Received: 13 July 2022 | Revised: 2 September 2022 | Accepted: 4 September 2022 DOI: 10.1002/appl.202200062 RESEARCH ARTICLE Infiltration of aluminum in 3D‐printed metallic inserts Helder Nunes1 | Omid Emadinia1 | José Costa2 Rui Soares1 | José Silva1 | Inês Frada3 Manuel Vieira1,2 | Ana Reis1,4 1 Advanced Manufacturing Processes, Institute of Science and Innovation in Mechanical and Industrial Engineering (LAETA/INEGI), Porto, Portugal 2 Department of Metallurgical and Materials Engineering, University of Porto, Porto, Portugal 3 Centre for Innovation and Technology N. Mahalingam, Águeda, Portugal 4 Department of Mechanical Engineering, University of Porto, Porto, Portugal Correspondence Helder Nunes, Institute of Science and Innovation in Mechanical and Industrial Engineering (LAETA/INEGI), 4200‐465 Porto, Portugal. Email: hnunes@inegi.up.pt | Rui Madureira1 | | Vitor Anjos3 | Filomena Viana1,2 | Abstract Aluminum structural composites through the infiltration process can be performed by vacuum, centrifugal, or squeeze casting, involving the infiltration of molten Al into fibers, particles, foams, or even porous preforms. This methodology creates hybrid structures of two distinct metal alloys that can be used to locally strengthen components or even to improve the properties of bulk materials, such as ultimate tensile strength and thermal conductivity. New approaches involve the infiltration of liquid Al into a three‐dimensional (3D)‐printed structure of the more rigid metal, such as steel, that the Al matrix. In the current study, stainless steel and copper inserts were produced by fused filament fabrication techniques with various geometries. Moreover, some 3D inserts were electrochemically coated with pure copper to enhance the wettability of the steel insert by Al. Then, the infiltration of these inserts was evaluated by gravity casting, centrifugal casting, and low‐pressure sand casting Funding information (LPSC). Evaluations involved microstructural characterization using optical micros- Fundação para a Ciência e a Tecnologia copy and SEM for interface analysis. It is revealed that centrifugal casting is highly reliable to infiltrate the inserts with Al, filling detailed cavities in depth. The copper coating aided in the creation of intimate interfaces. The infiltration at the insert's surfaces, curved‐like topography, is obtained through LPSC though it is affected by the direction in respect of material flow. KEYWORDS 3D‐printed inserts, aluminum, casting, infiltration I NTR O D U C TI O N rigid structure [1, 2], for example, aluminum (Al) foams obtained from a polyurethane sponge template, a plaster‐foam precursor, and then Aluminum metal matrix composites (Al‐MMCs) belong to the class of centrifugal casting [3]. However, the Al sponge can be obtained by advanced materials that have been continually researched and the counter‐gravity infiltration of a packed bed of expanded perlite developed to meet the needs of technological advancements for particles [4]. Singh et al. [5] performed the infiltration of aluminum by higher mechanical performance, considered lightweight, and cost‐ investment casting methodology into a mold loaded with Al‐Al2O3 effective. These composites are categorized depending on the particles, producing a function‐graded material for mechanical reinforcement type and shape, such as particles, whiskers, and fiber resistance applications. reinforced. Different methods can be used to produce these Shaga et al. [6] developed an Al‐MMC reinforced with a porous composites, varying from solid‐state to liquid‐metal‐infiltration laminated SiC scaffold, the Al matrix (Al—12 wt% Si—10 wt% Mg processes. The latter methodology involves the molten metal/alloy alloy) was infiltrated in the SiC structure by a pressureless technique, penetrating a preform, either being a packed bed or a porous and that is, placing an aluminum block on the top of the SiC scaffold in an Appl. Res. 2022;e202200062. https://doi.org/10.1002/appl.202200062 wileyonlinelibrary.com/journal/appl © 2022 Wiley‐VCH GmbH. | 1 of 9 2 of 9 | APPLIED RESEARCH alumina crucible, applying a vacuum of ~100 Pa followed by binder material, followed debinding to remove the binder and then introducing a high‐purity N2 gas flux at 950°C for 2 h. The authors sintering for densification [13–15]. attributed the success of