Modification of as-cast Al-Mg/B4C composite by addition of Zr

MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Modification of As-Cast Al-Mg/B 4 C Composite by Addition of Zr M. Rezayat a,b,* , M.R. Bahremand a , M.H. Parsa a , H. Mirzadeh a , J.M. Cabrera b a School of Metallurgy and Materials Engineering, College of Engineering, University of Tehran, P.O. Box 11155-4563, Tehran, Iran b Departamento de Ciencia de los Materiales e Ingeniería Metalúrgica, ETSEIB, Universitat Politècnica de Catalunya, Av. Diagonal 647, 08028, Barcelona, Spain * Corresponding author e-mail address: [email protected] Abstract Zirconium was used in Al-Mg/B 4 C composite to improve compocasting efficiency by increasing particle incorporation. Scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS) results revealed that by addition of zirconium a reaction layer containing Zr, Al, B and C is formed on the interface of B 4 C-matrix. X-ray diffraction (XRD) analysis of extracted particles unveiled that the ZrB 2 phase is the main constituent of this layer. Formation of ZrB 2 is an exothermic reaction which can rise temperature locally around particles and agglomerates. Rising temperature around agglomerates in conjunction with turbulent flow of melt facilitates agglomerates wetting and dissolving into molten aluminum. As the result, final product contains more uniformly distributed B 4 C particles. Besides enhancing compocasting efficiency, addition of Zr and formation of reaction layer by improving particle matrix bonding quality, led to increase in ultimate tensile strength and elongation of the composite around 8% and 30%, respectively. SEM observations of the fracture surfaces confirmed that a proper bonding presents at the interface of particles and matrix in presence of Zr. Keywords: Metal-matrix composites; Liquid-solid reactions; Mechanical properties; Scanning electron microscopy; X-ray diffraction. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT 1. Introduction The particle reinforced aluminum matrix composites (AMC) are widely used in transportation-related industries because of their desirable properties including high value of specific strength, improved elastic behavior, and high wear resistance [1-3]. Generally, the fabrication process of AMCs can be divided to casting and powder metallurgy routes [1]. Among various casting methods, stir casting has been known as an inexpensive process that is applicable for a wide range of materials [1, 2]. Boron carbide (B 4 C) is known as an attractive reinforcement for AMCs due to its lower density compared with aluminum, high hardness, and low thermal expansion in comparison with other common reinforcements such as SiC and Al 2 O 3 [4]. However, poor wettability of B 4 C with molten aluminum at temperatures below 1100 °C is a drawback of liquid state fabrication methods [5, 6]. Several researches have been conducted so far to improve the interaction and incorporation of B 4 C in molten Al, including preheating particles to form boron oxide, surface preparation techniques, coating with Ti, and applying Ti-containing flux [6-11]. As a result of the interfacial reaction between the reinforcement and matrix, an interface forms, which plays an important role in mechanical properties of composites [1214], and hence, lots of research has been conducted on this subject [2,3,12,13]. Zirconium has been widely used in aluminum alloys as a grain refiner in casting or microstructure stabilizer at high temperature [15]. It has been also proved that according to Eq. 1, Zr can react with boron carbide and produces Zirconium Carbide and Zirconium Boride [16], even when Zr is in the Al matrix as an alloying element [17]. 24 23 ZrBZrCCBZr + → + (1) Based on the facts that this reaction is exothermic [18] and also interaction in metal-ceramic system is rarely desirable [19], it seems that addition of Zr can improve the particle incorporation and mechanical properties. However, there is not an available comprehensive MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT study on the influence of Zr on the B 4 C incorporation in the Al matrix. Therefore, the aim of current work is to investigate the effect of Zr as an alloying element on the boron carbide incorporation into molten aluminum, particle-matrix interface, and mechanical properties of the compocast product. The stirring time and the casting temperature were considered as influential processing parameters and their impacts on the microstructure were investigated. Moreover, based on the Response Surface Methodology (RMS) [20], the mentioned parameters were optimized to reach the highest B 4 C yield, which has gained a great technological importance. 