Additive manufacturing (AM) of ceramics doesn’t seem to garner the same attention as AM for plastics and metals. Our industry is smaller, and our products are more often behind the scenes in applications. Ceramic processing is also more complicated in many ways than that for plastics and metals. The properties that make ceramics advantageous in the extreme conditions where other materials fail contribute to the challenges of their processing. In plastics and metals AM, many have come to expect that a finished part comes out of the printer. While true in some cases, those materials often require post processing, whether through the removal of support material, heat treating, or even hot isostatic pressing.

For ceramics, post processing is the only option. AM methods provide the opportunity to form green body ceramics without the constraints of molds and machining. After forming, ceramic AM parts must be fired, with binder removal and sintering, just as with other ceramic processes. The great challenge for AM ceramics is to produce a green body that will fire to the same high density as ceramics formed by more conventional methods, along with the realization of isotropic properties despite the layer-wise buildup of AM objects.

Additive Manufacturing Methods

Four main AM approaches exist for ceramics. Binder jetting prints binder into a spreadable powder bed to form a porous green body. Binder-jetted parts can be sintered to yield porous ceramics or infiltrated to produce dense ceramics such as reaction-bonded silicon carbide.

Laminated object manufacturing (LOM) uses lamination methods known in the electronic ceramics industry to stack precision-cut tapes into green parts. The tapes are made to fire to high density.

Robocasting uses the extrusion of rheologically controlled ceramic pastes to build components that fire to full density, bearing a characteristic coiled-rope appearance. Robocast components have found use as crucibles, gas burners, molten metal filters and catalyst supports.

Finally, lithography-based ceramic manufacturing (LCM) uses photolithography to produce complex, precision ceramics with smooth as-sintered surfaces and densities better than 99.6% for alumina, zirconia, and silicon nitride. LCM parts and post processing bear a strong resemblance to injection molding. The lithography process forms a green part from a slurry that contains photocurable binder, plasticizers and dispersants. The green ceramic is a composite of the ceramic powder and the crosslinked binder. The binder is removed in a careful burnout stage, and the part is subsequently fired to achieve the desired density.

Understanding LCM

The LCM process builds components layer by layer out of a slurry (see Figure 1). A dynamic mirror array projects blue light onto the underside of the transparent vat, curing the photopolymer only in the exposed area. The mirror array results in an x-y resolution of 40-60 µm. After each layer is cured, the part is carefully lifted out of the slurry, and new slurry is added and spread across the vat as needed. The part is lowered into the slurry again to expose the next layer.

Each exposure takes just a few seconds, and each layer takes about 30 sec to complete. The layer thickness can be adjusted to 25-100 µm, depending on the requirements and results in a build rate of 3-12 mm/h in z-direction. Due to the exposure via a mirror array, there is no influence of the cured area on the build rate. The entire layer is cured at once.

The LCM process is ideal for ceramics because the high solids loading in the slurries, from 40-60 vol%, yields high green densities. The slurries are optimized to meet the viscosity and stability requirements for the printing process. Expertise in developing the slurry formulations, in addition to the equipment and control software, is key to the success of the technology. The high quality of sintered LCM components is the result of the combination of these key parameters. By yielding high-quality green parts, LCM is able to produce high-quality sintered parts.

LCM Materials

Since 2011, success has been demonstrated with several standard materials, including alumina, yttria-stabilized zirconia, tricalcium phosphate (TCP), a silica-based casting core material, and silicon nitride. However, LCM is not limited to a few standard materials. As new applications have been presented, new formulations have been developed and new materials demonstrated.

Each material has its own characteristics that determine the exact formulation of the binder system. Slurries are tailored for light penetration, scattering, viscosity, rheology and slurry stability. The list of materials has grown to include magnesia, cordierite, porcelain, mullite, silica, Bioglass™, hydroxyapatite (HA), and various cermets. This experience has created a toolbox from which technologists can quickly determine and test new materials.

LCM Applications

The LCM method is ideal for ceramic applications where high density and precise, fine features are needed. Forms that are impossible or overly complex by conventional methods (e.g., undercuts and three-dimensional lattices) are simple to produce. Designers can consider topology optimization to create strong, lightweight ceramic structures.

With conventional forming methods, testing design iterations can occupy a long and costly cycle of tool design and production. In addition, the designs are limited to what can be achieved with tooling and machining. With AM, design iterations are as simple as editing the original CAD file.

Ceramic Casting Cores

The aerospace and power industries are continually seeking to produce designs with more complex internal cooling channels to improve turbine engine efficiency. Additive manufacturing methods offer an opportunity to enable efficient designs that are not currently producible. One example can be found with ceramic cores for investment casting, which have been actively developed over the past three years.

