Additive manufacturing capabilities for heat switch technology: Key challenges & knowledge gaps

Copyright © 2022 by Jakub Mašek et al. Published by the EUCASS association with permission. Additive manufacturing capabilities for heat switch technology: Key challenges & knowledge gaps MAŠEK Jakub1* 1 Brno University of Technology, Institute of Aerospace Engineering, Technicka 2, 61669 Brno, Czech Republic * [email protected] LÖFFELMANN František, POPELA Robert, KUBÍK Petr, ŠEBEK František, KOUTNÝ Daniel, MALÝ Martin, PANTĚLEJEV Libor2, PAMBAGUIAN Laurent3 2 Brno University of Technology, Faculty of Mechanical Engineering, Technicka 2, 61669 Brno, Czech Republic, Frantisek.Loffelma[email protected]; pop[email protected]r.cz; [email protected]tbr.cz; [email protected]; Daniel.Kou[email protected]; [email protected]; [email protected] 3 European Space Agency (ESA - ESTEC), Materials and Processes Section (TEC-MSP), Keplerlaan 1, NL-2200 AG Noordwijk, The Netherlands, Laurent.Pamb[email protected]t Abstract The paper is to provide an overview of the key challenges and knowledge gaps in additive manufacturing of metals applied to two parts of a novel heat switch technology – a Baseplate and a Flexible thermal structure. Additive design, optimization and manufacturing capabilities as well as quality of produced parts were investigated. The key challenge was to manufacture bio-inspired structure without internal supports and assure 190 surfaces to be in a contact at the same time for efficient heat transfer. Finally, modern trends in additive as multi-material design and manufacturing or effect of defects are discussed for further evolution of parts. 1. Introduction Additive manufacturing of metals, as a growing market, has a lot to offer in terms of design-manufacturing capabilities and sustainability [1]. Therefore, it attracts key players and entrepreneurs in the space industry to reach better performance of parts at a lower mass. One of the game-changing applications is thermal hardware [2; 3]. Thermal management is an integral part of each spacecraft. It is necessary to ensure the thermal regulation for internal electronics and payload to allow them to work in suitable conditions. A phase change material-based (PCM) heat switch [4; 5] is a one type of many [6; 7] studied in literature. The switch technology can be divided in passive and active devices that differ in temperature and conductivity parameters as well as in the physical principle of function. Selective laser melting (SLM) technology is a currently leading innovative manufacturing process among the additive technologies that accelerates changes in many engineering areas, particularly the aerospace [8; 9]. Several studies investigated space components optimized and produced by SLM technology from AlSi10Mg powder, as brackets [10; 11], CubeSat frame [12], antennas and waveguides [13] or heat pipes [14]. The benefit of printing the components layer-by-layer enables a different approach in design and manufacturing process. Costs, weight and material wasting can be reduced while producing parts with the same or even better mechanical properties [15; 16]. However, only few additive parts were qualified for space flight [11; 17; 18]. The technology itself creates micro and macro structure defects, porosity, thermal deformations and poor surface quality [19; 20; 21]. Effect of defects [22] and related inspection and post-processing technologies are among the most critical parameters for larger space application and qualification of additive parts. Higher amount of porosity, lower geometry and surface precision are the key disadvantages of aluminium-based materials despite of easier SLM manufacturing and larger knowledge base of material properties [23; 24]. DOI: 10.13009/EUCASS2022-6196 9ᵀᴴ EUROPEAN CONFERENCE FOR AERONAUTICS AND SPACE SCIENCES (EUCASS) Jakub Mašek et al. 2 A miniaturized heat switch [25; 26] (Figure 1) is a device under the development dedicated to efficient, autonomous and powerless temperature regulation based on thermo-physical properties of paraffin. The technology repeatedly couples or