Development of a build volume reduction kit for studying epitaxial re-solidification in laser powder bed fusion

Laser powder bed fusion (PBF-LB/M) is a promising additive manufacturing process that enables the production of complex and high-performance parts. However, the high cost of materials and the need for large quantities of powder in conventional industrial-grade systems pose challenges for experimental materials development and testing activities. This study focuses on the development of a modular build volume reduction kit for an existing EOS EOSINT M-series PBF-LB/M machine. The proposed build volume reduction kit can be customized and adapted for specific research needs, expanding the capabilities of existing infrastructure without significant capital investment. This study describes the design and characterization of the build volume reduction kit and a detachable Pt-heater module, which allows for preheating of the substrate material above 500 °C. The kit’s operation was validated by manufacturing simple cuboid samples using EOS 316L stainless steel powder on a 316L stainless steel substrate. The results demonstrate the feasibility of using the reduction kit for cost-effective experimental investigations, as well as highlighting its potential for studying the epitaxial solidification of PBF-LB/M-built functional materials.


Introduction
Laser powder bed fusion (PBF-LB/M) is a type of additive manufacturing (AM) process that uses a focused laser beam to selectively melt and fuse layers of powders (polymer or metal) to create complex three-dimensional (3D) structures with high geometrical precision [1].PBF-LB/M enables the production of complex parts with intricate geometries, to an extent that is not possible with conventional manufacturing processes.Additionally, in some cases, PBF-LB/M can produce parts with superior mechanical properties in comparison to those that have been conventionally manufactured, which results from having the capacity to control parts microstructure during the manufacturing process [2].Such capabilities have made PBF-LB/M an attractive option for industries requiring complex and highperformance parts, such as for sports cars, aerospace components, and biomedical implants [2,3].1296 (2023) 012018 IOP Publishing doi:10.1088/1757-899X/1296/1/012018 2 A large quantity of the research published on PBF-LB/M is focused on materials development, e.g.characterization of the properties of PBF-LB/M-built materials or the development of new materials suitable for PBF-LB/M [4,5].Indeed, PBF-LB/M has proven particularly useful in manufacturing metallic alloys that are otherwise difficult to achieve using traditional manufacturing methods, e.g.tungsten alloys [5] or Ni-Mn-Ga-based magnetic shape memory alloys [6,7] that, due to their high brittleness, are generally difficult to machine.Furthermore, PBF-LB/M also enables the processing of new materials with unique chemical compositions and properties [8].Although modern materials development significantly benefits from the computational materials engineering, developing new alloys and processes still heavily relies on experimental materials development and testing activities.The majority of the commercially available PBF-LB/M machines have build volumes above 150×150×150 mm 3 [9].Different research institutes and universities have typically purchased these types of larger industrial-scale machines due to the capabilities they offer in the manufacture of larger parts and patch sizes [10].However, although this is advantageous for general mechanical engineering and applicationoriented studies, a significant challenge remains when it comes to experimental materials development and testing activities.Most of these industrial-scale machines require the minimum usage of dozens of kilograms of metal powder to fill the build chamber and successfully operate the device.This scenario is problematic because the relatively high cost [11] of new experimental materials and precious alloys can make it difficult for researchers to conduct experimental research with these systems.
The costs associated with material usage within the manufacturing process can be reduced by exploring ways to reuse and recycle powder (minimizing waste), reducing the cost of used powders, and powder consumption too.These actions can also help achieving time savings and sustainability-related goals, such as minimizing material waste and the amount of used energy, during experimental materials development activities.In cases where the experimental development of new materials or expensive precious alloys is required, reducing the amount of required powder is the most effective approach.One option is using an PBF-LB/M system with a smaller build volume [12], although this may require significant capital investment if a new system needs to be purchased and installed.Another option is to develop a customized testbed PBF-LB/M system to enable precise control over the build volume, as well as to optimize the system for specific research tasks [8,12].There is a major disadvantage with this strategy, however, in that these types of testbed systems tend to be tailored for specific experiments, making them generally less reliable than commercially sold systems.A third option is to purchase or develop a build volume reduction kit for an existing PBF-LB/M machine [13,14].This kit would enable the machine to use less powder, hence, reducing the cost of materials involved with experimental materials development and testing activities.The advantage is that this option allows for expanding the capabilities of existing (reliable) infrastructure while reducing the need for large capital investment.
The aim of this study was the kit was designed to fit into an existing (modified) EOS EOSINT Mseries PBF-LB/M machine.The modular design of the developed reduction kit was specifically tailored to enable system customization and adaptation to align with research needs.By manufacturing simple cuboid samples using EOS 316L stainless steel powder, the operation of the manufactured reduction kit was finally validated.In this study, the reduction kit was designed to feature an additional detachable Pt-heater module, allowing for preheating the used substrate material above 500 °C.Taking this step enables the future studies on epitaxial solidification of PBF-LB/M-built functional materials.