infiltration to the presence of N2. These The casting process can be performed through different authors mentioned that the fracture under the tensile test occurred methodologies such as gravity or under external forces applied by in the Al matrix and not at the interface of Al–SiC. Another study centrifuge or high/low pressure. Low‐pressure sand casting (LPSC) presented the production of Al‐Ti bimetallic composite by casting in a seems to be a promising process since it can ensure a good filling and vacuum and squeeze infiltration to fill Ti preform. The first process decrease defects. The application of gas pressure to the melt surface included casting the A356 alloy on a Ti preform through a vacuum gently forces the molten material, such as Al, through a riser tube into chamber at a pressure of 10−5 Pa. Regarding squeeze casting, the Ti the mold cavity placed above the furnace [16, 17]. Choi et al. [18] preform and aluminum strips were stacked heating the material to employed low‐pressure casting to infiltrate pure Al into a preform of 730°C under 4.5 Pa pressure; thus, the molten Al seeped into the carbon–alumina short fibers. The authors used a melting temperature gaps between the Ti strips [7]. Concerning the success of infiltration, of 800°C and a pressure of 0.4 MPa to produce a highly dense some authors mentioned that the chemical composition of the Al cast composite with a 73% higher hardness (HV) than the Al matrix which is important, for example, Aghajanian et al. [8] demonstrated that Mg was ~19 HV. concentration should be between 0.5 and 1 wt.% for successful This study aims to primarily evaluate the Al infiltration behavior infiltration. Other elements, such as Si, can also improve infiltration. into steel/copper AM structures by applying three different casting This effect can be attributed to the reduction of viscosity by the methods. Then, this study proceeded with the interface characteri- addition of Mg or Si. zation, evaluating the influence of copper coating applied on steel Recent approaches involve additive manufacturing (AM) and inserts. Thus, it is intended to establish parameter and process casting, that is, first creating a complex lattice structure through limitations to develop robust Al structural composites in future AM process and then casting another material on the printed studies. structure. The steel‐Al hybrid structure has been experienced through the infiltration of molten Al into the steel insert [9]. This MATER IA LS AND METHO DS combination may cause phase evolution at the insert‐cast interface or nonintimate interfaces. If the latter condition occurs, Metal insert production the load transfer strengthening mechanism cannot be effective, lowering the mechanical response of the composite [10]. Regarding the formation of an intimate interface, in the case of The inserts used in this study were produced through the FFF dissimilar materials, such as Al and Fe, the formation of process. This AM methodology involves the design of the geometry intermetallic phases can occur at the interface (e.g., η‐Al 5Fe). of interest by CAD software, then it is exported as an STL file The chemical composition, the morphology (whether it is smooth transmitted to the printing machine, which in this study it is a or rough), and the thickness of the intermetallic layer effects the Markforged Metal X equipment. The chemical composition of inserts bonding strength [11, 12]. is represented in Table 1. Metallic inserts can be fabricated through AM processes since it Several 3D shapes were printed, as shown in Figure 1, to facilitates the production of complex and interconnected structures determine the optimal geometry for aluminum infiltration. The shapes such as lattices used for load‐transferring applications. Liquid/sinter‐ were developed considering the rules of design for AM, and based methodologies can be involved. The former ones can be laser recommendations from Markforged for the FFF, once 3D printing or electron beam techniques providing molten pool followed by is a geometry‐dependent process. Since FFF for metallic materials solidification whereas fused filament fabrication (FFF) process involves fabrication, debinding, and sintering, the rules mentioned involves densification of metal powders through sintering, that is, a before were used to optimize components to resist the different three‐dimensional (3D) structure is printed layer by layer using a production phases. They were also fabricated in the best orientation filament composed of a homogeneous mixture of powder and a to avoid the build of supports, considering overhangs reduction, by 17‐4 PH stainless steel Cr Ni Cu Si Mn Nb C P S Fe 15–17.5 3–5 3–5 <1 <1 0.14–0.45 <0.07 <0.04 <0.03 Bal. Copper Cu O Fe Other 99.8 <0.05 <0.05 Bal. Abbreviation: PH, precipitation‐hardening. T A B L E 1 Chemical composition (wt.