- Email: [email protected]
Characteristic behaviours of clad layer by a multi-layer laser cladding with powder mixture of Stellite-6 and tungsten carbide
Surface & Coatings Technology 201 (2006) 3385 – 3392 www.elsevier.com/locate/surfcoat
Characteristic behaviours of clad layer by a multi-layer laser cladding with powder mixture of Stellite-6 and tungsten carbide Guojian Xu a,⁎,1 , Munaharu Kutsuna a , Zhongjie Liu a , Liquan Sun b a
Graduate School of Engineering, Nagoya University, 1 Furo-cho Chikusa-ku, Nagoya 464-8603, Japan b Chang Chun Institute of Technology, China Received 11 May 2006; accepted in revised form 17 July 2006 Available online 1 September 2006
Abstract The powder mixture of Stellite-6 and tungsten carbide (WC) in the range of 0 wt.%–47 wt.% WC fraction with constant chemical composition multi-layer cladding (CCCMLC) and functionally gradient material multi-layer cladding (FGMMLC) were deposited on mild steel (JIS-SM400B) plates by a 2.4 kW cw CO2 laser. In the CCCMLC, the WC weight fractions at each layer are constant. In the FGMMLC, the WC weight fractions at each layer are varied by controlling the disk rotation speed of the feeder. The phase constitution, microstructure, hardness and wear resistance of the clad layer were investigated by an X-ray diffractometer, energy dispersion spectroscopy (EDS), scanning electron microscope (SEM), laser microscope, Vickers hardness tester and wear tester. According to the analyzed results, the microstructure of the clad layer consists of hypoeutectic structure, and undissolved tungsten carbides dispersed in the matrix of the Co-based alloy. The Vickers hardness increases with the increase of WC weight fraction. The clad layers of CCCMLC and FGMMLC have almost same wear resistance. On the other hand, to decrease the crack sensitivity of clad layer, the influence of a preheating temperature and WC weight fraction in the CCCMLC and FGMMLC on the crack sensitivity was investigated comparatively. As the result, the cracks of clad layer clad with the powder mixture of Stellite-6 and WC belonged to a quasicleavage fracture type crack. In the FGMMLC method with gradual increase of WC particles clad at three subsequent layers shifts the critical amount of WC at which cracking appears to the higher level, or that it reduces the required preheating temperature. Namely, the crack sensitivity of clad layer using FGMMLC method is lower than that of CCCMLC method. © 2006 Elsevier B.V. All rights reserved. Keywords: Laser cladding; Quasi-cleavage fracture; Crack sensitivity; Constant chemical composition multi-layer cladding (CCCMLC); Functionally gradient material multi-layer cladding (FGMMLC); Hypoeutectic structure
1. Introduction Advantages of laser cladding are to produce a clad layer of composite materials, reduced dilution and deformation and excellent wear resistance by a fast cladding process. Dilution can be controlled by varying defocusing distance, travel speed and powder feeding rate [1]. Therefore in recent years, there has been rapid development of the laser surface treatment technique ⁎ Corresponding author. E-mail addresses: [email protected] (G. Xu), [email protected] (M. Kutsuna), [email protected] (Z. Liu), [email protected] (L. Sun). 1 Present: Technical Development, LASER X CO., LTD. 7 Kofukada Shinbayashi-cho, Chiryu-shi Aichi 472-0017 JAPAN. 0257-8972/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2006.07.210
in the production domain [2]. It has already been put in practical use for engine valves and valve seats in the automotive industry in Japan [3]. The process involving injection of the powder mixture consisting of hard carbide particles and metallic ductile binders such as cobalt base alloys and nickel base alloys are very effective. It produces wear resistance layers on components subjected to large stresses involving complex loading conditions, such as abrasive wear, impact, of fatigue [4]. Before the carbide particles undergo melting of dissolution in the molten pool of ductile particles, rapid cooling leads to the rapid solidification producing fine composite structures which consist of undissolved carbides dispersed in the matrix of the ductile alloy. For example, the applications and researches about the mixed powder that adds WC particle are widely performed by
3386
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
roughness of 50 μm (Rz). Prior to laser cladding, the surfaces of substrates were cleaned using acetone. 2.2. Experimental facilities
Fig. 1. Image of multi-layer cladding. (a) Constant chemical composition multilayer cladding (CCCMLC) and (b) functionally gradient material multi-layer cladding (FGMMLC).
laser cladding such as Ni-based alloy powder + WC [5] and Cobased alloy powder + WC [6]. Though the wear resistance of the clad layer increases with the increase of WC weight fraction, the clad metal becomes brittle, and the brittle crack easily occurs due to the precipitation of complex carbides and residual tensile stress [6]. Preheating is adopted usually as an effective method to prevent the crack from occurring. The characteristic and crack in the clad layer with functionally gradient material multilayer cladding (FGMMLC) was also investigated by laser processes [7–11]. However, effect of preheating temperature for substrate and weight fraction of hard phase (such as carbide or boride) particles in the powder mixture on the crack sensitivity of the clad layer were not demonstrated with a detail spec. In the present research, the clad layers with high wear resistance and low crack sensitivity are needed to clad the machinery parts in the power plant. Therefore, the powder mixture of Co-based alloy powder (Stellite-6) and tungsten carbide (WC) was deposited on the mild steel plat (SM400B) by a 2.4 kW cw CO2 laser. Effect of the preheating temperature and WC weight fraction on the crack sensitivity and wear resistance were comparatively examined using a constant chemical composition multi-layer cladding (CCCMLC) and a functionally gradient material multi-layer cladding (FGMMLC). The images of constant chemical composition multi-layer cladding (CCC MLC) and functionally gradient material multi-layer cladding (FGMMLC) are shown in Fig. 1. In the CCCMLC, the WC weight fractions at each layer are constant. Namely, clad with powder mixture of constant chemical composition in the each layer. In the FGMMLC, the WC weight fractions at each layer (the 0 wt.% WC at 1st layer, 7 wt.% WC at 2nd layer and RWC wt.% (RWC: 10–47) at 3rd layer were used) are varied by controlling the disk rotation speed of the feeder. Namely, clad with powder mixture of functionally gradient chemical composition in each layer.