PBF-LB/M system
The PBF-LB/M machine used in this study is a modified research machine that represents the EOSINT M-series -see [14] for specific machine details.The machine consists of a laser source, a galvanometric scanner, movement and laser exposure control software, and a build chamber with a powder re-coater system.The system was retrofitted with a new single-mode IPG YLS 200W SM CW ytterbium fiber laser with a wavelength of 1,070 nm and a maximum continuous wave average power output of 200W.The laser beam is transported from the laser to the galvanometric scanner optics (Scanlab hurry-SCAN 20 with F-theta lens) via optical fiber.Figure 1 shows the power intensity distributions of the laser beam at its focal point.The laser beam produced by the system was capable of near-Gaussian power distribution with the intensity in the center of the beam.The resulting focal point diameter of the laser beam is 83 µm at laser power of 80 W and 92 µm at laser power of 200 W.The atmosphere control system in the build chamber allows for the use of inert process gas, such as argon or nitrogen.The system has an inbuilt nitrogen generator with an approximate productivity of 20 m 3 /h.3D file manipulation (including slicing and generation of scanning vectors) and laser beam manipulation were conducted using SCAPS SAMLight scanner software with 3D functionality, whereas the physical movements of the systems axis (re-coater, powder feed, and the build platform position) were controlled externally using a G-code based software.

Design and characterization of the build volume reduction kit and detachable Pt-heater module
The build volume reduction kit and the additional detachable Pt-heater module were designed using Dassault Systèmes SolidWorks.Based on discussions and the subsequent manufacture of prototypes for initial testing, the design was gradually iterated to the final setup presented in this study.Whereas the main frame, some of the other smaller bulk components, and the final assembly were all implemented at the LUT machine shop and the LUT laser processing laboratory.Characterization of the detachable heater module performance was implemented using a calibrated IR camera (Fluke Ti10, USA) with the uncertainty of ± 2 °C or 2 % (-20 -250 °C) and a basic electronic multimeter equipped with K-type thermocouples with the uncertainty of ± 2.2 % (∼ 0 -293 °C) and ± 0.75 % (∼ 293 -1250 °C).By applying voltage to the Pt-heater module, using a DC-supply Tenma 72-10495 (0-30 V ± 0.01, 5 A ± 0.01), circulating current was measured to study the electric current vs. electric voltage characteristics.

