This study shows multi-material additive manufacturing (AM) using fused filament fabrication (FFF) of stainless steel and zirconia.
Technical ceramics are widely used for industrial and research applications, as well as for consumer goods. Today, the demand for complex geometries with diverse customization options and favorable production methods is increasing continuously. With fused filament fabrication (FFF), it is possible to produce large and complex components quickly with high material efficiency. In FFF, a continuous thermoplastic filament is melted in a heated nozzle and deposited below. The computer-controlled print head is moved in order to build up the desired shape layer by layer. Investigations regarding printing of metals or ceramics are increasing more and more in research and industry. This study focuses on additive manufacturing (AM) with a multi-material approach to combine a metal (stainless steel) with a technical ceramic (zirconia: ZrO2). Combining these materials offers a broad variety of applications due to their different electrical and mechanical properties. The paper shows the main issues in preparation of the material and feedstock, device development, and printing of these composites.
According to ISO/ASTM, additive manufacturing (AM) is the general term for technologies that create physical objects based on a geometrical representation by successive addition of material1. Hence, these technologies offer the possibility of manufacturing components with extremely complex geometry, which cannot be attained by any other shaping technique known to the authors.
Ceramic materials have been studied since the early development of the different AM technologies in the past quarter century2,3; however, additive manufacturing of ceramic components is not state of the art in contrast to additive manufacturing of polymer or metal components. Several overviews about the AM technologies used for ceramic components are given by Chartier et al.4, Travitzky et al.5 and Zocca et al.6, which can be classified according to the state of the material that is used – powder materials, liquid materials, and solid materials4,5 or according to the kind of material deposition and solidification6. AM devices are available that allow the additive manufacturing of dense and high-quality ceramic components with the desired properties for most applications7,8,9,10,11.
Production of ceramic components requires complex processing, and this has stalled progress in the AM of ceramics. Nevertheless, ceramic components are indispensable for special consumer goods and medical devices, and AM opens up new horizons for the fabrication of novel components with "impossible" geometries12. For technical ceramic components, a subsequent thermal treatment of the manufactured components is required since the AM shaping of ceramics requires the use of powders suspended in organic binders that need to be removed (i.e., debinding) before the powder is fused together (i.e., sintering).
The AM of multi-material or multi-functional components combines the benefits of AM and functionally graded materials (FGM)13 into ceramic-based 4D-components14. Material hybrids allow property combinations such as electrically conductive/insulating, magnetic/non-magnetic, ductile/hard or different colorations. Hybrid components can exhibit sensor or actuator functions known from MEMS (micro electromechanical systems)15 as well. Furthermore, metal/ceramic composites can complement joining ceramic parts into machines since conventional weldable steel partners can be used.
The European Project cerAMfacturing (EU-Project CORDIS 678503) is developing AM technologies for single material components as well as a completely new approach for AM of multi-material components, which will allow serial production of customized and multifunctional components for various applications12. Three different suspension-based AM techniques are qualified to allow the AM of ceramic-ceramic as well as metal-ceramic components. The utilization of suspension-based AM techniques promises improved component performance in comparison to powder-based methods. Because the particle distribution of the powder in a suspension is more homogeneous and more compact than in a powder bed, these shaping methods yield higher green densities, which result in sintered components with dense microstructures and low surface roughness levels12.
Along with lithography-based ceramic manufacturing (LCM)7,8,9,10,11,16,17, fused filament fabrication (FFF) and thermoplastic 3D-printing (T3DP)12,14,18 are being developed. FFF and T3DP are more suitable for the AM of multi-material components than LCM because of the selective deposition and solidification of the certain material instead of the pure selective solidification of material deposited all over the entire layer14.
An additional benefit of FFF and T3DP compared to LCM is the use of thermoplastic binder systems instead of photo-curing polymers. The binder system enables the processing of powders independent of their optical properties such as absorption, emission and reflection of electromagnetic waves, e.g., dark and bright materials (in visible range), which is necessary for the production of metal-ceramic components19,20. Furthermore, low investment is required for the FFF equipment since a large variety of standard devices are available. This technique becomes economical due to the high material efficiency and the recyclable materials. Finally, FFF is easy to upscale for large parts since the process relies on moving print head on axles.
