  3. vibration of stainless steel foils resulting in cracking at the joint edges;
  4. more pronounced edge effects in stainless steel than in other materials;
  5. delamination of multilayer builds. 

 

EWI also tried laser welding.  The chief problem found with this technique was deformation from heat-induced stresses along the weld, which was solved by modifying the method of holding the assembly in place during welding.  Another technique was to make microchannels as a folded fin structure.  Such structures formed by sheet metal only or by sheet metal between wire mesh were firmer than mesh-on-mesh fins.  Well-formed fin structures had smaller fin heights and lengths, with more fins per unit width.  Both laser-welding and fin-folding were found “reasonably applicable” to manufacturing the distillation-unit conceptual design, while Ultrasonic Additive Manufacturing “did not show manufacturing promise and required long term development”. 

Other materials

 

Other reports show how additive manufacturing isn’t confined to metals. 

 

Additive methods for building objects from polymers[Wikipedia] are now being commonly used and extensively explored and advanced.  One such method is described in another report from investigators at ORNL’s Manufacturing Demonstration Facility. 

 

Computer designed 3D-models and parts can be printed using plastic as the “ink” resulting in rapid prototypes, models, and useful parts. Fused Deposition Modeling (FDM) is a specific manufacturing process in the general class of additive manufacturing, also known as 3-D printing. In FDM a thermoplastic filament is fed into a heated extrusion tip in the liquefier head. The liquefier head is moved under computer control in the X- and Y-plane as the molten polymer is extruded from the heated tip. The first printed layer is deposited on a disposable plastic build sheet held in place on the build platform by a vacuum. The build platform can move in the Z-direction and contains a grid of small holes through which a vacuum is applied to hold the release sheet in place. After each layer of thermoplastic is printed, the build platform is lowered by a small increment to allow the deposition of the next layer. As the molten thermoplastic is deposited on the thermoplastic below, it fuses to it making a solid bond. This process continues until the final model or part has been manufactured. (P. 1, “Real-time Process Monitoring and Temperature Mapping of the 3D Polymer Printing Process”[SciTech Connect].)

 

Aside from directly applying already-heated plastic polymer to the completed layers below instead of applying a metal powder and heating it with an electron beam, Fused Deposition Modeling resembles Electron Beam Melting.  The resemblance even extends to the temporary use of space-filling material at lower temperature to support protrusions of the object while it’s being built: 

 

The liquefier head contains two extrusion tips. The first one, as discussed above, is used to extrude the thermoplastic, which makes up the desired part. The second extrusion tip operates at a different temperature and is used to deposit fragile and dissolvable polymer used as support material. Support material is needed when the next layer to be printed would otherwise not have thermoplastic below to support it (i.e., overhangs, tubes, inverted bevels, etc.). The entire FDM process takes place inside a temperature controlled convection oven to help reduce thermal stresses in the parts being printed. (Ibid., p. 1.) 

 

Like the previously discussed report “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing”, the report just quoted describes an attempt to better understand a particular additive manufacturing process using data from an infrared camera.  Unlike the other investigation, though, this one didn’t involve a process whose side effects blocked infrared light.  Here, the investigators could proceed more directly to measure samples’ temperature variation during their manufacture and map the temperature within the oven’s build volume—data useful for quality assurance.  The map could indicate how manufactured parts’ properties depend on where they are built in the oven.  Observing how a part’s temperature varies during construction shows how the deposition process itself produces temperature distributions that can lead to the part’s failure. 

 

Figure 7. Temperature maps of the build envelope inside a Fortus 900mc Fused Deposition Modeling Production System manufactured by StrataSys, Inc.  Readings shown were taken (a) one inch from the top of the oven, (b) eleven inches from the top, (c) 21 inches from the top, and (d) 31 inches from the top.  (P. 5 of 9, “Real-time Process Monitoring and Temperature Mapping of the 3D Polymer Printing Process”[SciTech Connect].)