Manufacturing of EOS 316L stainless steel test samples
The developed build volume reduction kit was tested by manufacturing test samples using EOS 316L stainless steel powder.The samples were melted on a 316L stainless steel substrate that was cut from a circular rod and machined down to the final dimensions of Ø45×15 mm 3 .Table 1 shows the chemical compositions of the used powder and the substrate.Figure 2(a) shows that the volumeweighted particle size distribution (d0.1 = 29.6 μm, d0.5 = 46.2μm, and d0.9 = 63.3 μm) of the powder was measured using Panalytical Morphologi G3S as well as Figure 2(b) shows the SEM secondary electron image of the powder particles using Hitachi SU3500 electron backscatter microscope.The process parameters used for manufacturing the EOS 316L test samples are summarized in Table 2.The samples themselves were cuboids with approximate target dimensions of 7×7×1 mm 3 .All samples were manufactured using a unidirectional scanning strategy (aligned along one of the side facets of each) without rotation of the scanning pattern from layer to layer.A fixed laser power of 200 W and layer thickness of 20 μm was also utilized for manufacturing while varying the applied scanning speed and hatch distance in three levels.Contour scans were melted using the same parameters as the hatched areas.The center-point values were determined based on the known optimal values for this material, as previously used in the same system without the reduction kit.During the experiments, the laser beam was focused on the powder bed surface.The test samples were deposited in a single patch (nine samples) on the 316L substrate, leaving a 1 mm wide gap between each to minimize thermal interactions during the process.The melted cuboids were ground incrementally along the build direction using SiC abrasives and then polished using 1 µm alumina particle suspension and a napless cloth.Polished cross-sections of the samples were inspected using a customized Zeiss Axio Scope.A1 optical polarized light microscope, and the relative densities of the manufactured samples were determined optically by measuring the area fraction of the pores within each sample.To facilitate comparison, a 316L sample manufactured using the known optimal values (scanning speed of 1000 mm/s and hatch distance of 100 μm) in the same PBF-LB/M system without the reduction kit was also ground, polished, and measured.

Design and characterization of the modular build volume reduction kit
The developed build volume reduction kit's structure and main components are shown in Figure 3.The kit is enclosed in the shielding gas chamber of the initial system, which includes inlet and exhaust ports for filling the shielding gas and providing a gas flow across the powder bed during builds.The focused laser beam passes through a window located on the top surface of the shielding gas chamber.The build volume reduction kit comprises two main components: the main powder reservoir and the 'manufacturing zone' with an integrated build platform and fixed depth reservoir for excess powder.The main powder reservoir (powder feed) and the build platform (powder layer height) were designed to be operated using the original in-built movement axis (z-direction) of the PBF-LB/M machine.Adopting this design approach effectively eliminated the need for building a complex setup of external electrical drives to produce the required motions.The powder feed mechanism selected for the build volume reduction kit was a type of simple cantilever hinge mechanism.This mechanism is directly mounted on the top surface of the original powder feed plate, serving as a relatively inexpensive solution to be used in the initial device setup.This structure is separate and fully modular, allowing for technical design iterations and possible setup replacements with a more conventional piston-style approach in the near future.By moving the machine's original powder re-coater linearly in one direction, the powder is 6 spread from the main reservoir to the build platform.The powder layers of the desired thickness are produced by lowering the build platform and, thereby, restricting the spread powder layer thickness to a fixed value.The build plate is installed on a circular piston head that is mechanically connected to the initial machine's larger build platform surface.A Teflon ring around the piston's top edge prevents powder from falling through the gap between the frame of the assembly and the piston itself.The build plate is attached to the head of the piston using a screw.

