Embedding ceramic components in metal structures with hybrid directed energy deposition

The combined benefit of both additive and subtractive manufacturing within the same gantry system enables hybrid directed energy deposition to create complex geometries with smooth surface finish and superior dimensional accuracy. Moreover, with layer-by-layer access to the structure during both the addition and subtraction of material, the insertion of components is now possible, assuming the components can survive the high temperatures associated with the subsequent metal deposition. Ceramic inserts are of interest for a variety of reasons including (1) to create complex interwoven ductile/brittle composites for ballistics or high-temperature applications or (2) to integrate high-temperature strain or temperature sensors protected within ceramic substrate subsumed into a larger metal structure. In this work, stainless steel substrates were machined to create an internal cavity for the insertion of a ceramic component. During the investigation of several different over-the-ceramic deposition strategies, components were inserted, and different process sequences were allowed to continue to envelop the inserted ceramic with varying success. Unmelted powder was used to serve both as a thermal buffer and to provide a flush surface upon which the laser cladding could continue. Subsequent depositions were attempted with both dry and wet powder (addition of machining coolant to wet). The wet powder has previously been demonstrated to not significantly impact the mechanical properties of a final structure and provided a thermal barrier to protect the ceramic piece from the extreme temperatures of the final metal deposition. The wetting of the powder provided stability and minimized displacement caused by the powder flow from the laser cladding head. Finally, the use of an oblique angle for laser cladding allowed for the redirection of some fraction of the introduced thermal energy away from the ceramic component and, consequently, improved the survival of the ceramic inserts. With this combination of techniques, ceramic inserts survived full embedding within a 3D-printed stainless steel structure.


