Microstructural evolution and mechanical property of Ti-6Al-4V wall deposited by continuous plasma arc additive manufacturing without post heat treatment

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

Contents lists available at ScienceDirect

Journal of the Mechanical Behavior of Biomedical Materials journal homepage: www.elsevier.com/locate/jmbbm

Microstructural evolution and mechanical property of Ti-6Al-4V wall deposited by continuous plasma arc additive manufacturing without post heat treatment

MARK

⁎

Jianjun Lina,b, , Yaohui Lvb, Yuxin Liub, Zhe Sunb, Kaibo Wangb, Zhuguo Lia, Yixiong Wua, Binshi Xua,b a b

Shanghai Key Laboratory of Materials Laser Processing and Modification, Shanghai Jiao Tong University, Shanghai 200240, China National Key Laboratory for Remanufacturing, Academy of Armored Forces Engineering, Beijing 100072, China

A R T I C L E I N F O

A BS T RAC T

Keywords: Microstructural evolution Mechanical properties Ti-6Al-4V alloy Continuous plasma arc additive manufacturing

Plasma arc additive manufacturing (PAM) is a novel additive manufacturing (AM) technology due to its big potential in improving efficiency, convenience and being cost-savings compared to other AM processes of high energy bea\m. In this research, several Ti-6Al-4V thin walls were deposited by optimized weld wire-feed continuous PAM process (CPAM), in which the heat input was gradually decreased layer by layer. The deposited thin wall consisted of various morphologies, which includes epitaxial growth of prior β grains, horizontal layer bands, martensite and basket weave microstructure, that depends on the heat input, multiple thermal cycles and gradual cooling rate in the deposition process. By gradually reducing heat input of each bead and using continuous current in the PAM process, the average yield strength (YS), ultimate tensile strength (UTS) and elongation reach about 877 MPa, 968 MPa and 1.5%, respectively, which exceed the standard level of forging. The mechanical property was strengthened and toughened due to weakening the aspect ratio of prior β grains and separating nano-dispersoids among α lamellar. Furthermore, this research demonstrates that the CPAM process has a potential to manufacture or remanufacture in AM components of metallic biomaterials without post-processing heat treatment.

1. Introduction Titanium alloys, particularly Ti-6Al-4V, have been widely used in aerospace, aircraft, automotive, biomedical and chemical industries due to their excellent combination of strength, fracture toughness, low density, and very good corrosion resistance (Peters and Christoph, 2003; Donachie, 2000; Banerjee and Williams, 2013). However, material cost is regarded as the biggest barrier for further application considered the output of lower value and high price sensitive products (Lujering and Williams., 2007), the main reason is their poor machinability caused by the low thermal conductivity and high chemical reactivity with cutting tool materials (Davim, 2014). Fortunately, additive manufacturing (AM) technologies offer the potential to reduce cost, energy consumption and carbon footprint (Brooks and Molony., 2016; Paydas et al., 2016; Ding et al., 2015). Typical AM technologies include Tungsten Inter Gas welding process (TIG) (Brandl et al., 2010; Szost et al., 2016), Electron Beam Melting (EBM) (Edwards et al., 2013; Murr et al., 2009; Zhao et al.,

⁎

2016), Laser Beam Deposition (LBD) (Brandl et al., 2011; Miranda et al., 2008; Kelly and Kampe 2004), as well as Plasma Arc Additive Manufacturing(PAM). EBM and LBD processes are characterized by large temperature gradient and high cooling rate, and suit for precise parts or particular applications. Generally, the width of a certain thin component is larger than the spot diameter of laser beam or electrical beam, thus both processes need multiple passes for such components, which will result higher residual stress (Szost et al., 2016). Besides, a rapid solidification gives rise to the occurrence of segregation and the presence of coarse prior β-grain (Edwards et al., 2013; Brandl et al., 2011; Tan et al., 2015; Qiu et al., 2015a). Particularly, narrow prior β-grain with large aspect ratio always grows up to the top surface of deposited parts, which results in obvious anisotropy. Therefore, the parts normally need to be heat treated after deposition or hot isostatically pressed (Zhu, et al., 2015; Qiu et al., 2015b). With a big potential in improving efficiency, convenience and being cost-savings, PAM and TIG process are the kind of significant for AM

Corresponding author at: Shanghai Key Laboratory of Materials Laser Processing and Modification, Shanghai Jiao Tong University, Shanghai 200240, China. E-mail address: [email protected] (J. Lin).

http://dx.doi.org/10.1016/j.jmbbm.2016.12.015 Received 9 June 2016; Received in revised form 26 November 2016; Accepted 20 December 2016 Available online 21 December 2016 1751-6161/ © 2016 Elsevier Ltd. All rights reserved.

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

between deposited layers can help keep heat balanced, so that the molten pool size is controlled. Using AM technologies to fabricate components will experience complex cyclic thermal history, thus there is necessity to understand the relationship of microstructure, processing and properties. However, continuous PAM process (CPAM) is rarely reported before, in especial, the effect of microstructural evolution on mechanical property remains unclear for Ti-6Al-4V alloy deposited by the CPAM. In this research, several Ti-6Al-4V thin walls were deposited by optimized weld wire-feed CPAM, in which the heat input was gradually decreased layer by layer. The microstructural evolution and mechanical properties of Ti-6Al-4V wall deposited by wire-feed CPAM has been studied. The interactive effects between microstructural evolution, deposition process and mechanical properties should be clarified for the deposited thin wall without cooling substrate. It is expected that the mechanical properties of directly deposited walls at room temperature to be equal to or to exceed the standard level of forging. Furthermore, the results of the CPAM process can be added in development of an AM database of materials processing.