%) of the stainless steel and copper inserts | APPLIED RESEARCH 3 of 9 F I G U R E 1 Three‐dimensional printed metal inserts were used in the current study with: (a) body‐centered cubic with Z strut (BCCZ) geometry, (b) pyramid‐like geometry; and (c) columnar geometry. maintaining the structural integrity at the end of the process. The the design of the cavity required for casting. The ceramic cast dried body‐centered cubic with Z strut (BCCZ) geometry, shown in and sintered at 1000°C. Then, the insert selected was placed in the Figure 1a, was chosen as the first insert geometry because it exhibits mold cavity, Figure 2b, the mold was covered with a lid of similar good stiffness [19]. The two other geometries, however, were ceramic material and sealed with ceramic glue. This mold assembly adapted from the BCCZ while considering the constraints of the mold was dried at 200°C/1 h to remove moisture. The centrifugal casting design and infiltration process. As a result, the pyramid‐like process was performed by a Neutrodyn Easyti centrifugal casting geometry, Figure 1b, presented a higher free interspace, and similarly, equipment loading the A356 charge in an appropriate crucible and the geometry presented in Figure 1c was selected for producing long the mold assembly, Figure 2c. The fusion temperature was selected inserts. One type of insert was selected for electrochemical coating at 810°C, the lower possible temperature by that equipment, and the with pure copper (Figure 1b) to assess the effect of this coating on casting process were carried out in a low vacuum atmosphere. The the wettability and interfaces. The electrodeposition bath consisted centrifugal process began at a medium speed once the Al melted and of copper sulfate (220 g/L CuSO4.5H2O), sulfuric acid (63 g/L then, the melt was projected into the ceramic mold. H2SO4), hydrochloric acid (50 mg/L HCl), Cuflex 500 base (0.8 ml/ L), and a brightness agent (1.6 ml/L). A current of 12 A, a cylindrical copper electrode, and a deposition duration of 15 min were also used. Low‐pressure casting Figure 3a shows the LPSC configuration used in this study. The Al Aluminum infiltration alloy charge and the graphite crucible were dried at 200°C/1 h and then placed in a resistance furnace heated up to 720°C for melt Gravity casting preparation. Heating was performed under argon (Ar) gas atmosphere The Al alloy A356 was melted by a goldsmith furnace, in a graphite sand mold was prepared to place the inserts in the cavity of the mold crucible, at 700°C and then poured directly on the top of the insert (Figure 3b) and close the mold. to decrease melt oxidation and hydrogen pick‐up. Meanwhile, the placed in another crucible. Some inserts were preheated to 700°C to The slag was removed once the material had completely melted, assess the impact of the temperature on infiltration behavior. and the feeder tube and sand mold were installed on the top of the Besides, other experiments were performed by placing a piece of furnace. The casting process began with the molten metal rising the Al alloy on the top of the insert and heating it in a vacuum through the tube into the mold cavity applying an approximately atmosphere, using a furnace with a horizontal alumina tube, to 125 mbar Ar gas pressure on the liquid metal surface. evaluate whether the absence of air‐oxygen can increase infiltration. Three sets of processing conditions were used: low vacuum at 620°C/9 Pa/5 min and 700°C/9 Pa/5 min, and high vacuum at 700°C/6 × 10 −4 Interface characterization Pa/5 min. Cross sections were cut from casts followed by performing conventional metallographic preparation techniques, including the Centrifugal casting steps of grinding up to 1000 mesh and polishing with 6 and 1 µm diamond paste. Microscopic characterization for interfaces was This process requires a ceramic mold