The schematic drawing laser cladding layout is shown in Fig. 2. A 2.4 kW cw CO2 laser was used in the experiments. A TWIN10-SPG powder feeding system was used to supply powder by carrying gas, Ar [12]. The powder feeding rate was varied by controlling the disk rotation speed of the feeder. The wear resistance of clad layer was evaluated using Ogoshi type wear testing machine. Experimental set-up and geometrical parameters involved in the wear test are shown in Fig. 3. The wear disc is made of quenched and tempered tool steel (AISI D2, hardness HRC58) with 15 mm radius (r) and 3 mm thickness (B). The applied load (P) gradually increased from 0 to 18.9 kg as wear progresses. The sliding distance (L) was fixed on 400 m. The friction speed was set at 0.308 m/s. Under the specified test conditions, wear resistance is evaluated by measuring the width of wear mark (h0). Specific wear rate (WS) is calculated by the following Eq. (1): Ws ¼
Bh30 ðmm2 =kgÞ 8drdPdL
For the wear test, three-pass clad layers were made on top of each other in a rectangular shape of 50 mm length, 15 mm width and total height of 2.0–2.5 mm. The deposition sequence is shown in Fig. 4. After laser cladding, the surfaces of specimen were ground to make a flat clad layer with a thickness of 1.8– 2.3 mm for wear test. The position of wear test is in central surface of clad layer. An electric stove was used for preheating. The preheating temperature was measured using an S-423K-01-1-TPC-1-ASP type temperature sensor and a HFT-40 type display device. The experimental set-up and image of substrate temperature control during laser cladding was shown in Fig. 5. The temperature of substrate was controlled to preheating temperature by adjusting the flowing rate of water during laser cladding.
2. Experimental procedure 2.1. Materials used Mild steel (JIS-SM400B: 0.13 mass% C, 0.19 mass% Si, 0.66 mass% Mn, 0.016 mass% P, 0.005 mass% S, bal. mass% Fe) plates of size 80 L × 30 W × 9 T mm were used as substrate metal. Stellite-6 powder (1.08 mass% C, 1.27 mass% Si, 1.63 mass% Ni, 28.32 mass% Cr, 2.04 mass% Fe, 4.33 mass% W, bal. mass% Co) in diameter of 63–250 μm and 99.7 mass% tungsten carbide (WC) in size of 45 μm were used as cladding materials. Surfaces of substrate plate were machined with
ð1Þ
Fig. 2. Schematic drawing of CO2 laser cladding setup for steel plate.
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
3387
Fig. 3. Wear testing. (a) Experimental set-up of Ogoshi type tester and (b) geometrical parameters after wear test (unit: mm).
After cladding, the specimen were cut across, polished, and etched with a mixture of HCl and HNO3 in 3:1 proportion. The microstructures were observed using a laser microscope (Keyence VK8510) and a scanning electron microscope (SEM). The chemical composition was measured by an energy dispersion spectroscopy (EDS). The phases in clad layer were analyzed by X-ray diffractometer with CuKα radiation. A step of 0.02° was used to scan 2θ from 20° to 120°. The Vickers hardness was measured by AKASHI AAV-500 automotive hardness tester made by AKASHI Corporation, Japan. The position of Vickers hardness test is in central clad layer from the substrate to the clad surface. Liquid Penetrant Testing is a method that is used to reveal surface breaking flaws by bleedout of a colored dye from the flaw. The technique is based on the ability of a liquid to be drawn into a “clean” surface breaking flaw by capillary action. After a period of time called the “dwell,” excess surface penetrant is removed and a developer is applied. This acts as a “blotter.” It draws the penetrant from the flaw to reveal its presence. The process shall be (1) pre-cleaning using a R-1M(NT) (2) apply penetrant using a R-1A(NT) (3) remove penetrant using a R-1M (NT) (4) developing using a R-1S(NT) (5) inspect the surface.
Fig. 5. Experimental set-up and image of substrate temperature control during laser cladding.
distance of 15 mm, beam size of 2.5 mm in diameter on the top of the substrate, overlapping rate of 50% in the multi-layer cladding, travel speed of 350 mm/min, powder feeding rate of 14 g/min, powder feeding nozzle calibre of 2 mm, shielding gas (Argon) flow rate of 20 l/min, carrier gas (Argon) flow rate of 3 l/min, room temperature of 30 °C. The weight fraction of WC in the powder mixture was varied between 0 and 47 wt.%. In the methods of CCCMLC and FGMMLC, three-pass clad layers were deposited using the same conditions as that for cladding wear test specimen as shown in Fig. 4. 3. Experimental results and discussion 3.1. Microstructures of the clad layer A binary Co–WC phase diagram [13] is shown in Fig. 6. The eutectic structure consists of Co-rich γ solid solution (γ) and tungsten carbide (WC) at the 55 wt.% WC and 1320 °C. The microstructure consists of hypoeutectic structure (γ + (γ + WC)) (γ: primary Co-rich solid solution, γ + WC: eutectic structure) at
2.3. Experiment conditions Experimental conditions of the CCCMLC and FGMMLC has been same, those are laser power of 2.1 kW, defocusing
Fig. 4. Scan pattern of deposition sequence for wear testing specimen and multilayer cladding specimen.
Fig. 6. Binary Co–WC constitution diagram.
3388
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
Fig. 7. Microstructure of clad layer made with powder mixture of Stellite-6 and 26 wt.% WC clad at room temperature. (a) Optical microstructure and (b) SEM image of hypoeutectic structure.
the range of 0 wt.%–55 wt.% WC, and the microstructure consists of hypereutectic structure (WC + (γ + WC)) (WC: primary tungsten carbide, γ + WC: eutectic structure) at the range of 55 wt.%–100 wt.% WC. In the present work, the WC weight fraction was in the range of 0 wt.%–47 wt.% WC, thus, the microstructure of the clad layer consists of hypoeutectic structure (γ + (γ + WC)) (γ: primary Co-rich solid solution, γ + WC: eutectic structure). The micrograph of the clad layer by laser microscope and SEM image with 26 wt.% WC clad at room temperature are shown in Fig. 7(a) and (b). In the clad layer, undissolved WC particles due to its high melting point (3683 K) dispersed in the matrix of the Co-based alloy, as shown in Fig. 7(a). The X-ray diffraction analysis results for the clad layer with 26 wt.% WC clad at room temperature is shown in Fig. 8. The existence of Co-rich γ phase, Cr23C6, Co3W3C and WC phase was confirmed by X-ray diffractometer. The EDS analysis results for Co, Cr, W, C, Fe, and Ni for the clad layer made with a powder mixture of 26 wt.% WC clad at room temperature is shown in Fig. 9. It can be noted that the amounts of Co and Fe are higher in the primary dendrite as compared to that in the interdendritic eutectic structure. Also the
Fig. 8. X-ray diffraction analysis results of clad layer made with powder mixture of Stellite-6 and 26 wt.% WC clad at room temperature.
Fig. 9. EDS analysis results of clad layer made with powder mixture of Stellite-6 and 26 wt.% WC clad at room temperature.