decouples the heat source with external radiator that dissipates the excess of energy to the environment. It shall have two-position on/off thermal transfer up to 10 W. The switch shall work in deep space and Martian conditions to withstand temperatures from – 125 °C to + 60 °C. The implementation of additive technology is expected to enhance the design towards an operational sample. The paper is to provide an overview of the key challenges and knowledge gaps in additive SLM manufacturing applied to two parts of the novel heat switch technology – a Baseplate and a Flexible thermal structure. In addition, to provide a context of modern trends in additive for further evolution of parts. 1.1 Legacy issues of parts Baseplate (BP) part represents a half of paraffin pressure container (Figure 1a). The component (Ø32 mm and 17.6 mm in height) of mass roughly 30 g shall withstand an internal pressure of up to 16 MPa, while the deformations reach less than 0.02 mm. The Baseplate shall allow paraffin melting and on the contrary, prevent its leakage. The flatness of the plate is essential to provide a contact interface to an external heat source for efficient heat transfer. However, paraffin pressure driving the actuator causes in the ‘on’ mode swelling of the switch so that the contact surface might be significantly reduced (Figure 2b) [26]. The goal is therefore to optimize the thermal-stiffness properties by adding internal reinforcements at a minimum weight. Figure 2: Baseplate part; (a) former design; (b) temperature-to-deformation dependence based on paraffin pressure The second part, Flexible thermal structure (FTS), represents the path of heat transfer. The component (Ø56 mm and 24.1 mm in height) of mass roughly 70 g has a two-position on/off mode allowing a vertical linear movement and a thermal conductivity higher than 1.5 W⋅K-1. When the temperature of paraffin rises, the actuator pushes the flexible structure to prolong by 1.7 mm (7.6 %) and creates the heat conductive path. Figure 3: Flexible thermal structure; (a) former design; (b) design parameters The Flexible structure (Figure 3a) was initially made of copper braid soldered to two copper plates. However, the welded joint and the copper braid proved to be unacceptable. A micro-computed tomography inspection revealed that melted copper had not penetrated in between the wires and instead created a mechanical joint. It resulted in a low Temperature of paraffin [°C] Paraffin pressure* Deflection [MPa] [mm] 20 7.8 0.122 40 12.5 0.193 60 16.4 0.251 * MHS internal paraffin pressure was estimated based on the measured deflections; CAD models were analysed by finite element analysis FEA at different loads of paraffin pressure to reach the same deformations. Parameter Current value Requirement Thermal conductivity 0.365 W∙K-1 > 1.5 W∙K-1 FTS weight (3-parts assembly) 55.1 g < 55.1 g Specific thermal ratio 6.6 W∙K-1∙kg-1 > 27.2 W∙K-1∙kg-1 Figure 1: Miniaturized heat switch schema (dimensions in millimetres) (a) (b) (a) (b) DOI: 10.13009/EUCASS2022-6196 ADDITIVE MANUFACTURING CAPABILITIES FOR HEAT SWITCH TECHNOLOGY: KEY CHALLENGES & KNOWLEDGE GAPS 3 measured thermal conductivity of 0.36 W∙K-1, more than 4-times lower than required (Figure 3b). Additionally, the length of each individual wire was twice the height of the part and the heat was transferred along the wires rather than across the points where the wires were in contact [25]. 2. Material and Methods 2.1 AlSi10Mg powder The powder AlSi10Mg was produced by SLM Solutions Group AG, Lübeck, Germany using a gas atomization in nitrogen atmosphere. The morphology of similar powder was evaluated in the study of Vrána et al. [27] based on the scanning electron microscopy (Philips XL30 SEM, Amsterdam, The Netherlands) and showed almost a spherical shape of the particles. The powder particles had a large variability in size with a mean value of 41.4 μm and 90 % were up to 58.0 μm as investigated by the laser diffraction analysis (Horiba LA-960, Kyoto, Japan). Chemical composition defined by the vendor is shown in Table 1. Table 1: Chemical composition (wt %) of SLM material AlSi10Mg [28] Element Si Fe Cu Mn Mg Zn Others Al SLM powder 9.0-11.0 0.55 0.05 0.45 0.20-0.45 0.10 0.45 Balance Before the component fabrication started, the metal powder was dried to less than 5 % of residual humidity and subsequently filled into a recoater. All the samples were built on a platform (280 x 280 mm) made of aluminium alloy Al 3.3547 that was preheated to 150 °C. A 50 μm layer was applied corresponding to the particle size distribution. 