2. Experimental details 2.1. Stir-casting B 4 C particles with the average particle size of 25 and 80 µm were used as the reinforcement material in commercially pure aluminum (99.81 %) matrix. In order to examine the effect of Zr on the wettability, three different ratios of (wt.% Zr)/(wt.% B 4 C), namely 0, 0.05 and 0.1, were used by adding Al-15Zr master alloy into the molten aluminum. Particulate reinforced AMCs were produced by using 0.5 kg aluminum and adequate amount of master alloy which were melted and were held for 40 minutes in an alumina crucible. B 4 C particles were added into the melt via vortex formed from string with 570 rpm. Stirring was applied for 5, 12.5 and 20 minutes at the stirring temperatures of 750, 825 and 900 °C under Argon atmosphere. Just before pouring into a cold cylindrical steel mold, 3 wt. % Mg was added to the melt, and there is not enough time for take parting Mg in reactions, or in other words, there is not so much probability for an excessive reaction of Mg with B 4 C [7,21]. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT 2.2. Design of experiment for investigating particle incorporation Finding the effects of Zr and the processing related parameters such as stirring time and temperature, calls upon large numbers of experiments. In such experiments, one parameter is varying while the other factors are constant. A major demerit of this method is ignoring the interactions between parameters. Response surface methodology, which is based on the combination of mathematical and statistical techniques for designing experiments, provides a realistic optimization process by considering the effect of variable interactions. Reduction of the number of the experiments is the main merit of the RSM. In the current research, the central composite design (CCD) based on the Design-Expert® software (Ver. 8.0.1) [22] was utilized. CCD is an efficient way to provide a sufficient amount of information to test the fitness of a model. Considering three parameters (temperature, stirring time and amount of Zr) in three levels, the number of required experiments was obtained as 11 which are presented in Table 1. Table 1: Casting conditions and obtained results for different samples. Sample Time (min) Temperature (°C) D×W Zr /W B4C B 4 C incorporation % 1 5 750 0 16.41 2 12.5 750 1.25 26.12 3 20 750 2.5 32.87 4 5 825 1.25 17.13 5 12.5 825 0 8.06 6 12.5 825 1.25 18.08 7 12.5 825 2.5 20.34 8 20 825 1.25 24.00 9 5 900 2.5 5.29 10 12.5 900 1.25 7.00 11 20 900 0 2.00 To find the effect of parameters on particle incorporation (B 4 C yield%), the image analysis technique was used. Some corresponding micrographs of the composites, from which the data were obtained, are presented in Fig. 1 as representative examples. For microstructural MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT observations and energy dispersive X-ray spectroscopy (EDS), a Thermal-Field-Emission Scanning Electron Microscope (Jeol 7001f -0.1-30 kV) was employed. Fig. 1: Optical microghaphs of Al/B 4 C composites for (a) Sample 1, (b) Sample 6 and (c) Sample 10. After resolving the aluminum matrix in the NaOH solution, the X-Ray Diffraction (XRD) analysis was used to identify the phases present in the particles. The XRD experiment was performed by a Rigaku Ultima IV using Cu-Ka radiation. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT It was proved by Kennedt [5] that the weight fraction of alloying elements (here Zr) required for interfacial reaction depends on the weight fraction of the reinforcement (e.g. CBZr WW 4 / ) and the particle diameter (D). The critical mass of alloying element needed for wetting particles ( Zr m ) required to form a reaction layer is related to the surface of all of the particles (S), i.e. cSm Zr = [6]. The total surface of particles can be calculated by using Eq. 2; 2 DnS π = , (2) where n and D are the number and average diameter of particles, respectively. In the case of spherical particles, n can be evaluated by dividing total volume of particles to volume of one particle (Eq. 3). 