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
interdendritic eutectic structure contains more amounts of Cr, W, and C than in the primary dendrite. Moreover, the amount of Ni in the hypoeutectic structure is almost same. As analyzed earlier, the microstructure of the clad layer made with powder mixture of Stellite-6 and WC consists of hypoeutectic structure. The primary solidified dendrite phase in the hypoeutectic structure consists of Co-rich solid solution (γ). The eutectic structure consists of Co-rich γ phase mixed with complex carbides such as Cr23C6, Co3W3C and WC phase. Moreover, the undissolved WC particles dispersed in the matrix of the Co-based alloy were observed by the laser microscope and SEM image. The cross section and microstructures for clad layer made by FGMMLC (1st layer: Stellite-6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 26 wt.% WC) clad at preheating temperature of 450 °C are shown in Fig. 10 (In Fig. 10, the undissolved WC particles have been present actually in the clad layer. However, the undissolved WC particles in Fig. 10 were avoided to observe cleanly the matrix micrograph). The smooth clad layer was obtained, as Fig. 10(a). From fusion line to the clad surface, the solidification speed R increased gradually, and temperature gradient G decreased gradually [14–17], thus a planar growth, cellular growth, cellular dendrite growth, and equiaxial dendrite growth were observed in 1st clad layer, as shown in Fig. 10(d). A cellular growth, cellular dendrite growth, and equiaxial dendrite growth were observed in the clad layer of 2nd and 3rd, as shown in Fig. 10(b) and (c), respectively. In the heat-affected zone (HAZ) between clad layers, the coarse grain zone, mixed grain zone and fine grain zone due to the influence of the melting and reheating were observed by the laser microscope, as shown in Fig. 10(c). From the fusion line to the clad surface, the size of the primary solidified dendrite phase becomes small and small. It is thought that the thermal conductivity of the clad layer (SM400B: 30–60.5 J/m·s·K, Stellite-6: 30–45 J/m·s·K and WC: 29.3 J/m·s·K) may decrease gradually, thus temperature gradient G decreased, and, the constitutional supercooling [18,19] increased from the mild steel (SM400B) to the clad surface in the FGMMLC.
3389
ventional cladding processes, the laser beam cladding restricts the dilution to minimum due to the lower heat input. The more WC is added in the mixed powder from 1st layer to 3rd layer,
3.2. Hardness of the clad layer The microstructures of high hardness were obtained by laser cladding compared with other coating techniques have been shown and discussed in the literature [7,8,11,14,20]. However, no papers have mentioned the influence of overlapping track and clad layer. In the paper, the effect of the multi-layer deposition on the hardness was evaluated by measuring the hardness profiles. The hardness of clad layer for FGMMLC (1st layer: Stellite-6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 26 wt.% WC) clad at preheating temperature of 450 °C is shown in Fig. 11 (a). There was a gradual increase in the hardness from the substrate towards clad surface due to the variation of microstructure. The hardness near the substrate is low because of the dilution of the clad layer with Fe from the substrate. However, the width of the diluted clad layer was about 40 μm, which is very small as comparing to the con-
Fig. 10. Microstructures of clad layer in functionally gradient material multilayer cladding (FGMMLC) made with powder mixture of Stellite-6 and WC (1st layer: Stellite-6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 26 wt.% WC) clad at preheating of 450 °C. (a) Cross section of clad layer, (b) interface zone of 2nd layer and 3rd layer, (c) interface zone of 1st layer and 2nd layer, and (d) near the fusion line.
3390
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
used to evaluate the crack sensitivity as Eq. (2) for CCCMLC and Eq. (3) for FGMMLC.
Fig. 11. Vickers hardness of clad layer. (a) Functionally gradient material multilayer cladding made with 1st layer of Stellite-6 + 0 wt.% WC, 2nd layer of Stellite-6 + 7 wt.% WC, and 3rd layer of Stellite-6 + 26 wt.% WC clad at preheating of 450 °C, and (b) constant chemical composition multi-layer cladding (CCCMLC) made with Stellite-6 + 5 wt.% WC mixed powder clad at room temperature.
the higher the hardness of the clad layer. In hypoeutectic structures, the hardness of the clad layer increases due to the increase in the amount of eutectic structure, which contains hard complex carbides, as Cr23C6, Co3W3C and WC. On the other hand, hardness decreases at the interface between clad layers, this arises from the melting and reheating of a part of a track as the one adjacent to it is being deposited. Therefore, the coarse grain zone, mixed grain zone and fine grain zone was obtained at the interface between clad layers, as shown in Fig. 10(c). The hardness distribution of clad layer in CCCMLC made with a powder mixture of 5 wt.% WC clad at room temperature is shown in Fig. 11(b). The hardness in each layer does not change due to the same WC weight fraction. However, the hardness at interface between clad layers decreased as results of the remelting and reheating. 3.3. Cracks of the clad layer The effect of WC weight fraction and preheating temperature on crack sensitivity in CCCMLC and FGMMLC are shown in Fig. 12(a) and (b), respectively. The regression equations concerning preheating temperature and WC weight fraction can be
V −705 Tpr ¼ 105RWC
ð2Þ
V −1734 Tpr ¼ 84RWC
ð3Þ
′ : where, Tpr: critical preheating temperature (°C) and RWC critical WC weight fraction (wt.%). In the reproducibility experiments, the crack did not occur in the area above regression line, but occurred in the area below the regression line. The hatching area is a transition area in the experiments. The specimens of Liquid Penetrant Testing for CCCMLC and FGMMLC clad at room temperature are shown in Fig. 13(a) and (b), respectively. Crack did not appear in clad layer clad at room temperature when the WC weight fraction was lower than 7 wt.% in CCCMLC and when it was lower than 21 wt.% in FGMMLC at the 3rd layer. Clad at room temperature, the cracks appeared in the clad layer when the WC weight fraction was over than 11 wt.% in CCCMLC and its over than 26 wt.% WC in FGMMLC at the 3rd layer. The fractograph of the clad layer by SEM clad at room temperature with powder mixture of Stellite-6 and 32 wt.% WC are shown in Fig. 14. The facet pattern (Fig. 14(a)) and quasicleavage fracture (Fig. 14(b)) were observed. With increase of WC weight fraction, amount of W and C element increased in the clad layer due to the solution of the partial WC particles. The clad layer becomes brittle due to increases of W and C element because of the precipitation of carbides. Therefore, the quasicleavage fracture type crack occurred in the clad layer with the powder mixture of Stellite-6 and WC [12,21,22]. Compared with CCCMLC, the FGMMLC shows small regression line gradient. The WC weight fraction in FGMMLC at the 3rd layer was higher than that in CCCMLC when crack appeared. Thus, in FGMMLC with gradual increase of WC particles clad at three subsequent layers shifts the critical amount of WC at which cracking appears to the higher level, or that it reduces the required preheating temperature. Namely, the crack sensitivity of clad layer using FGMMLC method is lower than that of CCCMLC method. Therefore, compared with CCCMLC, the FGMMLC is an effective method to prevent the crack from occurring. It is thought that the residual plasticity strain from the 1st layer to the 3rd layer in FGMMLC becomes small gradually due to the expansion coefficients of the clad layer decreases when the WC weight fraction increases [23]. And, the residual stress from 1st layer to 3rd layer in FGMMLC becomes small gradually by the decrease of the residual plasticity strain, thus, the crack sensitivity becomes lower compared to CCCMLC. The expansion coefficient of Stellite-6 and WC are 1.45 × 10− 5/°C and 0.39 × 10− 5/°C, respectively. 3.4. Wear resistance of the clad layer The results of wear resistance for CCCMLC (in each layer: Stellite-6 + 21 wt.% WC) and FGMMLC (1st layer: Stellite-
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
3391
Fig. 12. Effect of WC weight fraction and preheating temperature on crack sensitivity of clad layer. (a) Constant chemical composition multi-layer cladding (CCCMLC) and (b) functionally gradient material multi-layer cladding (FGMMLC).