2.1.1 Mechanical and thermal properties The AlSi10Mg aluminium alloy metal powder was chosen over other materials due to its thermo-mechanical specific properties after artificial ageing. Different heat treatments were investigated by Vaverka et al. [29]. Two stage heat treatment - annealing and artificial ageing (T6: 520 °C/ 6 hours/ water quench/ 175 °C/ 4 hours) showed higher values of yield stress (YS) and on the contrary by 30 % lower ultimate tensile stress (UTS) compared to as-build. Table 2: Mechanical properties after heat treatment – artificial ageing [29] YS UTS Elongation R.S.* Thermal conductivity Rp0.2 [MPa] Rm [MPa] [%] [MPa] [W∙m-1∙K-1] 258 288 2.8 -17 163.1 * Compressive residual stress according to finite element (FE) analysis 2.1.2 Thermal conductivity analysis Thermal conductivity was investigated for SLM fabricated and artificially aged samples 10x10x2 mm of AlSi10Mg alloy. Two main surfaces were polished and matted with graphite powder. Two samples were measured at the Institute of Plasma Physics, Czech Technology Agency (IPP-CAS) by a laser-flash method in vacuum (Linseis LFA 1000, Robbinsville, USA) with 5 measurements per sample. A deviation of measurements per one sample was ± 4 %. The samples were analysed by micro-computed tomography to reveal the porosity of 0.10 % at 10 μm resolution. The result of average 163.1 W∙m-1∙K-1 matches well the study of Sélo et al. [30] that refers to values of 145 ÷ 173 W∙m-1∙K-1 for samples heat treated by annealing. 2.2 Manufacturing and inspection 2.2.1 Selective laser melting Additive production of both technology demonstrators was performed on an industrial 3D printer SLM 280HL (SLM Solutions Group AG, Lübeck, Germany) equipped with 400 W ytterbium fibre laser YLR-400-WC-Y11 (IPG Photonics, Oxford, USA) with a focus diameter of 82 µm and a Gaussian beam intensity distribution. All parts were produced on two platforms under the process parameters (Figure 4c) developed to achieve a homogeneous component, low porosity and high productivity. Nitrogen gas flow was used during fabrication and the oxygen level was kept under 0.2 % threshold. DOI: 10.13009/EUCASS2022-6196 Jakub Mašek et al. 4 Laser power Scanning speed – borders Scanning speed – volume Beam compensation Layer thickness Hatch distance Beam diameter 350 W 500 mm∙s-1 930 mm∙s-1 150 μm 50 μm 150 μm 82 μm (a) (c) [27] Figure 4: SLM manufacturing; (a) platform with Baseplate and Flexible thermal structure parts after SLM manufacturing; (b) two walls with different thickness - laser scanning strategy generated as default for thin wall structures; (c) SLM laser process parameters The ‘’Offset filling’’ scanning strategy was used for the thin walls and support structures, Figure 4b. All parts were printed 5 mm above the platform on supports consisting of perforated blocks and cones. Support structures were generated to support surfaces with an inclination lower than 35° and with an area higher than 0.1 mm2. 2.2.2 Porosity analysis Internal porosity was analysed using a micro-computed tomography (μCT, GE phoenix v|tome|x L240, GE, Wunstorf, Germany). By each μCT measurement, two Baseplates, two Flexible structures or four thermal samples were jointly analysed. The linear voxel size resolution of 1/1000 of the largest sample size was 60 μm for the Baseplate and 75 μm for the Flexible structure. The post-processing of reconstructed data was performed in the software VGStudio MAX 3.1, including the porosity analysis module. In order to inspect the internal invisible surfaces of parts, scanned 3D data were also exported in STL format for dimensional and deformation analyses in GOM Inspect software. 