Design and characterization of the detachable Pt-heater module
The main motivation for designing and building a detachable Pt-heater module can be explained by a special research need involving experimental investigations on the epitaxial solidification of PBF-LB/M-built functional materials.To significantly lower the thermal gradients during melting in PBF-LB/M [15], this approach requires the use of substantial substrate heating (above 500 °C).In the case of Ni-Mn-Ga-based magnetic shape memory alloys, the used substrate may be either polycrystalline or, in specific solidification studies, also single crystalline, which sets limits to the maximum size of the used substrate -typical Ni-Mn-Ga single crystals are less than 20 mm in diameter [16] The detachable Ptheater module was designed, therefore, to facilitate the use of small (below the aforementioned size) substrates.Figure 4 shows the designed detachable Pt-heater module, which includes a clamp with a top surface that is heated linearly using a PID controller.PID (Proportional-Integral-Derivative) is a control algorithm commonly used in engineering systems to achieve precise and efficient control of a heating process.The main body of the heater module consists of a Ø45×70 mm 3 alumina block that acts as a thermal isolator preventing the flow of heat from the Pt-heaters into the build platform piston and the reduction kit frame.The main heater assembly is placed on top of the alumina block and consists of a rectangular clamp measuring 20×10 mm 2 and four Pt heater elements sized 5.2×3.9×1mm 3 .The build plate, which is a 7×7×1.5 mm 3 substrate (single crystal or polycrystal), is fixed to a rectangular support assembly that is connected rigidly to the vertical stage (z-axis) using two screws.This module can also be mounted on top of the build platform piston -again, also by using screws.The piston (and the alumina block) requires holes for transferring the heater element and thermocouple wires through the piston, thus, preventing the wires from interacting with the powder re-coater or the powder itself during builds.
The temperature of the Pt microheaters is controlled by adjusting the voltage, whereas a K-type thermocouple is used to measure the temperature of the clamp and monitor the temperature increase.This facilitates compensating for any error caused by the physical separation between the clamp and the Pt-heaters.By monitoring the temperature of Pt heater elements using a calibrated IR camera and establishing a linear proportional relationship with the clamping surface, the temperature measurement of the samples was enhanced.Figure 5(a) shows the electric current vs. electric voltage characteristics of the used Pt-heater elements the heating rate of 55 °C min -1 (test 1) and 110 °C min -1 (test 2) to reach maximum temperatures of ∼272 °C, and ∼540 °C for in a direct current mode.The investigation of the current (I) versus voltage (V) characteristic of the sample reveals that the Pt heater elements exhibit nonlinear behaviour, as well as showing that the curve in Figure 5(a) can be divided into two distinct regions: 1) a nonlinear zone wherein the electrical conductivity of the Pt heater elements undergoes an abrupt change from a high value to a low value, resulting in a reduction of electrical resistance, and 2) an upturn zone that exhibits a moderate increase in electrical resistance.As the Pt heater elements can be utilized without experiencing a 24 V breakdown voltage, employing them for heating the sample as a substrate with low overvoltage is suggested.Figure 5(a) demonstrates that any rise in clamp temperature can be controlled by adjusting the I -V characteristics.As illustrated in Figure 5(b), the electrical resistance of the Pt heater elements was measured by applying input power.The resistance (R) of the Pt heater elements resulting from Joule heating with input power (power consumption) is plotted.The temperature of the clamp and Pt heaters, as well as their relationship with the input power, exhibit nearly identical linear behaviour -consistent with the data obtained from an IR camera.These measurements indicate that some heat losses occur due to the physical gap between the clamp and Pt heater elements.

Fabrication of the EOS 316L stainless steel cuboids
By manufacturing test samples using EOS 316L stainless steel powder on a 316L stainless steel substrate, the developed build volume reduction kit was tested.During this, two process deviations were observed that affected the manufacturing results.Firstly, it was found that the build platform pistonresponsible for the substrate movement -was not securely attached to the assembly.As a result, the piston started rotating during the build process, causing slight rotational drifting of the scanning pattern.By modifying and reinforcing the piston mounting to the main platform of the system, this issue can easily be resolved to ensure stable and precise movement without any rotational deviations.Secondly, the powder spreading process resulted in the powder being spread over two wide areas, causing some of the powder to flow sideways over the edges of the designed reduction kit frame.This resulted in the loss of powder that, from an economic standpoint, is undesirable.To address this issue, reducing the width of the active section of the powder re-coater that comes into physical contact with the powder is recommended.By reducing the width, the flow of powder to the sides can be minimized, thus, reducing wastage, and optimizing the powder utilization.Despite these minor process deviations, the results indicate that the build volume reduction kit was effective in controlling thermal interactions during the build.Figure 6(a, b) shows example crosssections and relative densities of each test sample and the reference sample.The shown errors correspond to the standard deviations of the measurements.As can be observed from the figure, the relative densities of the test samples manufactured using scanning speeds of 950 and 1000 mm/s and hatch distances of 90 and 100 μm were relatively close to the reference sample manufactured using the same PBF-LB/M machine without the build volume reduction kit installed.However, the samples manufactured using scanning speeds of 1100 mm/s and hatch distances of 110 μm exhibited significantly lower relative densities -see Figure 6(c).These samples also exhibited larger standard deviations of relative density.The observed drop of relative density values is drastic considering the relatively small change in the applied processing parameters, which implies that the optimal processing window for the used material shifts slightly towards lower scanning speed and hatching distance values when using the reduction kit.The observed shifting of the optimal processing window likely occurs due to the different thermal diffusion and heat accumulation characteristics between the substrates in these two setups (with and without the reduction kit).Besides influencing relative density, this effect may also induce minor microstructural differences between the samples manufactured using the initial system and the samples manufactured using the reduction kit.Future studies using this system will focus on benchmarking these differences and conduct more thorough microstructural investigations on the produced materials.
However, the successful test samples highlight the suitability of the designed build volume reduction kit for manufacturing precise and dimensionally accurate components, thus, facilitating the intended experimental materials development and testing activities.However, addressing the observed deviations and implementing the recommended modifications will further enhance the system's performance and efficiency.Overall, the results demonstrate the successful implementation of the build volume reduction kit.