with a cavity, Figure 2a, carried out using a Digital Microscope, Leica, DVM6 A, and a produced by pouring the ceramic paste into a silicon mold to replicate scanning electron microscope (SEM), FEI Quanta 400 FEG (ESEM) 4 of 9 | APPLIED RESEARCH (a) F I G U R E 2 (a) Plaster mold scheme; (b) ceramic mold with the insert placed; and (c) centrifugal casting setup. (b) (c) (a) F I G U R E 3 (a) Low‐pressure sand casting process scheme and (b) sand mold with inserts placed in the cavities. (b) equipment. The chemical composition distribution and localized had significant porosity caused by inadequate Al penetration; the analysis were determined by backscattered electron (BSE) mode interfaces between the cast Al and the insert were not intimate, and energy‐dispersive X‐ray spectroscopy (EDS), and EDAX Genesis which was attributed to the Al shrinkage during solidification not X4M (Oxford Instrument). allowing metallurgical bonding at the interface, as can be seen in Figure 4. Thus, the study proceeded by performing a set of casting to R E S ULT S AN D D I SCU SS I O N determine the critical diameter for the Al infiltration in similar steel. Three steel cups with holes, in the base part, ranging in diameters Gravity casting from 1 to 4 mm were produced by machining. The casting was performed on the cups at room temperature, preheated to almost In the current study, the efficiency of gravity casting for infiltration 350°C and 700°C though the real temperatures were 190°C and was evaluated by applying different strategies such as casting on an 370°C, respectively, in the initial moment of casting. This was caused insert at room temperature and preheating to 350°C and 700°C. by the insert cooling after it was taken out of the electric furnace and Moreover, the inserts were positioned in different orientations to before the molten aluminum was poured. As seen in Figure 5, the determine the infiltration optimal conditions. However, these best scenario was obtained in the insert at room temperature, experiments revealed that the gravity casting technology had showing that the critical diameter is roughly 2 mm since not all the numerous restrictions for Al infiltration. The produced composites 1.5 mm holes were infiltrated. The worst‐case scenario was obtained | APPLIED RESEARCH 5 of 9 by the cup preheating at 700°C. This change can be attributed to the observations, Figure 7b, confirmed metallurgical bonding at the expansion of steel. interface, and the formation of intermetallic phases as seen in Figure 7c. The predominant intermetallic compounds generated are γ‐Al4Cu9 and θ‐Al2Cu, according to EDS analysis (Table 2). However, Centrifugal casting the Al melt did not fill all the topography details of the Cu insert produced by FFF. The cross‐section observation reveals that the infiltration was In the current study, infiltration was evaluated for another complete; however, microscopic observations on the polished geometry, Figure 1c. The centrifugal casting was performed surfaces, Figure 6c, revealed that interfaces were not completely successfully for that length. Cross sections were cut along the height intimate being influenced by the curved‐like geometries. The micro of the cast, and some differences were observed specifically in the porosities observed in the cross‐section of these inserts are caused types of interfaces. Figure 8b,c represents, respectively, sections I by inefficient sintering during the FFF process. and II indicated in Figure 8a, which correspond to zones filled in the Since the copper is well‐known for having good wettability with earliest (section I) and latest (section II) steps of casting. These Al, the infiltration by centrifugal casting proceeded in a copper insert. microscopic observations reveal that cross sections closer to the The polished cross‐section of the Al–Cu structure composite, insert base (such as section I) have the majority of intimate interfaces Figure 7a, shows that infiltration is good and the SEM/BSE (Figure 8d) though some nonintimate interfaces (Figure 8f) are observed. The presence of intimate interfaces decreases, Figure 8c,e,g, by reaching the top region of the cast (such as section II). This difference can be attributed to the temperature reduction of the last molten Al filled the top region. Moreover, the influence of increased centrifugal force on