2.2.3 Dimensional and deformation analysis Visible surfaces of the Baseplate and Flexible structure were digitized by the optical 3D scanner Atos Triple Scan 8M (GOM GmbH, Braunschweig, Germany) with the optics MV170 to understand the deformations of parts in as-build and after artificial ageing. To successfully scan Al-alloy material that is highly reflective, parts were matted by Ti02 titanium powder with a thickness of 2 - 3 μm. The blue-light scanner resolution was then ± 0.005 mm. After the optical measurement, GOM Inspect software was used to analyse the dimensional deviations from 3D CAD model. 2.2.4 Post-processing of parts after SLM fabrication After the production of SLM, all parts underwent a sequence of post-processes: de-powdering - inspection - heat treatment T6 - inspection - cut-off parts from platform - supports removal - inspection and surface treatment by sand blasting. Two inspections by the blue light 3D scanning were planned before & after heat treatment to evaluate the rising inaccuracies. The support structures were removed by a combination of Wire Electric Discharge Machining (Wire cut EDM, CHMER G32S, Taiwan) and manually. Subsequently, the parts were inspected by the microtomography to understand the precision of SLM production as well as the internal defects and invisible surfaces. The outer surfaces were treated by sandblasting (Suction Blast Cabinet SBC 420L with manual blast gun) to visually unify the parts. The direction of beam and time exposure was controlled manually by using abrasive of corundum particles. 2.3 Design of additive parts for heat switch 2.3.1 Baseplate design definition The flatness of the Hot interface is essential to provide a contact surface for efficient heat transfer. Structural reinforcement encourages couple of design solutions, such as topology, parametric or multi-material optimization. The aim was to reach the best mass-to-stiffness ratio and to demonstrate the capabilities of conventional software tools and Contour Fill-contour (offset filling) Possible lack of fusion Path of laser scanning (b) Hatch DOI: 10.13009/EUCASS2022-6196 ADDITIVE MANUFACTURING CAPABILITIES FOR HEAT SWITCH TECHNOLOGY: KEY CHALLENGES & KNOWLEDGE GAPS 5 manufacturing processes. Based on the stiffness requirement (deflection < 0.02 mm), the height of the design domain was artificially increased by 5 mm compared to the former part, see dimensions in Figure 5. 2.3.2 Flexible thermal structure design definition FTS mechanism, as part of the Miniaturized heat switch technology, represents the path of heat transfer. It shall be a thermally optimized structure with a specific thermal conductivity > 2.72 W∙K-1∙kg-1. The design space (Figure 6) was derived from the prior design with three cut-outs for pylons. Outer diameter was enlarged compared to the baseline design, taking into account the new additive technology application and a high thermal conductance requirement. The cross-section area of 1400 mm2 was taken as the maximum size of the former heat switch [25]. (a) (b) Figure 5: Baseplate non-design and design geometry (available for optimization); (a) BP baseline definition and comparison to former design (dimensions in millimetres); (b) ISO view of BP geometry (a) (b) Figure 6: FTS design space available; (a) design space geometry cross-section [25]; (b) ISO view 2.4 FE numerical models The topology optimization and finite element analysis (FEA) of the Baseplate were done using Tosca 2019 and Ansys 2019 R3. One sixth symmetry was used. The model contained the half of the pylon made from Torlon and simplified screw, all discretized by hexahedral elements C3D8R of 0.035 mm. Normal displacements were constrained to nodes on the planes of symmetry. The pressure of 16 MPa was applied to inner surfaces of the paraffin box. The nodes in upper surface of the Baseplate were tied with the reference point (Figure 7a) which enabled only vertical axial displacement. A reaction force was applied to the reference point in the axial direction and was equivalent to the distributed pressure load. The pylon, screw and base plate were connected with tie