This paper presents the first results of manufacturing metal-ceramic composites using FFF. Additionally, the technical combination of FFF and T3DP units is presented, although it is still under development. In the FFF process, filaments of thermoplastic polymers are melted and selectively extruded by the action of two counter rotating elements. Once the material is extruded through the nozzle, it solidifies by cooling, enabling the production of components layer-by-layer. To produce final ceramic and metallic components, a variant of the process has been developed21,22,23,24,25,26. The polymeric compounds, known as binders, are highly filled with a ceramic or metallic powder. Once the shaping of the components has been conducted using the conventional FFF approach, two additional steps are required. First, the polymeric components must be completely removed from the specimens in the debinding stage, generating a structure with numerous micro-sized pores. To attain the final properties, the powder compacts are subsequently sintered at a temperature below the melting point of the material. Using this approach, the production of materials such as silicon nitride, fused silica, piezoelectric ceramics, stainless steels, tungsten carbide-cobalt, alumina or titanium dioxide23,24,25 has been successfully conducted elsewhere.
The use of highly-filled polymeric filaments and the characteristic of the process impose certain requirements in the materials21. Good compatibility must be provided between the thermoplastic binder components and the powder, which must be homogeneously distributed using compounding techniques at temperatures above the melting point of the organic binder components, such as kneading or shear rolling. Since the solid filament has to act as a piston in the print head to push the molten material, a high stiffness and a low viscosity are required to enable the extrusion of the material through the nozzle with typical diameters ranging from 0.3 to 1.0 mm. Meanwhile, the material must possess enough flexibility and strength to be shaped as a filament that can be spooled. To combine all these properties while having a high load of powder, different multi-component binder systems have been developed21,22,26.
In addition to the use of adequate binder formulation, a new driving system has been employed in this work. Commonly, toothed drive wheels are used to push the filament through the nozzle. These teeth can damage the brittle filament. In order to reduce the mechanical requirements of the filaments and increase the extrusion pressure during the FFF process, the conventional FFF system of toothed drive wheels was replaced by a special dual belt system. High friction and guidance is generated due to the length, the shape, and the special rubber coating of the belts. The most important issue was preventing any buckling of the filament through the print head. The filament must be guided all the way to the nozzle, no free space is allowed, and the necessary transitions between the components have to be considered.
After leaving the feeding unit, the filament enters the nozzle unit. The main goals were designed temperature management and gapless guidance. The developed print head is shown in Figure 1.
Figure 1: CAD model of the new belt drive unit (top) and image of the real unit (bottom). Please click here to view a larger version of this figure.
Another big challenge to be addressed for the production of metal-ceramic components is the selection of powders that allow co-processing during the thermal treatment (comparable coefficient of thermal expansion (CTE), temperature regimes, and atmosphere) and particularly the adjustment of the shrinkage behavior of both materials during the sintering step. In this work, an attempt is made to combine zirconia and modified stainless steel 17-4PH since they have a comparable CTE (approx. 11 x 10-6/K) and can be sintered at the same conditions (reducing hydrogen atmosphere, sintering temperature: 1350 – 1400 °C). However, for adjusting the shrinkage behavior, a special milling procedure for the metallic powder is required19,20.
1. Materials Used
2. Filament Production
Figure 2: Filament production line. The material is extruded on a controlled manner by the regulation of the extrusion speed and temperature. Afterwards, it is collected and driven by a conveyor belt and haul-off unit. The filament diameter is measured and if the values are within the desired range, the filament is spooled. To regulate the filament dimensions, the pulling and spooling speeds must be progressively adjusted. Please click here to view a larger version of this figure.
3. Additive Manufacturing of Green Components
Figure 3: Manufacturing process for metal-ceramic component with tower structure. Please click here to view a larger version of this figure.
Figure 4: Virtual print of a component with surrounding ooze-shield. Please click here to view a larger version of this figure.
4. Debinding and Sintering of Components
The best fitting results for stainless steel sintering behavior were obtained with an attrition milling time of 180 minutes and a planetary ball mill (PBM) milling time of 240 minutes. Figure 5 shows a SEM-image of the untreated powder (left), the deformed particles after the attrition milling (middle), and the chopped particles after the PBM milling step (right).
Figure 5: Untreated stainless steel <38 µm (D90) (left), stainless steel powder after attrition milling (middle), and stainless steel powder after PBM milling (right) Please click here to view a larger version of this figure.
The sintering behavior of the initial and milled steel powder are compared with the sintering behavior of the zirconia powder in Figure 6, all measured with an optical dilatometer.
Figure 6: Dilatometric curves of the zirconia powder (TZ-3Y-SE) and the stainless steel powder (17-4PH) in the initial state and after a high-energy milling treatment of the stainless steel powder. Please click here to view a larger version of this figure.