infiltration in the earliest steps of casting, due to the higher mass of the molten material, can be effective. Low‐pressure casting Regarding infiltration through LPSC, this methodology is highly influenced by the size of internal cavities. The Al did not FIGURE 4 Gravity infiltration composite with significant voids. penetrate in small interspaces of the insert with complex F I G U R E 5 Tests of the influence of insert preheating temperature on infiltration: (a) insert at room temperature; (b) insert preheated to ~350°C; and (c) insert preheated to ~700°C. 6 of 9 | APPLIED RESEARCH F I G U R E 6 (a) Composites resulting from centrifugal infiltration; (b) cross‐section of the insert with complete infiltration; and (c) the same section in (b) but at higher magnification. F I G U R E 7 (a) Al–Cu structure cross‐section; (b) scanning electron microscope image of the same cross‐section; and (c) a higher magnification of one region. Elemental composition of the zones identified in TABLE 2 Figure 7c Zone Al (at.%) Si (at.%) Cu (at.%) Possible phase Z1 12 – 88 (Cu) Z2 30 1 69 γ Z3 64 2 34 θ Z4 93 2 5 (Al) Z5 7 90 3 (Si) Al as well. In the meantime, the effect of the position of the inserts in the mold cavity was evaluated as shown in Figure 3b. Placement number 2 presented the largest penetration as seen in Figure 10a. This filling behavior might be related to changes in velocity and temperature during cavity filling. Nevertheless, the interface of the insert‐Al is not continuously intimate. As seen in Figure 10b,c, the presence of an intimate interface seems to vary from side to side. This change can be attributed to the direction of melt flow during filling. Chelladurai et al. [20] applied Cu coating to create a barrier between the steel and the molten Al to prevent any formation of iron aluminides, in the meantime, geometry, Figure 9a. This penetration partially increased in the creating a better interface bonding achieved through squeeze insert embedding a larger interspacing; however, it was not casting. completed. Regarding the effect of Cu coating at the Al–steel interface, This study proceeded by applying a Cu coating on the microscopic observations and chemical composition profiles are produced inserts to enhance the wettability and penetration of illustrated in Figure 11. As seen, the Cu coating is still detected at the | APPLIED RESEARCH 7 of 9 F I G U R E 8 (a) The printed insert designed for tensile test specimens; Al–steel cast produced by centrifugal casting: (b, c) different sections observed along the height; (d, e) intimate interfaces; and (f, g) nonintimate interfaces. interface whereas this coating seems to be diffused slightly in the steel and dissolved in the adjacent Al matrix. Intermetallic compounds were not found in any of the types of interfaces, suggesting that the copper coating may prevent the formation of these compounds though some micro cracks are observed. CONCLUSION This study covered the infiltration of Al AA356 alloy into steel inserts produced by the FFF technique as an AM approach. Three casting techniques were applied: gravity, centrifugal, and LPSC. The following conclusions can be drawn: F I G U R E 9 Cross section of the inserts placed in the mold cavity for infiltration by low‐pressure sand casting. • Gravity casting is not reliable for infiltration of semi to complex geometries. 8 of 9 | APPLIED RESEARCH F I G U R E 10 Cross sections of the composite structure of the copper coated steel insert filled with Al by low‐pressure sand casting: (a) as cut surface; (b, c) as polished surfaces observed by optical microscopy. F I G U R E 11 Interfaces of the insert 2 after casting: optical microscope images from the (a) steel–Cu–Al interface; (b) steel–Al interface; and (c, d) energy‐dispersive X‐ray spectroscopy analysis profiles of the corresponding interfaces. • Centrifuge casting can be a promising method to infiltrate Al melt into small details and complex geometries; however, the production of intimate interfaces varies along the insert height. • LPSC is limited for infiltration of Al melt into complex geometries. • The copper coating can enhance the formation of locally intimate interfaces without intermetallics formation. • The topography of 3D‐printed inserts produced by FFF affects the infiltration at the surface details. | APPLIED RESEARCH • The geometry of