Fig. 13. Results of Liquid Penetrant Testing for multi-layer cladding specimens clad at room temperature. (a) Constant chemical composition multi-layer cladding (CCCMLC) specimens and (b) functionally gradient material multi-layer cladding (FGMMLC) specimens.
Fig. 14. Fractograph of clad layer made with powder mixture of Stellite-6 and 32 wt.% WC clad at room temperature. (a) Pattern of facet, and (b) quasi-cleavage fracture.
3392
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 21 wt.% WC) clad at preheating temperature of 450 °C are almost same about 8.7 × 10− 9 mm2/kg due to the same microstructures of clad layer in the 3rd layer. 4. Conclusion (1) In the range of 0 wt.%–47 wt.% WC, the microstructure of clad layer made with powder mixture of Stellite-6 and tungsten carbide consists of hypoeutectic structure. The primary phase consists of Co-rich γ solid solution, and the eutectic structure consists of Co-rich γ solid solution mixed with complex carbides such as Cr23C6, Co3W3C and WC phase. The eutectic structure was increased with the increase of WC weight fraction due to the more melted WC particles. On the other hand, undissolved WC particle due to its high melting point (3683 K) dispersed in the matrix of Co-based alloy. (2) In the multi-layer cladding, there was a gradual increase in the hardness from the substrate towards clad surface due to the variation of microstructure. The hardness near the substrate is low because of the dilution of the clad layer with Fe from the substrate. On the other hand, the more WC is added in the mixed powder, the higher the hardness of the clad layer. Therefore, in FGMMLC, from 1st clad layer to 3rd clad layer, the hardness increased with the increase of WC weight fraction, and the hardness near interface between clad layers decreased due to the melting and reheating. In CCCMLC, though the hardness in each layer does not change due to the same WC weight fraction, at interface between clad layers it decreased as results of the melting and reheating. (3) The powder mixture of Stellite-6 and WC was deposited to mild steel (SM400B) using a CCCMLC and FGMMLC. The cracks of clad layer belonged to quasi-cleavage fracture type crack. (4) In FGMMLC with gradual increase of WC particles clad at three subsequent layers shifts the critical amount of WC at which cracking appears to the higher level, or that it reduces the required preheating temperature. Therefore, the crack sensitivity of clad layer using FGMMLC method is lower than that of CCCMLC method. The
FGMMLC is an effective method to prevent the crack from occurring. (5) The wear resistance of clad layer for CCCMLC and FGMMLC is almost same due to the same microstructures by the same WC weight fraction in the 3rd layer. References [1] Koujirou Kobayashi, Weld. Technol. 47 (8) (1999) 66 (in Japanese). [2] Teru Nagai, Tadasi Inoue, Hirofumi Shimura, J. Laser Appl. 14 (3) (2002) 191. [3] Kutsuna Muneharu, Weld. Technol. 50 (5) (2002) 127. [4] A. Techel, A. Luft, A. Muller, S. Nowotny, Opt. Quantum Electron. 27 (1995) 1311. [5] P. Wu, H.M. Du, X.L. Chen, Z.Q. Li, H.L. Bai, E.Y. Jiang, Wear 257 (1–2) (2004) 142. [6] Minlin Zhong, Kefu Yao, Wenjin Liu, J. Laser Appl. 13 (6) (2001) 247. [7] P. Wu, H.M. Du, X.L. Chen, Z.Q. Li, H.L. Bai, E.Y. Jiang, Wear 257 (1–2) (July 2004) 142. [8] A. Yakovlev, E. Trunova, D. Grevey, M. Pilloz, I. Smurov, Surf. Coat. Technol. 190-1–3 (January 2005) 15. [9] Xun Cai, Xiaoyu Yang, Tao Zhao, Liuhe Li, Qiulong Chen, Key Eng. Mater. 297–300 (part 1–4) (2005) 2813. [10] K. Mohammed Jasim, R.D. Rawlings, D.R.F. West, J. Mater. Sci. 28–10 (1993) 2820. [11] J.H. Abboud, D.R.F. West, R.D. Rawlings, J. Mater. Sci. 29–13 (1994) 3393. [12] Muneharu Kutsuna, Watanabe Masayoshi, Guojian Xu, Katsusige Yamada, Q. J. Jpn. Weld. Soc. 22 (1) (2004) 37 (in Japanese). [13] Japan Metal Society, Metallic Handbook (reissue), Maruzen Corporation, Tokyo, 1960, p. 1344, (in Japanese). [14] Guojian Xu, Muneharu Kutsuna, Katsusige Yamada, Q. J. Jpn. Weld. Soc. 23 (2) (2005) 286 (in Japanese). [15] Seiji Katayama, J. Light Metal Weld. Constr. 38 (3) (2000) 9 (in Japanese). [16] Shirou Ohara, Metal material generalization (Enlarged edition), Asakura Bookstore, Tokyo, 2001, p. 193, (in Japanese). [17] Guojian Xu, Muneharu Kutsuna, Hong Zhang, Zhongjie Liu, Mitsuaki Haneda, Hiroo Koide, Q. J. Jpn. Weld. Soc. 23 (3) (2005) 412 (in Japanese). [18] S. Kou, Welding Metallurgy, John Wiley & Sons, New York, 1987, p. 167. [19] W.F. Savage, Weld. World 18 (5/6) (1980) 89. [20] Guojian Xu, Muneharu Kutsuna, Zhongjie Liu, Hong Zhang, Mater. Sci. Eng., A 417 (1–2) (2006) 63. [21] Qiang Li, Dawei Zhang, Tingquan Lei, Chuanzhong Chen, Wenzhe Chen, Surf. Coat. Technol. 137 (2–3) (2001) 122. [22] Mingju Chao, Erjun Liang, Surf. Coat. Technol. 179 (2–3) (2004) 265. [23] Koichi, Masubuchi, Analysis of Welded Structures, Printed and bound in Great Britain by Willian Clowes (Beccles) Limied, Beccle and London 1980, pp. 400–448.