constraint. Material properties are listed in Table 3. Objective function of the topology optimization was defined to minimize mass subjected to constraints for deformation (< 0.02 mm), maximum stress (σvonMises ≤ 110 MPa) and manufacturability (minimum wall thickness 0.3 mm or 2.0 mm prescribed as minimum element size). Restriction of maximum vertical relative displacement dz [mm] was prescribed for nodes on the Hot interface (Figure 7a) according to equation −0.02 mm < ∆(dzmax – dzmin) < 0.02 mm. (1) Pylon pads Former design Non-design geometry Pylon and screw connection Hot interface - flat contact Design geometry Paraffin box cap and sealing DOI: 10.13009/EUCASS2022-6196 Jakub Mašek et al. 6 (a) (b) Figure 7: FE model set-up used for analyses and topology optimization; (a) BP boundary conditions; (b) FTS boundary conditions Heat transfer analysis of the Flexible structure was carried out by Ansys 2019 R3 solver. Rough mesh consists of second order tetrahedral elements TET10 with element size 0.5 mm. Heat flux Q [W] is prescribed to one of the interfaces (10 W) and temperature (T1 = 15 °C) is prescribed to the opposite interface. The average surface temperature (T2) is used to calculate thermal conductivity q [W∙K-1] of the FTS structure from temperature difference according to equation 𝑞 = 𝑄 𝑇2−𝑇1 [W∙K-1]. (2) The thermal contact within the mechanism is set to bonded with surface heat transfer coefficient of 2127.9 W∙m-2⋅K-1 which equals to Al-Al contact of 0.4 MPa contact pressure and both surface roughness of Ra 1.6 µm. Heat leakage by radiation and convection was not considered. Table 3: Material properties used in topology optimization and FEA analyses [28] AlSi10Mg (additive) Torlon 4203® AISI 316L Thermal Conductivity [W∙m-1∙K-1] 150 0.26 16 Density [g∙cm-3] 2.69 1.42 7.8 Young´s Modulus [GPa] 70.9 4.48 210 Poisson´s Ratio [-] 0.33 0.45 0.27 3. Results 3.1 Baseplate Design Multi-material design of Baseplate was successfully optimized, however was not to meet design requirements [26]. Therefore, out of many trials, three competitive designs were selected. Two geometry concepts are based on topology optimization with different restriction for minimum wall thickness 0.3 and 2.0 mm (Figure 8a,b) and one geometry is based on the conventional design of reinforcing ribs based on the designer’s experience (Figure 8c). To improve the stiffness of geometry BP(2) and BP(3), ISO grid (w/h 1:3) was designed with height of 1.5 mm and spacing 1.43 mm. The grid under 75° angle was limited by 1/6 symmetry and thus each line is not continuous in the neighbouring section. Generated ISO grid can be observed in Figure 8-1c (partially also in Figure 8-2,3b) and was identical for both aforementioned geometries. Design parameters of the ISO grid were not optimized. 3.1.1 Finite element analysis of Baseplate Finite element analyses of the design concepts were performed taking into account its 1/6 symmetry (Table 4). BP(1) and BP(3) concepts meet the design requirement of deflection in vertical axis, whereas BP(2) reaches improvement by roughly 12 % due to the ISO grid and has stiffness lower by 50 % compared to the design target. Nevertheless, the geometry could be improved by the increased height of the ISO grid and by the application of reinforcement near the connection of the pylon pads to the vertical wall. Analyses of fatigue life were performed but the stress concentrators were not suppressed, primarily for the uncertainty in the pre-selected paraffin pressure. Therefore, low-cycle-fatigue results are much lower than the required one hundred thousand. It corresponds to results of the static equivalent stress (von-Mises) that show strong concentrations of stress. More specifically, located at the edges of powder holes and in areas where small or none radius (fillet radius) exist. Reference point ½ of pylon and simplified screw Max. displacement on Hot interface Heat flux Q = 10 W T1 = 15 °C Al-Al bonded thermal contact with 0.4 MPa pressure DOI: 10.13009/EUCASS2022-6196 ADDITIVE MANUFACTURING CAPABILITIES FOR HEAT SWITCH TECHNOLOGY: KEY CHALLENGES & KNOWLEDGE GAPS 7 