Conclusions and further studies
Laser powder bed fusion (PBF-LB/M) serves as a promising additive manufacturing process that enables the production of complex and high-performance parts.Nevertheless, the high material costs and the need for large quantities of powder in conventional industrial-grade systems pose challenges for experimental materials development and testing activities.To address these challenges, this study developed a modular build volume reduction kit that was then integrated into an existing (modified) EOS EOSINT M-series PBF-LB/M machine.The reduction kit allows for the use of smaller quantities of powder, thus, reducing material costs and enabling cost-effective research on new exotic alloys or precious metals.The kit was successfully designed and characterized before its operation was validated by manufacturing test samples using EOS 316L stainless steel powder.The results demonstrate the feasibility of using the reduction kit to conduct experimental investigations in a cost-effective manner.Moreover, a detachable Pt-heater module was developed to facilitate studies on the epitaxial solidification of PBF-LB/M-built functional materials.This module enables precise control of the substrate temperature above 500 °C, which is crucial for the formation of specific microstructures.The module's thermal behaviour was characterized using an IR camera and K-type thermocouples, with results showing that the module effectively provided the desired temperature control and could be used for experiments involving the solidification of materials like Ni-Mn-Ga-based shape memory alloy.In summation, the developed build volume reduction kit and detachable Pt-heater module offer valuable tools for researchers working with PBF-LB/M, specifically for enabling cost-effective materials development and precise temperature control for specialized studies of functional materials.

Figure 1 .
Figure 1.Power intensity distribution of the near-Gaussian-shaped laser beam in focal point at (a) 80 W and (b) 200 W.

Figure 2 .
Figure 2. (a) Powder particle distribution of the used EOS 316L stainless steel powder; (b) SEM secondary electron image of the powder particles.

Figure 3 .
Figure 3. Schematic of the designed build volume reduction kit.(a) the excess powder reservoir and the build platform with an installed stainless-steel substrate, (b) secondary (Argon or Nitrogen) gas inlet and the powder re-coater, (c) operation of the main powder reservoir, and (d) build platform assembly with the z-axis piston and the detachable Pt-heater module.The flat grey sections in the central image show the structure and positions of the original PBF-LB/M system's main chambers.

Figure 4 .
Figure 4. (a) 3D model of the designed detachable Pt-heater module, and (b) a manufactured heater module with an externally mounted K-type thermocouple used for characterizing the module's thermal behaviour.The inset shows a capture from the IR camera data used for monitoring thermal distributions during heating.

Figure 5 .
Figure 5. Characterization of the performance of the used Pt-heater elements during test 1 (maximum temperature ∼272 °C), test 2 (maximum temperature ∼540 °C).(a) electric current vs. electric voltage and the electric current vs. heater temperature (T), and (b) electric resistance vs. power consumption and the electric resistance vs. heater temperature.

Figure 6 .
Figure 6.(a) Optical images of the EOS 316L test samples built using the reduction kit.The inset markings indicate relative densities for each sample, wherein the shown error corresponds to the standard measurement deviation.(b) Optical image of the EOS 316L reference sample built without using the build volume reduction kit.(c) Lowess surface fit showing the main effects of the applied parameters on relative density of the built samples.

Table 2 .
The applied process parameters for manufacturing the EOS 316L cuboids.

Table 1 .
Compositions of the used materials in weight percentage (wt.%).