Characteristic behaviours of clad layer by a multi-layer laser cladding with powder mixture of Stellite-6 and tungsten carbide Guojian Xu a,⁎,1 , Munaharu Kutsuna a , Zhongjie Liu a , Liquan Sun b a
Graduate School of Engineering, Nagoya University, 1 Furo-cho Chikusa-ku, Nagoya 464-8603, Japan b Chang Chun Institute of Technology, China Received 11 May 2006; accepted in revised form 17 July 2006 Available online 1 September 2006
Abstract The powder mixture of Stellite-6 and tungsten carbide (WC) in the range of 0 wt.%–47 wt.% WC fraction with constant chemical composition multi-layer cladding (CCCMLC) and functionally gradient material multi-layer cladding (FGMMLC) were deposited on mild steel (JIS-SM400B) plates by a 2.4 kW cw CO2 laser. In the CCCMLC, the WC weight fractions at each layer are constant. In the FGMMLC, the WC weight fractions at each layer are varied by controlling the disk rotation speed of the feeder. The phase constitution, microstructure, hardness and wear resistance of the clad layer were investigated by an X-ray diffractometer, energy dispersion spectroscopy (EDS), scanning electron microscope (SEM), laser microscope, Vickers hardness tester and wear tester. According to the analyzed results, the microstructure of the clad layer consists of hypoeutectic structure, and undissolved tungsten carbides dispersed in the matrix of the Co-based alloy. The Vickers hardness increases with the increase of WC weight fraction. The clad layers of CCCMLC and FGMMLC have almost same wear resistance. On the other hand, to decrease the crack sensitivity of clad layer, the influence of a preheating temperature and WC weight fraction in the CCCMLC and FGMMLC on the crack sensitivity was investigated comparatively. As the result, the cracks of clad layer clad with the powder mixture of Stellite-6 and WC belonged to a quasicleavage fracture type crack. In the FGMMLC method with gradual increase of WC particles clad at three subsequent layers shifts the critical amount of WC at which cracking appears to the higher level, or that it reduces the required preheating temperature. Namely, the crack sensitivity of clad layer using FGMMLC method is lower than that of CCCMLC method. © 2006 Elsevier B.V. All rights reserved. Keywords: Laser cladding; Quasi-cleavage fracture; Crack sensitivity; Constant chemical composition multi-layer cladding (CCCMLC); Functionally gradient material multi-layer cladding (FGMMLC); Hypoeutectic structure
1. Introduction Advantages of laser cladding are to produce a clad layer of composite materials, reduced dilution and deformation and excellent wear resistance by a fast cladding process. Dilution can be controlled by varying defocusing distance, travel speed and powder feeding rate [1]. Therefore in recent years, there has been rapid development of the laser surface treatment technique ⁎ Corresponding author. E-mail addresses: [email protected] (G. Xu), [email protected] (M. Kutsuna), [email protected] (Z. Liu), [email protected] (L. Sun). 1 Present: Technical Development, LASER X CO., LTD. 7 Kofukada Shinbayashi-cho, Chiryu-shi Aichi 472-0017 JAPAN. 0257-8972/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2006.07.210
in the production domain [2]. It has already been put in practical use for engine valves and valve seats in the automotive industry in Japan [3]. The process involving injection of the powder mixture consisting of hard carbide particles and metallic ductile binders such as cobalt base alloys and nickel base alloys are very effective. It produces wear resistance layers on components subjected to large stresses involving complex loading conditions, such as abrasive wear, impact, of fatigue [4]. Before the carbide particles undergo melting of dissolution in the molten pool of ductile particles, rapid cooling leads to the rapid solidification producing fine composite structures which consist of undissolved carbides dispersed in the matrix of the ductile alloy. For example, the applications and researches about the mixed powder that adds WC particle are widely performed by
3386
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
roughness of 50 μm (Rz). Prior to laser cladding, the surfaces of substrates were cleaned using acetone. 2.2. Experimental facilities
Fig. 1. Image of multi-layer cladding. (a) Constant chemical composition multilayer cladding (CCCMLC) and (b) functionally gradient material multi-layer cladding (FGMMLC).
laser cladding such as Ni-based alloy powder + WC [5] and Cobased alloy powder + WC [6]. Though the wear resistance of the clad layer increases with the increase of WC weight fraction, the clad metal becomes brittle, and the brittle crack easily occurs due to the precipitation of complex carbides and residual tensile stress [6]. Preheating is adopted usually as an effective method to prevent the crack from occurring. The characteristic and crack in the clad layer with functionally gradient material multilayer cladding (FGMMLC) was also investigated by laser processes [7–11]. However, effect of preheating temperature for substrate and weight fraction of hard phase (such as carbide or boride) particles in the powder mixture on the crack sensitivity of the clad layer were not demonstrated with a detail spec. In the present research, the clad layers with high wear resistance and low crack sensitivity are needed to clad the machinery parts in the power plant. Therefore, the powder mixture of Co-based alloy powder (Stellite-6) and tungsten carbide (WC) was deposited on the mild steel plat (SM400B) by a 2.4 kW cw CO2 laser. Effect of the preheating temperature and WC weight fraction on the crack sensitivity and wear resistance were comparatively examined using a constant chemical composition multi-layer cladding (CCCMLC) and a functionally gradient material multi-layer cladding (FGMMLC). The images of constant chemical composition multi-layer cladding (CCC MLC) and functionally gradient material multi-layer cladding (FGMMLC) are shown in Fig. 1. In the CCCMLC, the WC weight fractions at each layer are constant. Namely, clad with powder mixture of constant chemical composition in the each layer. In the FGMMLC, the WC weight fractions at each layer (the 0 wt.% WC at 1st layer, 7 wt.% WC at 2nd layer and RWC wt.% (RWC: 10–47) at 3rd layer were used) are varied by controlling the disk rotation speed of the feeder. Namely, clad with powder mixture of functionally gradient chemical composition in each layer.
The schematic drawing laser cladding layout is shown in Fig. 2. A 2.4 kW cw CO2 laser was used in the experiments. A TWIN10-SPG powder feeding system was used to supply powder by carrying gas, Ar [12]. The powder feeding rate was varied by controlling the disk rotation speed of the feeder. The wear resistance of clad layer was evaluated using Ogoshi type wear testing machine. Experimental set-up and geometrical parameters involved in the wear test are shown in Fig. 3. The wear disc is made of quenched and tempered tool steel (AISI D2, hardness HRC58) with 15 mm radius (r) and 3 mm thickness (B). The applied load (P) gradually increased from 0 to 18.9 kg as wear progresses. The sliding distance (L) was fixed on 400 m. The friction speed was set at 0.308 m/s. Under the specified test conditions, wear resistance is evaluated by measuring the width of wear mark (h0). Specific wear rate (WS) is calculated by the following Eq. (1): Ws ¼
Bh30 ðmm2 =kgÞ 8drdPdL
For the wear test, three-pass clad layers were made on top of each other in a rectangular shape of 50 mm length, 15 mm width and total height of 2.0–2.5 mm. The deposition sequence is shown in Fig. 4. After laser cladding, the surfaces of specimen were ground to make a flat clad layer with a thickness of 1.8– 2.3 mm for wear test. The position of wear test is in central surface of clad layer. An electric stove was used for preheating. The preheating temperature was measured using an S-423K-01-1-TPC-1-ASP type temperature sensor and a HFT-40 type display device. The experimental set-up and image of substrate temperature control during laser cladding was shown in Fig. 5. The temperature of substrate was controlled to preheating temperature by adjusting the flowing rate of water during laser cladding.