Introduction
The future of advanced manufacturing will include precise digital control of material, geometry, and functionality in three dimensions to provide the next generation of customized, high-value, and functionally graded products [1]. With traditional subtractive processing providing the foundational process, hybrid additive manufacturing (as defined in this work) includes the integration of additive manufacturing into CNC machining systems. Leveraging common directed energy deposition (DED) technology, hybrid systems deposit material with wire or powder feedstock to create complex geometries [2][3][4][5]; subsequent machining renders superior dimensional accuracy and surface finish. Collaboratively, the integrated processes can fabricate structures through the incorporation of multiple manufacturing techniques to create products within a single, digitally driven build chamber. The interleaving of the processes can occur at each layer or can be multiple layers depending on the access and reachability of surfaces after an extended deposition cycle of N layers. Furthermore, repairs can be directly applied to existing structures through the subtraction of suspect material, 3D scan capture, deposition of new structure, 3D scan capture, and a final subtraction of material as necessary [4,[6][7][8][9] to finish with an intended geometry. Figure 1 illustrates the Okuma MU8000 Laser EX system equipped with both machining and additive deposition capability in the same five-axis gantry.
Starting in the 1990s, DED has been investigated generally [2,[10][11][12][13] including with both metal and ceramic feedstock [14][15][16][17][18][19][20][21] in both powder or wire form, either of which is fed into a melt pool created by directed energy sources such as laser, electric arc, or electron beam [3]. Material is deposited in a selective manner in order to build up a three-dimensional object [22]. With powder-based systems, metal powder is blown by an inert gas to supply the melt pool with additional feedstock while providing protection from the otherwise oxidizing atmosphere [17]. The tool head includes the directed energy (laser beam in this example) coincident with the surface and gas-delivered powder. A unique advantage of blown-powder DED is the capability not only to change materials but also to vary the composition for functionally grading the structure with on-demand alloying with 3D spatial control [20,[23][24][25][26].
Hybrid manufacturing has been defined as the integration of multiple processes with layerwise access to a structure during fabrication to create a single material structure with improved dimensional accuracy or more complex multi-material and multicomponent structures [27][28][29]. Hybrid additive manufacturing in this paper is defined as the layerwise combination of AM with other traditional processes within a single build envelope [28], and in this case, it included the following: machining, directed energy deposition, and the insertion of components. Hybrid DED with machining has been researched for decades [16,27,[30][31][32]. By further enhancing the system with the integration of robotic placement within the build chamber, subcomponents can be integrated into the fabricated structure by machining geometry-specific cavities, inserting components, and, finally, capturing these sub-components within the structure with subsequent metal deposition. Motivations for embedding components within structures include the modulation of density, the improvement of impact energy management [33], or even the integration of embedded sensing and programmability for structural health diagnostics [34]. The inserted components can be fully encapsulated within the metal structure, assuming that the inserts can survive the subsequent high-temperature metal deposition process. However, this requirement dramatically limits the range of components that can be integrated without introducing a thermal barrier or altering the process to minimize the exposure of the subcomponent to the high temperatures of the laser cladding process.
Process interruptions in additive manufacturing for the purpose of embedding components within polymer structures were first reported in the 90 s [1,[35][36][37][38][39][40][41][42][43][44]. On the contrary, component integrations in metal additive processes have been less common due to the high temperatures required for melting different alloys-often destroying all but the simplest high-temperature sensors. Of the successful attempts to embed components in metal objects, electron beam melting was used for a ceramic sensor [45], sheet lamination was used and provides the lowest processing temperature of any metal additive manufacturing with ultrasonic welding of metal tape strips together [37,46], a combination of selective laser melting and directed energy deposition was used in a clever multi-process approach in [47], and the earliest case was in 2000 with a form of directed energy deposition referred to as shape deposition manufacturing [48]-prior to the standardization of terminology with ISO/ ASTM 52,900 [49]. Sensors were welded onto directed energy deposition structures in this case.
The current work advances the understanding of the digital manufacturing paradigm by using a commercially available hybrid directed energy deposition to create a multimaterial, multi-component demonstration to serve as a proof of concept. The process included (1) machining a geometryspecific cavity in a 316L stainless steel substrate, (2) manually placing a ceramic insert into the cavity, (3) introducing a thermal barrier and providing a flush surface upon which the introduction of unmelted powder in the occupied cavity, (4) wetting the powder with available machining coolant to provide temporary powder stability and adhesion, and then subsequently (5) depositing metal to fully encapsulate the ceramic component and trapped powder within a larger steel structure. Future work will include introducing ceramic components with embedded thermocouples with windowed access to sensor terminals on the outer surface of the 3D-printed metal structure. The introduction of ceramics into metal substrates could provide significant utility for a diversity of applications including impact energy management, thermal control, embedded sensors, and electronics to name a few.

Materials and methods
A series of fabrication experiments were conducted of a single intended composite geometry with a 316L stainless steel substrate entrapping a 3D-printed ceramic piece. An Admaflex printer can create complex ceramic geometries with a spatial resolution of 35 microns to fabricate dielectric substrates for electronic components and the possibility of subsequently introducing additional conductive traces. Figure 2 depicts the CAD rendition (left) of the inserted ceramic structure used in this study (right). The trench was included as a feature that could enclose a 3D-printed thermocouple into the ceramic that could be used to monitor the temperature inside of the final stainless steel and ceramic composite fabrication. The two metal inks are high-temperature resistant and are hypothesized to be capable of surviving subsequent high temperatures introduced during the entrapping metal depositions. Figure 3 illustrates an example of 3D-printed ceramic electronics with the Admaflex printer and motivates this work with the possibility of entrapping active electronics within the steel structure-the subject of future work and requiring additional effort in the thermal protection of the delicate and temperature-sensitive electronic components.

Experimental design and survival of a ceramic insert
A progression of four result-guided experiments were conducted to successfully entrap the 3D-printed ceramic substrate inside the stainless steel structure. With each build, a commercial 316L stainless steel metal powder produced from Oerlikon with a particle size distribution of 106 + / − 45 um. Processing conditions for each build are shown in Table 1.

Experiment 1-simple insertion
An initial attempt was conducted where the ceramic substrate was inserted into a machined pocket where the top surface was flush with the substrate. With the top surfaces of the two substrates coincident, the DED process occurred directly on top of the ceramic without any thermal barrier.