The improvement of the feedstock mechanical properties in the high shear compounding step was characterized for the zirconia feedstock. A feedstock produced in a single compounding step of 75 min in a roller rotors mixer (RM) was compared with the one produced by the method described in the protocol. Filaments were extruded using a high pressure capillary rheometer with a die of 1.75 mm diameter, a piston speed of 1 mm/s and a temperature of 190 °C. The filaments were collected with a conveyor belt and tested with a universal tensile testing machine. At least 5 repetitions were conducted per material. Figure 7 shows a comparison of both materials concerning the ultimate tensile strength (UTS), the elongation at UTS, and the secant modulus.
Figure 7: Influence of the compounding method in the mechanical properties of the zirconia feedstock. Feedstock was compounded in an internal roller mixer (RM) or in combination with a co-rotating twin screw step (TSE). The strength, flexibility and stiffness of filaments produced with a capillary rheometer were determined using the mean value and the correspondent standard deviation of the ultimate tensile strength (UTS), the elongation at UTS and the secant modulus, respectively. Please click here to view a larger version of this figure.
In Figure 8, the diameter values obtained during the production of the filaments made of zirconia (left) and stainless steel (right) feedstocks are presented. The diameter of the extruded filament was recorded during the production process via single-screw extrusion. For the zirconia filaments, a good control of the dimensions could be achieved with a mean diameter of 1.75 mm and a standard deviation of 0.02 mm. For the filaments containing the modified stainless steel powder, a higher variability of the average filament diameter was observed. A possible reason for this could be an inhomogeneous particle distribution within the feedstock resulting from the platelet-like shape of the metallic particles (Figure 5). In this case, a higher number of measurement points were found outside the desired range of 1.75 mm ± 0.05 mm, and the mean diameter value was 1.74 mm with a standard variation of 0.03 mm. For both types of filaments, the ovality values were considerably smaller than the 0.1 mm limit.
Figure 8: Histograms of the filament diameter for the studied materials. Please click here to view a larger version of this figure.
Figure 9 shows the suitable metal and zirconia filaments to manufacture green sandwich structures with the composition steel-zirconia-steel (left) as well as zirconia-steel-zirconia (right).
Figure 9: Green steel-zirconia-steel (left) and zirconia-steel-zirconia components (right) additively manufactured by FFF. Please click here to view a larger version of this figure.
Due to the similar binder system of both materials, it is possible to fuse certain layers to a monolithic composite part. A larger round shaped part with sharp transitions is shown in Figure 10.
Figure 10: Structure with sharp transitions between Zirconia and stainless steel. Please click here to view a larger version of this figure.
Figure 11 shows other green single- and multi-material components that were further processed. Figure 12 shows a pure zirconia sample on the left side, The middle shows a pure stainless steel sample, and finally a sintered and well joined steel-ceramic composite is pictured on the right side.
Figure 11: Green test samples manufactured by FFF; top: zirconia-steel-composites with stainless steel on top; middle: stainless steel; bottom: zirconia. Grid box 5 mm. Please click here to view a larger version of this figure.
Figure 12: Sintered zirconia sample (left), sintered stainless steel sample (middle), and sintered zirconia-stainless steel-composite (right). All scales in mm. Please click here to view a larger version of this figure.
In Figure 13, a typical structure of FFF-components with crotches (or sub-perimeter) between two deposited filaments is shown, which resulted from an ordinary slicing (tool path) and the continuous way of material deposition.
Figure 13: Typical structure of FFF-components resulting from slicing and continuous material deposition. Please click here to view a larger version of this figure.
By raising the extrusion multiplier in the slicing software, which leads to a higher volume deposition, the sub-perimeter can be reduced as well as by adapting the tool paths. Nevertheless, due to the high content of particles in the filaments, it is evident that the deposition behavior differs from ordinary printing of thermoplastics. Therefore, a software modification to close such defects is desirable.
After solvent debinding, thermal debinding and subsequent sintering, all the different samples showed no significant deformation or bloating. The sintered pure zirconia and stainless steel FFF specimens have a good geometrical stability both with and without compressive load and they do not buckle. The total mass loss was 14.8-14.9%, indicating complete debinding.
The metal-ceramic samples showed a good macroscopic adhesion of both materials. The mass loss after the sintering of the composites was found to be 14.1-14.4%, which also indicates a full debinding. Further analysis and process adjustments will follow. The electron microscope characterization of the composites is intended to provide insight into the quality of the composite. The desired formation of the composite has taken place successfully as shown in Figure 14.
Figure 14: SEM Image of microstructure in the metal-ceramic interface showing the material joint. Please click here to view a larger version of this figure.