BP(1) BP(2) BP(3) (a) (b) (c) Figure 8: Three Baseplate competitive geometries (a) BP(1) topology optimized with min. wall thickness 0.3 mm; (b) BP(2) topology optimized with min. wall thickness 2.0 mm; (c) BP(3) with conventional design of reinforcing ribs; (1) supports structures default settings; (2) pre-print modifications and supressed internal supports; (3) optical 3D scan results after heat treatment – artificial ageing Table 4: Results of Baseplate design concepts - FEA analyses overview BP(1) BP(2) BP(3) Thermal Conductivity [W∙K-1] 2.99 2.61 2.72 Weight [g] 30.56 26.55 28.78 Deflection [mm] 0.0166 0.0299 0.0192 Specific deflection [mm∙kg-1] 0.543 1.126 0.667 Fatigue Life [-] 1600 5500 25000 3.1.2 Baseplate design to be printed Prior to the fabrication, geometry was modified for manufacturing constraints, eliminating closed cavities, too small features < 0.3 mm and surface angles < 35°. On the contrary, holes for powder removal, identification labels and support structures were added (red arrows in Figure 8-1,2). The holes for powder removal and paraffin filling were cut into a paraboloid wall in geometry BP(1) and BP(2); 3 holes per part, radially oriented in a location of pylon pads (Figure 8-2a,b). Analyses showed in this region lower loads and therefore also less reinforcement structure appears. The cut-outs for holes do not affect inner reinforcement and therefore do not decrease the mechanical properties of BP. Based on the orientation of struts, walls and surfaces, the most promising orientation of 0° was selected to minimize the volume of supports needed. Generated supports solely inside the BP design are not removable by default (Figure 8-1a,b). Therefore, the primary aim for successful manufacturability was to supress the necessity of internal supports structures and avoid a powder locking inside the closed cavities. 3.2 Design of Flexible thermal structure Out of many design concepts and optimization trials, a two-part mechanism design with 5 contact levels was selected (Figure 9). The design based on stems and branches in a “tree shape” with variable thickness is to meet the design requirements. All contact surfaces shall be in contact at the same time, which demands strict requirements for the (1) (2) (3) DOI: 10.13009/EUCASS2022-6196 Jakub Mašek et al. 8 additive manufacturing itself and the precision of the surface finishing operations. The higher the surface roughness, the higher is the thermal contact resistance and the lower the efficiency of the FTS mechanism. The build orientation has also a fundamental effect on the surface quality and manufacturability. (b) (c) Figure 9: FTS mechanism; (a) FTS geometry modification according to surface treatment technology; (b) FTS(WEDM) in 0° orientation after SLM fabrication; (c) FTS(EDM) in 90° orientation – preparation of supports; (d) FTS(ECM) in 90° orientation after heat treatment – optical scanning results 3.2.1 Finite element analysis of FTS mechanism Figure 10c shows, among others, the specific thermal ratio (thermal conductivity divided by weight) to meet the design target of 27.2 W∙K-1∙kg-1. Thermal conductivity and contact force were designed roughly 30 % higher than required however having negative impact on weight. The specific thermal ratio represents the key property of each FTS structure (i.e. of the distribution of material in horizontal and vertical direction within the design volume). Assuming that the specific parameter is constant and that the contact force and weight are directly proportional to the conductivity, the FEA results were recalculated with a target conductivity of 1.5 W∙K-1. The parameters decreased by 27.9 % and are newly to meet the target values. The theoretical analytical calculation would be then converted to a real design modification by eliminating some thermal stem-branch structures. The specific thermal ratio is 10.3 % higher than required and might mitigate material and manufacturing uncertainties. Additionally, a variable thickness of stems increases the overall part´s thermal conductivity by roughly 10.9 % by optimization of the heat flux while keeping the same weight, see