2. Experimental procedure 2.1. Materials used Mild steel (JIS-SM400B: 0.13 mass% C, 0.19 mass% Si, 0.66 mass% Mn, 0.016 mass% P, 0.005 mass% S, bal. mass% Fe) plates of size 80 L × 30 W × 9 T mm were used as substrate metal. Stellite-6 powder (1.08 mass% C, 1.27 mass% Si, 1.63 mass% Ni, 28.32 mass% Cr, 2.04 mass% Fe, 4.33 mass% W, bal. mass% Co) in diameter of 63–250 μm and 99.7 mass% tungsten carbide (WC) in size of 45 μm were used as cladding materials. Surfaces of substrate plate were machined with
ð1Þ
Fig. 2. Schematic drawing of CO2 laser cladding setup for steel plate.
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
3387
Fig. 3. Wear testing. (a) Experimental set-up of Ogoshi type tester and (b) geometrical parameters after wear test (unit: mm).
After cladding, the specimen were cut across, polished, and etched with a mixture of HCl and HNO3 in 3:1 proportion. The microstructures were observed using a laser microscope (Keyence VK8510) and a scanning electron microscope (SEM). The chemical composition was measured by an energy dispersion spectroscopy (EDS). The phases in clad layer were analyzed by X-ray diffractometer with CuKα radiation. A step of 0.02° was used to scan 2θ from 20° to 120°. The Vickers hardness was measured by AKASHI AAV-500 automotive hardness tester made by AKASHI Corporation, Japan. The position of Vickers hardness test is in central clad layer from the substrate to the clad surface. Liquid Penetrant Testing is a method that is used to reveal surface breaking flaws by bleedout of a colored dye from the flaw. The technique is based on the ability of a liquid to be drawn into a “clean” surface breaking flaw by capillary action. After a period of time called the “dwell,” excess surface penetrant is removed and a developer is applied. This acts as a “blotter.” It draws the penetrant from the flaw to reveal its presence. The process shall be (1) pre-cleaning using a R-1M(NT) (2) apply penetrant using a R-1A(NT) (3) remove penetrant using a R-1M (NT) (4) developing using a R-1S(NT) (5) inspect the surface.
Fig. 5. Experimental set-up and image of substrate temperature control during laser cladding.
distance of 15 mm, beam size of 2.5 mm in diameter on the top of the substrate, overlapping rate of 50% in the multi-layer cladding, travel speed of 350 mm/min, powder feeding rate of 14 g/min, powder feeding nozzle calibre of 2 mm, shielding gas (Argon) flow rate of 20 l/min, carrier gas (Argon) flow rate of 3 l/min, room temperature of 30 °C. The weight fraction of WC in the powder mixture was varied between 0 and 47 wt.%. In the methods of CCCMLC and FGMMLC, three-pass clad layers were deposited using the same conditions as that for cladding wear test specimen as shown in Fig. 4. 3. Experimental results and discussion 3.1. Microstructures of the clad layer A binary Co–WC phase diagram [13] is shown in Fig. 6. The eutectic structure consists of Co-rich γ solid solution (γ) and tungsten carbide (WC) at the 55 wt.% WC and 1320 °C. The microstructure consists of hypoeutectic structure (γ + (γ + WC)) (γ: primary Co-rich solid solution, γ + WC: eutectic structure) at
2.3. Experiment conditions Experimental conditions of the CCCMLC and FGMMLC has been same, those are laser power of 2.1 kW, defocusing
Fig. 4. Scan pattern of deposition sequence for wear testing specimen and multilayer cladding specimen.
Fig. 6. Binary Co–WC constitution diagram.
3388
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
Fig. 7. Microstructure of clad layer made with powder mixture of Stellite-6 and 26 wt.% WC clad at room temperature. (a) Optical microstructure and (b) SEM image of hypoeutectic structure.
the range of 0 wt.%–55 wt.% WC, and the microstructure consists of hypereutectic structure (WC + (γ + WC)) (WC: primary tungsten carbide, γ + WC: eutectic structure) at the range of 55 wt.%–100 wt.% WC. In the present work, the WC weight fraction was in the range of 0 wt.%–47 wt.% WC, thus, the microstructure of the clad layer consists of hypoeutectic structure (γ + (γ + WC)) (γ: primary Co-rich solid solution, γ + WC: eutectic structure). The micrograph of the clad layer by laser microscope and SEM image with 26 wt.% WC clad at room temperature are shown in Fig. 7(a) and (b). In the clad layer, undissolved WC particles due to its high melting point (3683 K) dispersed in the matrix of the Co-based alloy, as shown in Fig. 7(a). The X-ray diffraction analysis results for the clad layer with 26 wt.% WC clad at room temperature is shown in Fig. 8. The existence of Co-rich γ phase, Cr23C6, Co3W3C and WC phase was confirmed by X-ray diffractometer. The EDS analysis results for Co, Cr, W, C, Fe, and Ni for the clad layer made with a powder mixture of 26 wt.% WC clad at room temperature is shown in Fig. 9. It can be noted that the amounts of Co and Fe are higher in the primary dendrite as compared to that in the interdendritic eutectic structure. Also the
Fig. 8. X-ray diffraction analysis results of clad layer made with powder mixture of Stellite-6 and 26 wt.% WC clad at room temperature.
Fig. 9. EDS analysis results of clad layer made with powder mixture of Stellite-6 and 26 wt.% WC clad at room temperature.