Experiment 2-inclusion of dry and wet unfused metal powder
To evaluate the use of a thermal barrier, a second attempt was performed with unmelted metal powder between the ceramic component and the top surface of the stainless steel substrate. As portrayed in Fig. 4, the machined pocket was intentionally machined 2.5 mm deeper (the thickness of the ceramic component) than experiment 1. This void was subsequently filled with unmelted powder to provide a thermal barrier as well as a flush surface for subsequent deposition  . 3 The example of a 3D-printed alumina circuit board as potential inserts for increased functionality in metallic structures fabricated with directed energy deposition. With sufficient thickness serving as a protective thermal barrier, active electronics could potentially survive to occur. The hypothesis was that the unfused powder would improve thermal and mechanical stability by providing a thermal barrier between the ceramic and directed energy as well as provide mechanical support during and after the solidification process.

Experiments 3 and 4-insertion with wet and dry powder thermal barrier and oblique laser deposition
The final experimental build attempt included all of the details of experiments 1 and 2; however, to further minimize the disruption of the thermal barrier powder due to gas flow during directed energy deposition, deposition occurred with the print head at an oblique angle of 35° to redirect the input thermal energy to the powder and reduce the thermal input orthogonally to the underlying ceramic substrate to promote survival of the insert (Fig. 5). Leveraging previous lessons learned, a single bead was deposited on the longitudinal sides of the pocket to help mitigate delamination concerns. These deposited beads help create bonding between the substrate and the deposited geometry. Previous studies conducted at Oak Ridge National Laboratory using a similar deposition strategy concluded that while this deposition strategy can be used for encapsulation purposes, delamination due to the excess powder is a concern. However, if partial delamination occurs, it can be mitigated in subsequent depositions once the metallic thermal barrier is in place.

CT scan
X-ray computed tomography (CT) was conducted on the two successful builds using a ZEISS METROTOM 800 X-ray CT system operating at a tube voltage of 225 kV, tube power of 500 W, and a resolution of 6 µm. CT images were analyzed using Dragonfly Pro 2022.1 (Object Research Systems, Montreal, Canada). Image de-noising was performed using a combination of non-local means filters and median filters, followed by contrasting tools in 2D and 3D.