6 3 44 D m n CBCB π ρ = (3) By combining abovementioned equations, the ratio of masses required for wetting can be expressed as: CBCB Zr c m m D 44 6 ρ = (4) Therefore, CBZr mDm 4 or CBZr WDW 4 are independent of the particle size. In the other words, the effects of Zr, B 4 C, and D were considered in a unique parameter presented as CBZr WDW 4 / . Finally, the response (output of the model) was the B 4 C yield percentage, i.e. fraction of particles incorporation into the molten Al. 2.3. Investigating the impact of interface on mechanical properties Chemical interactions between the matrix and the reinforcement determine the interface adhesion, modify the characteristics of the composite components and affect its mechanical characteristics [13]. In other words, the interface quality has an influential impact on the load MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT transfer from matrix to particles. Therefore, the higher the strength of the interface, the higher the load that the composite can withstand. Moreover, it is expected that increasing the quality of interface results in enhanced tensile elongation. Indeed, a weak interface is a suitable site for formation of cavities because of low strength of this area in comparison with either of particle or matrix [14]. In order to investigate the impact of Zr-containing layer on the mechanical properties, two composites, one with addition of Zr (COMP Zr ) and another without Zr (COMP), were fabricated. The casting condition was optimized to produce a Al-3 wt.% Mg - 10 vol% B 4 C composite (exact volume fraction of B 4 C was measured from microstructure) and then the product was remelted to fabricate COMP Zr and COMP by adding Zr. The details are presented in Table 2. The corresponding microstructures of COMP Zr and COMP are shown in Fig. 2, which confirms that the particle size and volume fraction are the same for both composites. After homogenization at 580 °C for 2 hours and hot extrusion at 470 °C with the ratio of 14:1, samples were prepared via machining according to ASTM E8 standard (sub size). Universal tensile test machine (SANTAM 2000) with crosshead speed of 1 mm/min was used for performing three tensile tests for each condition. Table 2: details of fabrication process for COMP Zr and COMP. Stirring temperature Stirring time Stirring temperature after remelting for adding Zr Stirring time after remelting for adding Zr Amount of added Zr COMP zr 720 ºC 10 min 800 ºC 20 min 0.2 wt%( D×W Zr /W B4C =1.6 ) COMP 800 ºC 20 min 0 MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Fig. 2: Optical micrographs of (a) COMP Zr and (b) COMP. 3. Results and discussion 3.1. Particle incorporation To evaluate the effect of Zr, casting temperature, and stirring time variables on the percentage of B 4 C yield in the product, experiments were performed on the basis of central composite design and the results are presented in Table 1. After the analyzing of the experimental results, Eq. 5 was obtained from Design-Expert software based on the response of different input variables: ( ) 2 24 44 100039.1/565.01526.0696.51 TWDWTionIncorporatCB CBZr − ×−++−= (5) While this model is based on regression analysis, some useful remarks can be deduced from it. For example, the model reveals that, in comparison with temperature and amount of Zr, the stirring time has negligible effect on the B 4 C yield. Interestingly, the model also predicts no interaction between parameters. Based on Eq. 5, it can be seen that the effect of Zr is MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT significant. Fig. 3 shows the effect of Zr and stirring temperature on the particle’s incorporation (yield) as predicted by Eq. 5. It can be concluded form this figure that Zr plays an influential role in wettability of the boron carbide particles and enhancement of particle’s incorporation at a given stirring temperature. Fig. 3: Variation of particle incorporation versus stirring temperature and Zr based on Eq. 5. 3.2. The interaction layer Fig. 4 shows a typical particle from sample 6 (stirring temperature of 820 ºC, stirring time of 12.5 min and CBZr WDW 4 / of 1.25) with a fine white layer at the interface of particle-matrix. EDS map analysis from this area represents a layer which contains Zr surrounding particle. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Table 3: Average mechanical properties of COMP and COMP Zr Offset Yield Stress (MPa) Ultimate Tensile Strength (MPa) Elongation ( % ) COMP 123 ± 6 176 ± 2 4.1±0.6 COMP Zr 122 ± 5 192 ± 3 5.3±0.4 To investigate the effect of Zr-containing layer on the properties of the particle-matrix interface, the fracture surfaces were taken into account. Fig. 10 illustrates the fracture surface of composites subjected to tensile test. According to this figure, fracture proceeded in both specimens by ductile rupture mechanism in the matrix. Dimples are the result of microvoid coalescence [26]. The fracture surface of COMP shows some dimples containing nonfractured particles and the separation have occurred from the particle-matrix interfaces (Fig. 10a). However, Fig. 10b shows that the numbers of fractured particle are greater in COMP Zr , which suggests that the particle-matrix interfaces have adequate strength. These results confirm that the addition of Zr can increase the quality of the interface. Fig. 10: Fracture surface of a) COMP and b, c) COMP Zr . 4. Conclusion Zirconium as an alloying element was introduced to aluminum in order to improve the incorporation of boron carbide particles during stir casting. The influential parameters of MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT compo-casting process including stirring temperature, stirring time and also the amount of Zr were optimized by applying the response surface methodology. The results and corresponding analyses showed that the addition of Zr enhances the particle yield during fabrication process. This fact was attributed to the formation of a reaction layer at the interface of matrix and B 4 C. EDS and XRD analyses revealed that this layer is mainly formed by ZrB 2 . Formation of ZrB 2 by increasing the temperature around agglomerates during stirring, leads to dissolving them into the molten aluminum. It was found that although stirring at elevated temperature promotes the reaction and leads to form a thicker interface, the intensified oxidation of molten Al has an adverse effect on the casting efficiency. Moreover, addition of Zr led to 8% increase in the ultimate tensile strength and around 30 % increase in elongation of the composite. SEM observations of the fracture surfaces confirmed that a proper bonding presents at the interface of particles and matrix of composite containing Zr. References [1] N. Chawla, K.K. Chawala, Metal Matrix Composites. New York, Springer. (2013). [2] D. Mandal, S. Viswanathan, Effect of re-melting on particle distribution and interface formation in SiC reinforced 2224Al matrix composite, Mater Charact. 86 (2013) 22-27. [3] Y.Z. Li, Q.Z. Wang, W.G. Wang, B.L. Xiao, Z.Y. Ma, Interfacial reaction mechanism between matrix and reinforcement in B 4 C/6061Al composites, Mater. Chemist Physic. 115 (2015) 107-117. [4] B. Manjunatha, HB. Niranjan, KG.Satyanarayana, Effect of mechanical and thermal loading on boron carbide particles reinforced Al-6061 alloy, Mater Sci Eng A. 632 (2015) 47155. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT [5] AR. Kennedy, B.Brampton, The reactive wetting and incorporation of B4C particles into molten aluminium, Scripta Materiala. 44 (2001) 1077-1082. [6] Ibrahim MF, Ammar HR, Samuel AM, Soliman MS, Samuel FH, On the impact toughness of Al-15 vol.% B 4 C metal matrix composites, Comp Part B 2015; 79; 83-94. [7] Zhang Z, Fortin K, Charette A, Chen XG, Fluidity and Microstructure Evolution of Al-12 % B 4 C Composites Containing Magnesium, Mat Trans 2011; 52; 928-933. [8] B. Zou, P. Shen, Q. Jiang, Reaction synthesis of TiC – TiB 2 / Al composites from an Al– Ti–B 4 C system. J Mater Sci., 1 (2007) 9927-9933. [9] Z. Zhang, XG. Chen, K. Fortin, A. Charette, Effect of titanium on microstructure and fluidity of Al–B 4 C composites, J Mater Sci. (2011) 3176-3185. [10] A. Canakci, F. Arslan, I. Yasar, Pre-treatment process of B 4 C particles to improve incorporation into molten AA2014 alloy, J Mater Sci. 42 (2007) 9536-9542. [11] I. Kerti, F. Toptan, Microstructural variations in cast B 4 C-reinforced aluminium matrix composites (AMCs), Mater Letters. 