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
interdendritic eutectic structure contains more amounts of Cr, W, and C than in the primary dendrite. Moreover, the amount of Ni in the hypoeutectic structure is almost same. As analyzed earlier, the microstructure of the clad layer made with powder mixture of Stellite-6 and WC consists of hypoeutectic structure. The primary solidified dendrite phase in the hypoeutectic structure consists of Co-rich solid solution (γ). The eutectic structure consists of Co-rich γ phase mixed with complex carbides such as Cr23C6, Co3W3C and WC phase. Moreover, the undissolved WC particles dispersed in the matrix of the Co-based alloy were observed by the laser microscope and SEM image. The cross section and microstructures for clad layer made by FGMMLC (1st layer: Stellite-6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 26 wt.% WC) clad at preheating temperature of 450 °C are shown in Fig. 10 (In Fig. 10, the undissolved WC particles have been present actually in the clad layer. However, the undissolved WC particles in Fig. 10 were avoided to observe cleanly the matrix micrograph). The smooth clad layer was obtained, as Fig. 10(a). From fusion line to the clad surface, the solidification speed R increased gradually, and temperature gradient G decreased gradually [14–17], thus a planar growth, cellular growth, cellular dendrite growth, and equiaxial dendrite growth were observed in 1st clad layer, as shown in Fig. 10(d). A cellular growth, cellular dendrite growth, and equiaxial dendrite growth were observed in the clad layer of 2nd and 3rd, as shown in Fig. 10(b) and (c), respectively. In the heat-affected zone (HAZ) between clad layers, the coarse grain zone, mixed grain zone and fine grain zone due to the influence of the melting and reheating were observed by the laser microscope, as shown in Fig. 10(c). From the fusion line to the clad surface, the size of the primary solidified dendrite phase becomes small and small. It is thought that the thermal conductivity of the clad layer (SM400B: 30–60.5 J/m·s·K, Stellite-6: 30–45 J/m·s·K and WC: 29.3 J/m·s·K) may decrease gradually, thus temperature gradient G decreased, and, the constitutional supercooling [18,19] increased from the mild steel (SM400B) to the clad surface in the FGMMLC.
3389
ventional cladding processes, the laser beam cladding restricts the dilution to minimum due to the lower heat input. The more WC is added in the mixed powder from 1st layer to 3rd layer,
3.2. Hardness of the clad layer The microstructures of high hardness were obtained by laser cladding compared with other coating techniques have been shown and discussed in the literature [7,8,11,14,20]. However, no papers have mentioned the influence of overlapping track and clad layer. In the paper, the effect of the multi-layer deposition on the hardness was evaluated by measuring the hardness profiles. The hardness of clad layer for FGMMLC (1st layer: Stellite-6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 26 wt.% WC) clad at preheating temperature of 450 °C is shown in Fig. 11 (a). There was a gradual increase in the hardness from the substrate towards clad surface due to the variation of microstructure. The hardness near the substrate is low because of the dilution of the clad layer with Fe from the substrate. However, the width of the diluted clad layer was about 40 μm, which is very small as comparing to the con-
Fig. 10. Microstructures of clad layer in functionally gradient material multilayer cladding (FGMMLC) made with powder mixture of Stellite-6 and WC (1st layer: Stellite-6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 26 wt.% WC) clad at preheating of 450 °C. (a) Cross section of clad layer, (b) interface zone of 2nd layer and 3rd layer, (c) interface zone of 1st layer and 2nd layer, and (d) near the fusion line.
3390
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
used to evaluate the crack sensitivity as Eq. (2) for CCCMLC and Eq. (3) for FGMMLC.
Fig. 11. Vickers hardness of clad layer. (a) Functionally gradient material multilayer cladding made with 1st layer of Stellite-6 + 0 wt.% WC, 2nd layer of Stellite-6 + 7 wt.% WC, and 3rd layer of Stellite-6 + 26 wt.% WC clad at preheating of 450 °C, and (b) constant chemical composition multi-layer cladding (CCCMLC) made with Stellite-6 + 5 wt.% WC mixed powder clad at room temperature.
the higher the hardness of the clad layer. In hypoeutectic structures, the hardness of the clad layer increases due to the increase in the amount of eutectic structure, which contains hard complex carbides, as Cr23C6, Co3W3C and WC. On the other hand, hardness decreases at the interface between clad layers, this arises from the melting and reheating of a part of a track as the one adjacent to it is being deposited. Therefore, the coarse grain zone, mixed grain zone and fine grain zone was obtained at the interface between clad layers, as shown in Fig. 10(c). The hardness distribution of clad layer in CCCMLC made with a powder mixture of 5 wt.% WC clad at room temperature is shown in Fig. 11(b). The hardness in each layer does not change due to the same WC weight fraction. However, the hardness at interface between clad layers decreased as results of the remelting and reheating. 3.3. Cracks of the clad layer The effect of WC weight fraction and preheating temperature on crack sensitivity in CCCMLC and FGMMLC are shown in Fig. 12(a) and (b), respectively. The regression equations concerning preheating temperature and WC weight fraction can be
V −705 Tpr ¼ 105RWC
ð2Þ
V −1734 Tpr ¼ 84RWC
ð3Þ
′ : where, Tpr: critical preheating temperature (°C) and RWC critical WC weight fraction (wt.%). In the reproducibility experiments, the crack did not occur in the area above regression line, but occurred in the area below the regression line. The hatching area is a transition area in the experiments. The specimens of Liquid Penetrant Testing for CCCMLC and FGMMLC clad at room temperature are shown in Fig. 13(a) and (b), respectively. Crack did not appear in clad layer clad at room temperature when the WC weight fraction was lower than 7 wt.% in CCCMLC and when it was lower than 21 wt.% in FGMMLC at the 3rd layer. Clad at room temperature, the cracks appeared in the clad layer when the WC weight fraction was over than 11 wt.% in CCCMLC and its over than 26 wt.% WC in FGMMLC at the 3rd layer. The fractograph of the clad layer by SEM clad at room temperature with powder mixture of Stellite-6 and 32 wt.% WC are shown in Fig. 14. The facet pattern (Fig. 14(a)) and quasicleavage fracture (Fig. 14(b)) were observed. With increase of WC weight fraction, amount of W and C element increased in the clad layer due to the solution of the partial WC particles. The clad layer becomes brittle due to increases of W and C element because of the precipitation of carbides. Therefore, the quasicleavage fracture type crack occurred in the clad layer with the powder mixture of Stellite-6 and WC [12,21,22]. Compared with CCCMLC, the FGMMLC shows small regression line gradient. The WC weight fraction in FGMMLC at the 3rd layer was higher than that in CCCMLC when crack appeared. Thus, in FGMMLC with gradual increase of WC particles clad at three subsequent layers shifts the critical amount of WC at which cracking appears to the higher level, or that it reduces the required preheating temperature. Namely, the crack sensitivity of clad layer using FGMMLC method is lower than that of CCCMLC method. Therefore, compared with CCCMLC, the FGMMLC is an effective method to prevent the crack from occurring. It is thought that the residual plasticity strain from the 1st layer to the 3rd layer in FGMMLC becomes small gradually due to the expansion coefficients of the clad layer decreases when the WC weight fraction increases [23]. And, the residual stress from 1st layer to 3rd layer in FGMMLC becomes small gradually by the decrease of the residual plasticity strain, thus, the crack sensitivity becomes lower compared to CCCMLC. The expansion coefficient of Stellite-6 and WC are 1.45 × 10− 5/°C and 0.39 × 10− 5/°C, respectively. 3.4. Wear resistance of the clad layer The results of wear resistance for CCCMLC (in each layer: Stellite-6 + 21 wt.% WC) and FGMMLC (1st layer: Stellite-
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
3391
Fig. 12. Effect of WC weight fraction and preheating temperature on crack sensitivity of clad layer. (a) Constant chemical composition multi-layer cladding (CCCMLC) and (b) functionally gradient material multi-layer cladding (FGMMLC).