Results and discussion
For experiment 1, no unfused powder serving as support and thermal barrier was included. As seen in Fig. 6, the ceramic component was obliterated due to the thermal input from the additive process. As the ceramic degraded, pieces of the  Fig. 4 Schematics of a ceramic piece inserted into a milled cavity on the build plate as described in experiment 1 (left). For experiment 2, a slightly deeper cavity was milled for the deposition of unfused pow-der above the inserted piece to create a thermal barrier and provide additional support when the following layer is printed above (right) component were carried away by the DED processing gas flow. As seen in the figure, a cavity was left in place of where the ceramic was inserted. As shown in Fig. 7 (left), it was concluded that the loose metal powder in the cavity was blown away due to the additive manufacturing processing gasses. The nozzle and carrier gas created a crater in the surface of the powder, rendering the thermal barrier useless. To mitigate the loose powder evacuating the cavity due to gas flow during subsequent metal deposition, another attempt was conducted where the powder was wetted with machining cooling fluid. Machining coolant is readily available in hybrid manufacturing systems, and the use of machining coolant has been shown to not impact mechanical performance [50]. As shown in the right image for Fig. 7 (right), the delivery of powder orthogonal to the cavity filled with wet powder resulted in little to no displaced powder. The wetting was hypothesized to aid in powder adhesion necessary to minimize the powder disruption due to the airflow from the process gasses in subsequent powder-blown laser cladding. To test this hypothesis, experiment 1 was re-ran while using a thermal barrier of wet powder. As shown in Fig. 8, this experiment resulted in a more uniform deposition. However, the deposition is still caved in where the cavity is located. It is hypothesized that the heat from the deposition process dried the wet powder, where it was subsequently removed from the cavity by the process gas pressure.
To further reduce the thermal load into the thermal barrier, the processing angle was modified to be 35° from the surface normal. This experiment was conducted for both dry and wet powder conditions. For the dry powder experiment (Fig. 9, left), the ceramic was placed in the cavity, and the excess   . 7 The delivery of powder to a cavity filled with dry powder (left). The delivery of powder to a cavity filled with wet powder (right) powder from the process was used to fill the void in situ to the deposition process. For the wet powder experiment (Fig. 9, right), the powder/coolant mixture was placed in the cavity before deposition occurred. When compared, the wet case provides a better thermal barrier. The authors conclude that due to the nature of having the excess powder from the process to provide the feedstock for the thermal barrier in the dry case, there is insufficient time to properly fill the channel at the end of the deposition process. As hypothesized after experiment 2, the use of an oblique processing angle was sufficient to reduce the heat input due to the presence of the thermal barrier. Thus, the thermal barrier for the wet case is intact and provides a successful proof-of-concept of embedding ceramic structures in metal structures. Figure 10 shows a longitudinal cross-section of the CT scan for both experiments 3 and 4. As seen in the figure, the ceramic insert and thermal barrier are still intact. As the build progressed for both cases, delamination was observed. This is due to a combination of the buildup of excess powder from the process and the oblique processing angle. However, this is not a concern for the authors since the deposited structure provides a thermal barrier for subsequent additive operations where the delamination will be mitigated. By depositing a larger surface area around the ceramic insert, as expected in most applications, the proper metallurgical bonding will be generated. Future work will address this delamination and explore mitigation strategies.
Experiments 3 and 4 were cross-sectioned to better understand the effect of deposition on the ceramic insert. The edge of each sample was first cut to remove the ceramic insert without damage. Figure 11 shows the removed ceramic inserts. As seen in the figure, there is no damage or visible discoloration to either ceramic insert.
The samples were further cross-sectioned for optical imaging. Figure 12 shows the cross-section view of each sample. It can be seen in the figure that each sample is still intact after cutting.
Each sample was prepared using standard metallographic preparation techniques, and the samples were electrolytically etched with 10% oxalic acid solution. The last polishing step  CT videos are included as supplementary material was vibratory polishing using Buehler VibroMet 2 Vibratory Polisher using diluted colloidal silica solution for 12 h. Figure 13 shows the optical images from each sample. It can be seen in the figure that a hollow cavity was manufactured for each experiment. Experiment 4 resulted in a more uniform layer compared to experiment 3. It is hypothesized that this is due to the pre-placed wet powder that acted as a support structure during the deposition process.
Hardness and microstructure analysis was not conducted on either experiment as the results would not be pertinent to an industrial insertion of this processing scheme. After the first layer is deposited at an oblique angle with a small processing laser spot diameter (as demonstrated in this article), the ensuing deposition scheme would utilize orthogonal deposition at a large laser spot diameter that enables faster deposition rates. The analysis of material properties using this type of processing scheme is described by [51][52][53][54].

Conclusions
The work has advanced the understanding of digital hybrid manufacturing by leveraging an existing commercial system with hybrid-directed energy deposition to fabricate a multicomponent proof-of-concept. The major findings include the following: (1) A demonstration of ceramic insert fully entrapped in a stainless steel directed energy deposition process but with required mitigating operations to improve survivability: (a) The introduction of wet powder to provide a flush surface and mechanical support while simultaneously providing a thermal barrier to protect the ceramic substrate during the subsequent metal laser cladding, and The immediately subsequent layer of metal should be deposited with an oblique angle for the print head to divert introduced thermal energy into the powder and less directly to the ceramic insert.
(2) By adding geometrically complex and multi-functional 3D-printed ceramic inserts into 3D-printed metal structures with conforming cavities and then entrapping the ceramic, a diversity of applications can be enabled. The ceramic subcomponent allows for chemical-and hightemperature-resistant internal structures contained fully or partially within a larger metal structure in order to increase functionality. Thermocouples, strain gauges, fluidic valves and channels, and even embedded electronics could be embedded and survive to enhance a structural geometrically complex metal component.