Figure 10a,b. This effect was not included in the results of FEA presented in Figure 10c. FTS FTS (theoretical) Weight [g] 69.26 49.98 Thermal conductivity [W∙K-1] 2.08 1.50 Specific Thermal Ratio [W∙K-1∙kg-1] 30.01 30.01 Force* [N] 1071.0 773.1 *To create 0.4 MPa contact pressure (a) (b) (c) Figure 10: FTS mechanism; (a) heat flux of FTS geometry cut-out; (b) optimized heat flux of FTS geometry cut-out with variable thickness of stems; (c) FEA results of Flexible structure Stems Branches Lower and upper plate (interfaces) Machining allowances Supports – cones and perforated blocks Cut-outs for pylons Technological holderscubes 0° 90° WEDM modification EDM/ECM modification (a) (d) 0 DOI: 10.13009/EUCASS2022-6196 ADDITIVE MANUFACTURING CAPABILITIES FOR HEAT SWITCH TECHNOLOGY: KEY CHALLENGES & KNOWLEDGE GAPS 9 3.2.2 Flexible structure design to be printed To ensure that all 190 surfaces are at the same time in contact for efficient heat transfer, only three unconventional surface-finishing technologies were identified as applicable – Electrical Discharge Machining (EDM), Wire-EDM and Electro-Chemical Machining (ECM). However, each finishing technology demands its own specific geometry modification. The geometry for WEDM technology was printed as one part and subsequently was cut by the wire into two sections. On the contrary, for ECM and EDM technologies, the mechanism was printed as two parts ‘assembled’ together with a pre-defined clearance (Figure 9a). All manufacturing steps are leading to the same geometry at the end of the production cycle. Four configurations of mechanism, WEDM modification in 0° and 90° (Figure 11e) orientations and EDM and ECM in 90° orientation (Figure 11d), were selected to assure manufacturability, removal of supports and surface treatment, while considering the position of all local surfaces. Despite the similar geometry, it was not possible to print the EDM and ECM configurations in 0° orientation due to the position of the branches in the upper part. Prior to the fabrication, each geometry was modified for manufacturing constraints, adding machining allowances, support structures, technological holders-cubes, cut-outs for pylons and identification labels to assure fixation, clamping, easier machining and tracking of parts during all processes. Machining allowances were added to all contact surfaces to be subsequently machined (Figure 9b,c and 11e,f). 3.3 SLM manufacturing Elements of concurrent design and manufacturing engineering were implemented to save time over the development cycle of parts [31]. Therefore, different geometry-based concepts of the Baseplate and manufacturability-based modifications of the Flexible structure were designed and subsequently manufactured by SLM. Production of parts (Figure 11) was roughly 8.5 hours per platform, consisting of 480 layers of the Baseplate and 882 or 1220 layers of the Flexible structure according to orientation. (a) (b) (c) (d) (e) (f) Figure 11: SLM manufactured parts; (a) BP(1) with 0.3 mm wall thickness constraint; (b) BP(2) with 2.0 mm wall thickness constraint; (c) BP(3) with ribs and ISO grid; (d) FTS build in 90° orientation in WEDM/EDM/ECM modification; (e) FTS build in 0° orientation in WEDM modification; (f) FTS after machining 3.4 Dimensional and porosity inspection An overview of dimensional inaccuracy progression after the heat treatment is given in Table 5. Based on the 3D scanning of parts attached to the platform, majority of deviations in as-build are within ± 0.10 mm, while after the heat treatment larger deformations were measured mostly lower than ± 0.20 mm. However, the deviations in specific thinwall locations reach up to ± 0.25 mm, as in Figure 8-3c. The part ECM (90°) apparently released strong residual stresses into an extensive deformation of the top-free end, up to + 0.19 mm in as-build and up to ± 0.70 mm after artificial ageing (Figure 9d). The effect is alike for all parts, but in smaller scale. No threshold was set-up for maximum deformation acceptance. No crack was observed. WEDM ECM EDM Identification labels 0° 90° WEDM DOI: 10.13009/EUCASS2022-6196