Fig. 13. Results of Liquid Penetrant Testing for multi-layer cladding specimens clad at room temperature. (a) Constant chemical composition multi-layer cladding (CCCMLC) specimens and (b) functionally gradient material multi-layer cladding (FGMMLC) specimens.
Fig. 14. Fractograph of clad layer made with powder mixture of Stellite-6 and 32 wt.% WC clad at room temperature. (a) Pattern of facet, and (b) quasi-cleavage fracture.
3392
G. Xu et al. / Surface & Coatings Technology 201 (2006) 3385–3392
6 + 0 wt.% WC, 2nd layer: Stellite-6 + 7 wt.% WC, 3rd layer: Stellite-6 + 21 wt.% WC) clad at preheating temperature of 450 °C are almost same about 8.7 × 10− 9 mm2/kg due to the same microstructures of clad layer in the 3rd layer. 4. Conclusion (1) In the range of 0 wt.%–47 wt.% WC, the microstructure of clad layer made with powder mixture of Stellite-6 and tungsten carbide consists of hypoeutectic structure. The primary phase consists of Co-rich γ solid solution, and the eutectic structure consists of Co-rich γ solid solution mixed with complex carbides such as Cr23C6, Co3W3C and WC phase. The eutectic structure was increased with the increase of WC weight fraction due to the more melted WC particles. On the other hand, undissolved WC particle due to its high melting point (3683 K) dispersed in the matrix of Co-based alloy. (2) In the multi-layer cladding, there was a gradual increase in the hardness from the substrate towards clad surface due to the variation of microstructure. The hardness near the substrate is low because of the dilution of the clad layer with Fe from the substrate. On the other hand, the more WC is added in the mixed powder, the higher the hardness of the clad layer. Therefore, in FGMMLC, from 1st clad layer to 3rd clad layer, the hardness increased with the increase of WC weight fraction, and the hardness near interface between clad layers decreased due to the melting and reheating. In CCCMLC, though the hardness in each layer does not change due to the same WC weight fraction, at interface between clad layers it decreased as results of the melting and reheating. (3) The powder mixture of Stellite-6 and WC was deposited to mild steel (SM400B) using a CCCMLC and FGMMLC. The cracks of clad layer belonged to quasi-cleavage fracture type crack. (4) In FGMMLC with gradual increase of WC particles clad at three subsequent layers shifts the critical amount of WC at which cracking appears to the higher level, or that it reduces the required preheating temperature. Therefore, the crack sensitivity of clad layer using FGMMLC method is lower than that of CCCMLC method. The
FGMMLC is an effective method to prevent the crack from occurring. (5) The wear resistance of clad layer for CCCMLC and FGMMLC is almost same due to the same microstructures by the same WC weight fraction in the 3rd layer. References [1] Koujirou Kobayashi, Weld. Technol. 47 (8) (1999) 66 (in Japanese). [2] Teru Nagai, Tadasi Inoue, Hirofumi Shimura, J. Laser Appl. 14 (3) (2002) 191. [3] Kutsuna Muneharu, Weld. Technol. 50 (5) (2002) 127. [4] A. Techel, A. Luft, A. Muller, S. Nowotny, Opt. Quantum Electron. 27 (1995) 1311. [5] P. Wu, H.M. Du, X.L. Chen, Z.Q. Li, H.L. Bai, E.Y. Jiang, Wear 257 (1–2) (2004) 142. [6] Minlin Zhong, Kefu Yao, Wenjin Liu, J. Laser Appl. 13 (6) (2001) 247. [7] P. Wu, H.M. Du, X.L. Chen, Z.Q. Li, H.L. Bai, E.Y. Jiang, Wear 257 (1–2) (July 2004) 142. [8] A. Yakovlev, E. Trunova, D. Grevey, M. Pilloz, I. Smurov, Surf. Coat. Technol. 190-1–3 (January 2005) 15. [9] Xun Cai, Xiaoyu Yang, Tao Zhao, Liuhe Li, Qiulong Chen, Key Eng. Mater. 297–300 (part 1–4) (2005) 2813. [10] K. Mohammed Jasim, R.D. Rawlings, D.R.F. West, J. Mater. Sci. 28–10 (1993) 2820. [11] J.H. Abboud, D.R.F. West, R.D. Rawlings, J. Mater. Sci. 29–13 (1994) 3393. [12] Muneharu Kutsuna, Watanabe Masayoshi, Guojian Xu, Katsusige Yamada, Q. J. Jpn. Weld. Soc. 22 (1) (2004) 37 (in Japanese). [13] Japan Metal Society, Metallic Handbook (reissue), Maruzen Corporation, Tokyo, 1960, p. 1344, (in Japanese). [14] Guojian Xu, Muneharu Kutsuna, Katsusige Yamada, Q. J. Jpn. Weld. Soc. 23 (2) (2005) 286 (in Japanese). [15] Seiji Katayama, J. Light Metal Weld. Constr. 38 (3) (2000) 9 (in Japanese). [16] Shirou Ohara, Metal material generalization (Enlarged edition), Asakura Bookstore, Tokyo, 2001, p. 193, (in Japanese). [17] Guojian Xu, Muneharu Kutsuna, Hong Zhang, Zhongjie Liu, Mitsuaki Haneda, Hiroo Koide, Q. J. Jpn. Weld. Soc. 23 (3) (2005) 412 (in Japanese). [18] S. Kou, Welding Metallurgy, John Wiley & Sons, New York, 1987, p. 167. [19] W.F. Savage, Weld. World 18 (5/6) (1980) 89. [20] Guojian Xu, Muneharu Kutsuna, Zhongjie Liu, Hong Zhang, Mater. Sci. Eng., A 417 (1–2) (2006) 63. [21] Qiang Li, Dawei Zhang, Tingquan Lei, Chuanzhong Chen, Wenzhe Chen, Surf. Coat. Technol. 137 (2–3) (2001) 122. [22] Mingju Chao, Erjun Liang, Surf. Coat. Technol. 179 (2–3) (2004) 265. [23] Koichi, Masubuchi, Analysis of Welded Structures, Printed and bound in Great Britain by Willian Clowes (Beccles) Limied, Beccle and London 1980, pp. 400–448.