 

Also as in the case of electron-beam melting, these investigators developed two different ways to image Fused Deposition Modeling with infrared light.  One way views the entire build space through the oven’s front window with an extended range infrared camera, allowing measurements of temperature variations in the oven and in individual parts as they’re printed.  The other way uses a small lightweight infrared camera mounted on the liquefier head for a closeup view of the heated extrusion tip (Figure 8), allowing higher-resolution imaging and temperature measurements of the polymer as it’s extruded and begins cooling.  Temperatures were measured as the build platform was lowered.  The oven’s right side was typically hottest, with decreasing temperatures toward the bottom of the oven; the largest temperature difference measured was 28°C—not huge for an oven, though in temperate climates it would distinguish a summer from a winter day.  Thermal gradients in printed parts, which as we just noted can lead to distortions and thermal stresses, were also found.  Small design modifications in the extrusion tips were also found to have large effects on the extrusion temperatures. 

 

Figure 8. Top left: drawing of the liquefier head with the FLIR A35 infrared camera mounted on top, imaging down through a NaCl[Wikipedia] window at a gold mirror, which aims the camera at the extrusion tip.  Top right: the infrared camera mounted on the liquefier head, aimed down into the oven.  Bottom:  view of the extrusion tip and liquefier head from this camera.  (Pp. 4 and 8 of 9, “Real-time Process Monitoring and Temperature Mapping of the 3D Polymer Printing Process”[SciTech Connect].)

 

A slide presentation from Los Alamos National Laboratory (“An Overview of Polymer Additive Manufacturing Technologies”[SciTech Connect]) takes us from this specific problem of Fused Deposition Modeling to a more general look at several AM techniques for different kinds of materials (especially polymers).  According to this January 2014 presentation, there are 30 specific technologies available for depositing material and building parts, each representing one of seven basic additive methods.  Three of the seven methods were used in the projects described above:  powder bed fusion (exemplified by electron-beam melting), material extrusion (the type of process represented by fused deposition modeling of polymers), and sheet lamination (two methods being the laser welding and ultrasonic additive manufacturing tried by Velocys on stainless steel, with differing success).  The presentation lists the seven methods (Figure 9), briefly describes the nature and salient points of each, and enumerates any instruments for each method that Los Alamos National Lab has.  The methods described most extensively in the slides are powder bed fusion, material extrusion, and material jetting, which involves selective deposition of build-material droplets of one or more different materials (Figure 10). 

Additive manufacturing technologies (from ASTM F2792-08)

 

 

 

 

 

 

 

 

 

 

Figure 9.  Standard names of seven basic additive manufacturing technologies listed in “An Overview of Polymer Additive Manufacturing Technologies”[SciTech Connect]. 

 

Figure 10.  Schematic of a material jetting system for photopolymers[Wikipedia].  In this system the polymer layers are printed as jets of ink and hardened (cured[Wikipedia]) by exposure to light. 

 

 

While the author’s attention is on polymers generally, he gives some special attention to how well the different methods work for syntactic materials[Wikipedia], which are foams made by filling a metal, polymer, or ceramic matrix with hollow particles.  The author presents advantages and disadvantages of powder bed fusion, material extrusion, material jetting, and binder jetting (in which a liquid binding agent is selectively deposited to join powder materials).  He notes that, while the latter two methods “might work, but have significant limitations” due to the materials and the technology, and that extrusion methods could be viable if layers can be made to adhere before curing, powder bed methods “are likely the best approach” for syntactic materials. 

 

Many techniques have been developed and a large variety of equipment is available for additive manufacturing with polymers.  Techniques and equipment are both relatively scarce for ceramics, as the same author points out in another slide presentation (“Additive Manufacturing for Ceramics”[SciTech Connect]).  This presentation describes different techniques for executing certain ceramic-processing steps, illustrates some commercial equipment, and tabulates advantages and disadvantages of three different processing methods for ceramics (material extrusion, binder jetting, and material jetting). 