62 (2008) 1225 - 1228. [12] QQ. Zhang, GQ. Wu , Z. Huang, Y. Tao, Effects of particle/matrix interfaces on the mechanical properties for SiCp or YAl2p reinforced Mg–Li, J Alloy Compounds. 588 (2014) 1-6. [13] J. Yin, D. Yao, Y. Xia, K. Zuo, Y. Zeng,, The effect of modified interfaces on the mechanical property of β-silicon nitride whiskers reinforced Cu matrix composites, J Alloy Compounds. 615 (2014) 983–988. [14] L. Ceschini, G. Minak, A. Morri, Tensile and fatigue properties of the AA6061/20 vol.% Al 2 O 3 p and AA7005 /10vol.% Al 2 O 3 p composites, Compo Sci Tech., 66 (2006) 333-342. [15] F. Wang, D. Qiu, ZL. Liu, JA. Taylor, MA. Easton, MX. Zhang, The grain refinement mechanism of cast aluminium by zirconium, Acta Materialia. 61 (2013) 5636–5645. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT [16] WB. Johnson, AS. Nagelberg, E. Breval, Kinetics of Formation of a Platelet-Reinforced Ceramic Composite Prepared by the Directed Reaction of Zirconium with Boron Carbide, J American Ceramic Society. 74 (1991) 2093-2201. [17] J. Lai, Z Zhang, XG.Chen, Effect of Sc, Zr, and Ti on the interfacial reactions of the B 4 C/Al system, J Mater Sci. (2011) 451-459. [18] M. Zhang, Y. Huo, Q. Hu, P. Zhang, B. Zou, Reaction behavior and formation mechanism of ZrC and ZrB 2 in the Cu– Zr– B4C system, Int J Ref Metal Hard Mater. 43 (2014) 102-108. [19] AR. Kennedy, AE. Karantzalis, The incorporation of ceramic particles in molten aluminium and the relationship to contact angle data, Mater Sci Eng A., 264 (1999) 122-129. [20] RL. Mason, RF. Gunst, JL. Hess, Statistical Design and Analysis of Experiments with Applications to Engineering and Science, John Wiley and Sons., USA, (2003). [21] K.B. Lee, H.S. Sim, S.Y. Cho, H. Kwon, Reaction products of Al-Mg-B4C composite fabricated by pressureless infiltration technique, Mater Sci Eng A. 302 (2001) 227–234. [22] Design-Expert® Software Version 9, www.statease.com. [23] E. Trujillo-Vázquez, M.I. Pech-Canul, J.C. Guía-Tello, M.A. Pech-Canul, Surface chemistry modification for elimination of hydrophilic Al 4 C 3 in B 4 C/Al composites, Mater Des. 89 (2016) 94-101. [24] PK. Ghosh and S. Ray, A model study on the particle dispersion and fluid particle interaction in slurry of liquid alloy and ceramic particle, Trans Japan Inst Metals. 29 (1999) 502-508. [25] O. Dezellus, B. Gardiola, J. Andrieux, On the Solubility of Group IV Elements (Ti, Zr, Hf) in Liquid Aluminum Below 800°C, J Phase Equilibria and Diffusion 35 (2014) 120-126. [26] G. Dieter, Mechanical Metallurgy. 3rd ed. McGraw-Hill; 1976. MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT [27] B. Ko, Y.Yoo ,Hot-deformation behaviour of AA2224 composites reinforced with both particles and whiskers of SiC, Comp Sci Tech. 3538 (1998) 479-485. . MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Figure captions Fig. 1: Optical microghaphs of Al/B 4 C composites for (a) Sample 1, (b) Sample 6 and (c) Sample 10. Fig. 2: Optical micrographs of (a) COMP Zr and (b) COMP. Fig. 3: Variation of particle incorporation versus stirring temperature and Zr based on Eq. 5. Fig. 4: EDS map for variation of Zr near the B 4 C-matrix interface for sample 6 (stirring temperature of 820 ºC, stirring time of 12.5 min and CBZr WDW 4 / of 1.25). Fig. 5: The morphology of interface layer including crystals in different areas of a particle. Fig.6: variation of elements through the interface of matrix –particle. Fig. 7: XRD analysis of bulk extracted B 4 C particles (sample #7stirring temperature of 820 ºC, stirring time of 12.5 min and CBZr WDW 4 / of 2.5). It should be noted that the main peaks of B 4 C and AlB 12 C 2 are near to each other. Fig. 8: Interface of B 4 C-matrix at different casting conditions, (a) sample#3 (stirring temperature of 750 ºC), (b) sample#10 (stirring temperature of 900 ºC). Fig. 9: Engineering stress-strain carves for COMP Zr and COMP. Fig. 10: Fracture surface of a) COMP and b, c) COMP Zr . MANUS CRIP T ACCEP TED ACCEPTED MANUSCRIPT Highlights • Using Zr in Al-Mg/B 4 C system improves the particles incorporation by promoting the particle matrix interfacial reaction. • Based on SEM, EDS and XRD, the fine reaction layer mainly consists of ZrB 2 . • The reaction layer strengthens the particle matrix interface which in turn, improves the mechanical properties and fracture behavior of product.