the inserts greatly influences the infiltration [4] behavior. • The application of Cu inserts can cause the formation of copper aluminides at the interface. To achieve the main objectives of producing aluminum structural components locally reinforced with a steel insert the authors proposes future works to accomplish, including: [5] [6] [7] [8] • Optimization of the mold design or using a close‐up sand mold to avoid heat and pressure losses during the LPSC process. [9] • Production of composites through centrifugal casting suitable for machining tensile test specimens to evaluate the mechanical [10] properties. [11] A C KN O W L E D G M E N T S The authors would like to acknowledge AAPICO S.A., and Fundação [12] para a Ciência e a Tecnologia (FCT), I.P., through UIDB/50022/2020 and UIDP/50022/2020 project. They would also like to acknowledge the Centro de Materiais da Universidade do Porto for microscopy assistance. This study was funded by the Project PAC (PO‐CI‐01‐ [13] [14] 0247‐FEDER‐046095), cofinanced by COMPETE 2020 through FEDER and by National Funds through FCT. [15] CO NFL I CT OF INTERES T [16] The authors declare no conflict of interest. D A TA A V A I L A B I L I T Y S T A T E M E N T Data sharing is not applicable to this article as no new data were [17] [18] created or analyzed in this study. [19] ORCID Helder Nunes http://orcid.org/0000-0002-8110-2572 [20] 9 of 9 M. Taherishargh, I. V. Belova, G. E. Murch, T. Fiedler, Mater. Sci. Eng.: A 2014, 604, 127. https://doi.org/10.1016/j.msea.2014. 03.003 S. Singh, R. Kumar, R. Kumar, J. S. Chohan, N. Ranjan, Proc. Inst. Mech. Eng., Part L: J. Mater.: Des. Appl. 2021, 23, 674. https://doi. org/10.1177/14644207211054143 A. Shaga, P. Shen, C. Sun, Q. Jiang, Materi. Sci. Eng.: A 2015, 630, 78. https://doi.org/10.1016/j.msea.2015.02.012 R. Gecu, A. Karaaslan, Int. J. Metalcasting 2019, 13, 311. https://doi. org/10.1007/s40962-018-0253-0 M. K. Aghajanian, M. A. Rocazella, J. T. Burke, S. D. Keck, J. Mater. Sci. 1991, 26, 447. https://doi.org/10.1007/BF00576541 M. Ghasri‐Khouzani, X. Li, A. A. Bogno, Z. Chen, J. Liu, H. Henein, A. J. Qureshi, J. Manuf. Process. 2021, 69, 320. https://doi.org/10. 1016/j.jmapro.2021.07.047 S. Senge, J. Brachmann, G. Hirt, A. Bührig‐Polaczek, AIP Conf. Proc. 2018, 1960, 040019. https://doi.org/10.1063/1.5034873 S.‐L. Shang, H. Sun, B. Pan, Y. Wang, A. M. Krajewski, M. Banu, J. Li, Z. K. Liu, Sci. Rep. 2021, 11, 24251. https://doi.org/10.1038/ s41598-021-03578-0 X. Yu, J. Huang, T. Yang, D. Fan, Materials 2022, 15, 3563. https:// doi.org/10.3390/ma15103563 B. Dutta, S. Babu, B. H. Jared, Science, Technology and Applications of Metals in Additive Manufacturing, Elsevier Science, USA 2019. G. Poszvek, G. Stattler, E. Markl, R. Seemann, M. Lackner, Digital Conversion on the Way to Industry 4.0 (Eds: N. M. Durakbasa, M. G. Gençyılmaz), Springer International Publishing, USA 2021, p. 140. T. Kurose, Y. Abe, M. V. A. Santos, Y. Kanaya, A. Ishigami, S. Tanaka, H. Ito, Materials 2020, 13, 2493. https://doi.org/10.3390/ ma13112493 G. T. Kridli, P. A. Friedman, J. M. Boileau, Materials, Design and Manufacturing for Lightweight Vehicles, 2nd ed. (Ed: P. K. Mallick), Woodhead Publishing, USA 2021, p. 267. A. I. H. Committee, ASM Handbook: Casting, 10 ed. Vol. 15, ASM International, USA 2008. Y. Choi, Y. Kataoka, Z. Zu, K. Matsugi, G. Sasaki, Mater. Trans. 2017, 58, 1097. https://doi.org/10.2320/matertrans.M2017075 A. Seharing, A. H. Azman, S. Abdullah, Adv. Mech. Eng. 2020, 12, 1687814020916951. https://doi.org/10.1177/1687814020916951 S. J. S. Chelladurai, R. Arthanari, K. Krishnamoorthy, K. S. Selvaraj, P. Govindan, Trans. Indian Inst. Met. 2018, 71, 1041. https://doi. org/10.1007/s12666-017-1235-2 REFERENCES [1] [2] [3] F. Nturanabo, L. Masu, J. Kirabira, Aluminium Alloys and Composites (Ed: K. Cooke), IntechOpen, London 2019. R. Etemadi, B. Wang, K. M. Pillai, B. Niroumand, E. Omrani, P. Rohatgi, Mater. Manuf. Processes 2018, 33, 1261. https://doi.org/ 10.1080/10426914.2017.1328122 K. Chang, J.‐T. Gao, Z. Wang, Z. C. Guo, J. Mater. Process. Technol. 2018, 252, 705. https://doi.org/10.1016/j.jmatprotec.2017.10.032 How to cite this article: H. Nunes, O. Emadinia, J. Costa, R. Madureira, R. Soares, J. Silva, I. Frada, V. Anjos, F. Viana, M. Vieira, A. Reis, Appl. Res. 2022, e202200062. https://doi.org/10.1002/appl.202200062