 

Each of the reports described above focuses on projects that involve a single additive manufacturing method.  In view of the fact that no single method or technology is compatible with the entire range of materials, a collaboration of researchers at the University of Illinois at Urbana-Champaign, the Massachusetts Institute of Technology (MIT), the University of Wisconsin—Madison, and Lawrence Livermore National Laboratory (LLNL) is advancing three methods that each have different strengths with a goal of integrating them into a single technology.  While commercial additive manufacturing achieves roughly 100-micrometer[Wikipedia] resolution with just a few materials, the multilab collaboration’s methods improve on this by incorporating several materials with feature sizes in the few-micrometer to submicrometer range.  This work is described in LLNL’s March 2012 Science and Technology Review[SciTech Connect] in an article entitled “Materials by Design”. 

 

The three additive manufacturing methods have complementary advantages.  Projection microstereolithography, a vat photopolymerization method, has only been made to work on a few materials, but the method reliably produces three-dimensional structures, even if those structures have empty spaces.  The material-extrusion method of direct ink writing produces structures with empty spaces even more easily and works with more materials, but its 3-D capability is inferior to that of projection microstereolithography’s.  Electrophoretic deposition also works with more materials than projection microstereolithography but has less 3-D capability, and empty spaces in components have to be made by burning out excess material. 

 

Projection microstereolithography works by projecting images of each layer of the product onto a liquid photopolymer, which hardens the portions of photopolymer within that image, thus making the corresponding layer of the product.  The quality of products made by this method has been limited by the uniformity of the light at the image source or at the photopolymer, by the photopolymer’s own characteristics, and by the system’s optical resolution.  Researchers have been mathematically modeling the process to determine which parameters to adjust to produce particular shapes at the desired resolution. 

 

Figure 11.  Octet-truss space frames[Wikipedia] made by projection microstereolithography—single unit cell (left, middle) and a 33 array (right).  The 33 array is approximately 1 millimeter wide and 1 millimeter high.  The layer-by-layer construction of the frames is evident from the middle micrograph.  (From “Materials by Design”, Science and Technology Review, March 2012 [SciTech Connect], p. 20.) 

 

Direct ink-writing is done by depositing a product one layer at a time as a continuous ink filament from a computer-controlled print head.  Current systems can write up to 10 centimeters per second and can produce features as fine as 0.2 micrometers—finer than those produced by projection microstereolithography.  To achieve the even more complex 3-D structures that some artificial materials call for, new inks derived from metals, ceramics, and polymers have been designed to readily flow through nozzles without clogging and set rapidly afterward for minimal shrinkage and shape distortion.  Other advances toward greater control of products’ material arrangement and composition include mathematical modeling of ink-deposition dynamics and improving ink-delivery systems. 

 

Electrophoretic deposition, long used to produce material coatings, involves inducing electric charge on the surfaces of liquid-suspended particles to drive them toward electrodes, where they form a layer.  Researchers have advanced this method by an operation similar to one used in projection microstereolithography:  projecting product layers’ images onto a transparent photocounductive layer attached to an electrode, so that only the image area has an electric field to drive the suspended particles.  Altering the images results in the deposition of different shapes, allowing the construction of 3-D pattern as a succession of layers.  Mathematical models of deposition process details also show the investigators how to design electrodes to result in specific deposited shapes. 

 

The collaborators have additively manufactured things made of more than one material, and have even combined different techniques.  When discussing the construction of a lattice structure that combines one material with high thermal expansion and one with low thermal expansion, with the two materials arranged so that the lattice as a whole would have zero or graded thermal expansion while remaining structurally sound, the principal investigator for projection microstereolithography told how “we can strategically place void spaces or small amounts of bending or twisting in a local structural member to accommodate growth or shrinkage from temperature changes.”  In another project, direct ink-writing of an aluminum lattice was combined with electrophoretic deposition of copper oxide in the lattice spaces to produce improved thermites[Wikipedia], which are compositions of metal fuel and oxide that release energy exothermically[Wikipedia] when ignited.  The project’s first effort resulted in a thermite with about twice the power density and burn velocity and half the burn time of more conventional thermites. 

 

Figure 12.  In a recent experiment, Livermore researchers fabricated a highly ordered thermite structure by combining two additive manufacturing techniques. They applied direct ink writing to build an aluminum lattice structure complete with void space. They then filled the void with copper oxide using electrophoretic deposition.  (P. 19, “Materials by Design”, Science and Technology Review, March 2012 [SciTech Connect].) 