The results show that a promising approach to manufacture metal-ceramic composites using FFF generating electrical conductive and electrical isolating properties into one component. Furthermore, the implementation of ceramic parts into metallic environments becomes possible due to the good material bond and weldability of the stainless steel. Within the EU, project heating devices were manufactured by FFF containing an electrical conductive path made of stainless steel in a non-conductive ZrO2 matrix. Figure 15 shows the sintered samples. These multi-material components must be analyzed and tested in the future.
Figure 15: Sintered heating elements made of zirconia and stainless steel Please click here to view a larger version of this figure.
Figure 16 and Figure 17 show the new print head with two FFF-printing heads and two T3DP-printing heads as CAD-model (Figure 16) as well as implemented in the FFF device (Figure 17). One challenge is controlling of the output for both systems. For the micro dispensing units, the output is controlled by the frequency of a piezo-driven piston instead of the stepper motors speed for the belt drives within the FFF-printing heads. The interaction of both devices must be tested in the future.
Figure 16: CAD model of new print head with two FFF-printing heads and two T3DP-printing heads. Please click here to view a larger version of this figure.
Figure 17: Image of new print head with two FFF-printing heads and one T3DP-printing head (left). Please click here to view a larger version of this figure.
The zirconia and stainless steel used here are very suitable for the co-sintering of metal-ceramic components because of the comparable CTE, sintering temperature, and sintering atmosphere. The sintering behavior of the zirconia and the stainless steel feedstocks could be adjusted by the treatment of the stainless steel powder (Figure 9) successfully. By using the mentioned materials and methods, it is possible to manufacture macroscopic defect-free parts by FFF for the first time. To the authors' knowledge, no other comparable AM method is known to manufacture such parts except for T3DP19,20. One application for metal ceramic components is shown in Figure 17, which is a heating element with an electrical conducting looped stainless steel in an isolating zirconia matrix.
One of the major challenges for the FFF of metallic and ceramic components is the dramatic increase of the stiffness and brittleness of the filaments due to the high solid content. Therefore, the selection of the right binder components was a key factor for the success of the project. In addition, the strength and flexibility of the filaments could be improved by the use of a high shear mixing technique (Figure 7). According to previous studies with highly filled systems28, this improvement could be caused by a better powder dispersion and reduction of the agglomerates29,30.
The investigation and adjustment of the extrusion, pulling and spooling speeds during the filament production process allowed the production of highly particle-filled filaments with the proper dimensions. Other parameters like the temperature distribution within the extruder as well as the use of cooling devices significantly influenced the filament quality and were chosen carefully.
Both filaments were processed in the FFF-device successfully. The adhesion between the feedstocks was found to be very good in the green state (Figure 7-9). Only some small unfilled volumes were visible, which are typically for a state of the art FFF process (Figure 13). To close these critical volumes with thermoplastic materials, the FFF-device was equipped with two micro dispensing units known from T3DP18,19,20,31,32, which allow the deposition of single droplets to close the insufficient filled volumes as well as the manufacturing of finer structures (Figure 14 and 15).
Geometrical restrictions of the part complexity or resolution are strongly dependent on the printer setup the continuous material flow as well as the used slicing software. The design rules and the resulting part appearance are at most found to be similar to using FFF of plastics.
The authors have nothing to disclose.
This project has received funding from the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No 678503.
Zirconia | TZ-3YS-E | Tosoh, Europe B.V. | |
Stainless steel | UNS17400 -38 µm | Sandvik Osprey Ltd. | |
Table of Devices and Software | |||
slicing software | Simplify 3D | Simplify 3D, USA | |
roller rotors mixer | Plasti-Corder PL2000 | Brabender GmbH & Co. KG, Germany | |
3D printer | model Ceram | HAGE, Austria | |
cutting mill | SM200 | Retsch Gmbh Germany | |
corotating extruder | ZSE 18 HP-48D | Leistrutz Extrusionstechnik GmbH, Germany | |
laser measurementdevice | Diagnostic Laser 2010 | SIKORA AG, Germany | |
capillary rheometer | Rheograph 2002 | Göttfert Werkstoff-Prüfmaschinen GmbH, Germany | |
single screw extruder | FT-E20T-MP-IS | Dr. Collin GmbH, Germany | |
tungsten furnace | Hochtemperatur-Wolframofen WOHV 250/300-1900V | MUT Advanced Heating GmbH | |
debinding furnace | Retorten-Entbinderungsofen RRO 280 / 300-900V | MUT Advanced Heating GmbH | |
attrition mill | PE 1.4 | Erich NETZSCH GmbH & Co. Holding KG, Germany | |
PBM (planetary ball mill) | PM 400 | Retsch Gmbh, Germany |