Using a silica-based material, several turbine blade core demonstration pieces were created, as shown in Figure 2. Cores as tall as 300 mm have been produced. LCM core surfaces have excellent smoothness, ensuring that internal channels in the cast alloy have smooth walls. Complex shapes with features at least as fine as 250 µm have been proven.

A new casting core material* was developed to be optimized for LCM. The steps of the investment casting process include forming a wax turbine body model, which is used to create a silicone mold of the turbine. The original wax model is removed and the ceramic core is fixed in place. New wax is poured into the silicone mold to create the final wax pattern with the installed core. A ceramic investment mold is then formed around the wax pattern, which is subsequently molten out. Finally, an alloy is cast into the mold. After cooling, the mold is broken away and the core is leached out in a hot caustic bath. LCM-printed cores typically show a surface roughness of 1-3 µm Ra within the turbine passages.

*LithaCore 450

Biomedical Applications

State-of-the-art permanent bone implants are often made of titanium or cobalt-chrome alloys, which bear the disadvantage of requiring replacement over time. The use of bioresorbable TCP or HA allows the replacement of the implant by native bone tissue, and therefore the growth of healthy and stable bone during the healing process. TCP and HA are currently used for many medical applications, such as bone cement, and are known to be biocompatible.

Additive manufacturing is especially interesting for biomedical applications due to virtually no limitation concerning the complexity of the design and the possibility to create patient-specific products. AM is well-suited to producing ceramic scaffolds that mimic the complex spongey structure of bone, which allows it to be both lightweight and strong. Scaffolds allow the ingrowth of bone cells and provide pathways for vascularization, which accelerates the healing process by enabling the transport of nutrients and removal of metabolic waste. LCM can produce both precisely defined and highly reproducible macro-porosity and textured surfaces on the microporous structures, which together can improve osteoblast adhesion and coverage.

AM implants have an added benefit of ensuring a correct fit prior to surgery. In some cases, when an implant does not fit, the part must be machined during the surgery and tried again until a good fit is obtained. This lengthy process can be avoided by creating a 3D digital model from a CT scan of the original tissue or defect, and printing the desired part. A perfect fit can be achieved without the need for bedside adjustments, enabling significantly faster and less traumatic surgery.

Cranium (skull) injuries sometimes require trepanation (opening of the cranium) and subsequent closure of the opening with an implant. An implant tailored to the patient in terms of size and thickness is beneficial. In addition, it is necessary to design the implant similar to the native tissue in order to achieve a proper ingrowth of the surrounding cells. Several active research projects are underway to test cranial bone substitution designs. Figure 3 shows the typical structure of a human cranial bone and its corresponding LCM-produced TCP implant. Cranial bone consists of two dense cortical bone layers (5-10% porosity), with a very porous internal layer of trabecular bone (50-90% porosity). The displayed example has an overall porosity of 50% to mimic the natural bone tissue. The first human trials of LCM implants are planned for 2017.

Advanced Fluid Flow Structures

The ability to form complex, controlled 3D networks relates not just to bone scaffolds, but also to gas and liquid flow applications such as catalyst materials, static mixers, and heat exchangers. Catalysts typically comprise packed beds with macro- to nanoscale porosity, sponge-derived reticulated ceramics, and extruded honeycombs. Random structures may have wide flow and reaction variations. Laminar flow in honeycombs may prevent complete contact of fluids or gases with the catalyst.

LCM permits 3D design freedom to enable the implementation of new design features that can introduce controlled turbulence to improve catalyst performance and lifetime. Novel functional geometries are being explored that are derived from fluid and thermodynamic simulations to improve mixing efficiency. Such a structure, developed at Fraunhofer IKTS, incorporates channels of oscillating width and cross channels, allowing mixing between flow streams (see Figure 4).

The LCM process has produced monolithic 3D catalyst structures based on alumina, cordierite, mullite, magnesia and silicon nitride, with detailed feature sizes down to 150 µm. Such catalysts provide a high surface-to-volume ratio and significantly reduce the pressure drop suffered by standard pellet-type catalysts. 3D-printed catalysts show promise for satellite launch and propulsion, where a significant need exists for capable catalytic solutions.

Design Freedom

New applications continue to be investigated, specifically in terms of enabling new design thinking and added functionality for electronic, magnetic, and catalytically active ceramics. With ceramic AM now commercially available, ceramists, materials scientists and application engineers have increasing freedom in ceramic design. Harnessing additive capabilities in the ceramic industry will enable ceramics designed for functionality rather than manufacturability.
 

For more information, contact the lead author at sallan@lithoz-america.com or visit www.lithoz.com.