 

As we noted at the beginning of this article, additive manufacturing removes a constraint on designers by increasing the range of what it’s practical to make.  The rest of the article has indicated how additive manufacturing widens designers’ options in another way, whose significance can be thought of in terms of scatterplots like Figure 13.  Graphs like these, in which each coordinate axes represents some physical property that a material might have, aid one of the common basic tasks of engineering:  the selection of materials to perform particular functions under specified conditions.  To perform those functions, the material has to have one or another particular set of physical properties.  Since each point marked on the scatterplot represents at least one material, and each point’s coordinates represent the materials’ relevant physical properties, any material that can perform the required function will, if it exists, be represented by some mark on the scatterplot.  As some of the aforementioned reports indicate, one result of additive manufacturing is the production of materials with new combinations of properties, which can be represented as filling in areas of these scatterplots that earlier technology left empty.  The scatterplots thus serve as maps, illustrating one way that additive manufacturing is expanding the frontiers of our technology into previously unoccupied territory. 

Figure 13.  Materials designed with new additive manufacturing techniques exhibit high stiffness and low density, occupying a previously vacant area of the material selection[Wikipedia] scatterplot (an Ashby plot[Wikipedia]) for Young’s modulus (stiffness) versus density. The octet truss structure shown at the upper left and  in Figure 11 is a stretch-dominated lattice.  (From “Materials by Design”, Science and Technology Review, March 2012 [SciTech Connect], p. 16.) 

References

WikipediaResearch Organizations
• Hip replacement:  Acetabular cup• Oak Ridge National Laboratory (ORNL)
• Titanium aluminide• ORNL’s Manufacturing Demonstration Facility
• Hot isostatic pressing• Arcam AB
• Megapascal• Oak Ridge Institute for Science and Education (ORISE)
• Lamella (materials)• Velocys, Inc.
• Lamellar structure• Edison Welding Institute (EWI)
• Kapton• Mid Atlantic Technology, Research, and Innovation Center (MATRIC)
• CubeSat• Los Alamos National Laboratory
• Polymer



• University of Illinois at Urbana-Champaign
• Sodium chloride:  Optical usage• Massachusetts Institute of Technology (MIT)
• Photopolymer• University of Wisconsin—Madison
• Curing (chemistry)• Lawrence Livermore National Laboratory (LLNL)
• Syntactic foam
• Micrometre
• Space frame
• Thermite
• Exothermic
• Material selectionAshby plots

 

Reports Available through OSTI’s SciTech Connect

• “The Use of Additive Manufacturing for Fabrication of Multi-Function Small Satellite Structures” [Metadata and full text available from DoE’s SciTech Connect]

• “Thermographic In-Situ Process Monitoring of the Electron Beam Melting Technology used in Additive Manufacturing” [Metadata and full text available from DoE’s SciTech Connect]

• “Microstructural Properties of Gamma Titanium Aluminide Manufactured by Electron Beam Melting” [Metadata and full text available from DoE’s SciTech Connect]

• “Prediction of Material Thermal Properties and Beam-Particle Interaction at Meso-Scale during Electron Beam Additive Manufacturing” [Metadata and full text available from DoE’s SciTech Connect]

• “Distributive Distillation Enabled by Microchannel Process Technology” [Metadata and full text available from DoE’s SciTech Connect]

• “Real-time Process Monitoring and Temperature Mapping of the 3D Polymer Printing Process” [Metadata and full text available from DoE’s SciTech Connect]

• “An Overview of Polymer Additive Manufacturing Technologies” [Metadata and full text available from DoE’s SciTech Connect]

• “Additive Manufacturing for Ceramics” [Metadata and full text available from DoE’s SciTech Connect]

• “Materials by Design” in March 2012 Science and Technology Review (Lawrence Livermore National Laboratory) [Metadata and full text available from DoE’s SciTech Connect]


Additional References

FLIR

StrataSys, Inc.

ASTM F2792

Acknowledgement

Prepared by Dr. William N. Watson, Physicist

DoE Office of Scientific and Technical Information