Fig. 1. Schematic representation of the plasma arc deposited process.

Table 1 Summary of reported advantages of PAM compared with other technologies. AM technology

β grains

Cost

Utilization ratio

Precision

Laser beam

0.5–1.5 mm (0) 1.4–5.0 mm (－) 15–35 mm (－ －) (Maybe +, 0, or －)

$500,000 (－)

(－ －)

(+)

$1,000,000(－ －) < $7000 (+)

(－ －)

(+)

(+)

(－ －)

$7000 (0)

(+)

(0)/ (+)

Electron beam TIG Plasma arc

2. Experimental procedure 2.1. Experimental material and equipment A 1.0 mm diameter Ti-6Al-4V wire (ERTi-5) was used in deposition process, and namely chemical composition was 0.02 C, 0.14 O, 0.01 N, 0.007 H, 0.0 7 Fe, 6.11 Al, 3.95 V, Bal.Ti (in weight %), which is below the maximum content recommended by ASTM F1108−04 (2009). The substrate was 2000 × 1500×8 mm (length×width×height) hot rolled plate. The plates were grounded by 150 to 400 grit SiC papers and then degreased by acetone and ethanol before being used. A TransTig 4000 Job G/F and A Plasma module 10 were used as the power supply with a KD7000 D-11 wire feeder. The plasma arc welding torch was attached to a 6-axis KUKA robot linked to a 2-axis table. The PPAM process was carried out under inter gas (argon) shield.

(+)=good, (0) = neutral, (－) negative, and (－－) more negative.

technology. PAM process has advantages of high efficiency for the manufacture of high-cost structural components, such as disks and blades of aircraft gas turbine, which are normally produced by costly titanium alloys. These components can be fabricated by using one beam only of PAM process. In addition, the temperature zone of plasma arc transferred to the work piece in the concentrated beam can reach about 10000–16000 K (Aiyiti et al., 2006), so the energy density of plasma arc is much higher than gas tungsten arc welding and close to the lower limit of laser, which helps deposit in better precision compared to TIG process. Fig. 1 shows schematic representation of the plasma arc deposited process. Moreover, the price of plasma arc power source decreases to 1/7 of that of Laser Beam or 1/15 of Electron Beam (Stavinoha, 2012), as is shown in Table 1 (Stavinoha, 2012; Antonysamy, 2012; Martina et al., 2012). Table 1 gives a summary of reported advantages of PAM in comparison with other technologies. From the data of measured length of columnar β grains in samples deposited by AM technologies, it clearly shows that LBM is the best deposition process, and EBM is much better than TIG, and the PAM process needs further investigation. In PAM process, wire or powder is fed into the melt pool produced by constant plasma arc, pulsed plasma arc, or variable polarity plasma arc which is usually used in welding aluminum alloys in order to wipe off the oxide (Jiang et al., 2013). In addition, previous researches have confirmed that it can refine the structure by pulsed plasma arc welding, of which the reason is that the process enables the melt pool to agitate more tempestuously during welding (Correa et al., 2008). However, using constant parameters deposited by the pulsed plasma arc process needs to set more parameters, which increases the complexity in controlling and cooling time. Additionally, the heat input value of each deposited layer will easily result in heat accumulation (Stavinoha., 2012; Xu et al., 2013; Lin et al., 2016). Moreover, the first lot of layers were excluded because of the thermal effect of the base plate after finish deposited part (Martina et al., 2012), which will spent time, waste material and impel anisotropy. Hence, decreasing the heat input

2.2. Research methodology After 16-layer deposition, the total height with one bead width is shown in Fig. 2. In each specific trial, Ti-6Al-4V wire was deposited layer by layer onto the substrate by CPAM. The scanning direction of each layer was set X-coordinate, moreover, each additive height (ΔZ) was set at 1.5 mm in positive Z-coordinate direction. To prevent air oxidation, heat input of bead was gradually decreased layer by layer, and each layer would be deposited only if the temperature of the previous layers fell below 300 ℃. Thermocouple probes were inserted into the substrate, of which the distance from the deposited surface was 5 mm. The main processing parameters were constant: Travel Speed(Ts) was 0.25 m/min, Wire Feed Speed(WFS) was 3.5 m/min, Plasma gas was 0.2 L/min, and Argon atmosphere was

Fig. 2. Photograph of CPAM process deposited components.

20

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

Table 2 Details of parameters of each layer deposited by CPAM process. Layers

L1 L2 L3 L4-L6 L7-L10 L11-L16

Parameters I(A)

U(V)

250 230 210 190 170 150

21.7 21.5 20.3 19.3 19.1 18.2

20–30 L/min. The parameters of each layer and other experimental settings are detailed in Table 2. According to ASTM E8-04, the tensile coupons were extracted in deposited direction, i.e. X-coordinate. The dynamic strain gauge extensometer with a gage length of 25 mm were set up to measure 0.2 pct yield stress (Rp0.2) and ultimate tensile stress(UTS) with a displacement rate of 0.5 mm/min, being attached to each specimen. All the tensile property tests were carried out by an AG-25KNIS machine. The tested Vickers hardness coupons were incised from the formed thin wall, 20 mm away from the starting point of deposition, and were along the additive direction. The coupons of microstructure were obtained in the same way, as is shown in Fig. 2. Those microstructure analyzing coupons were polished with SiC papers (150,400,600,800,1000grit) by a BUEHLER machine grinding and polishing, then they were electrolytic polished in ethanol (purity 95%) 94%+ perchloric acid (purity 60%) 6% for about 3–5 min at a low temperature. After that, they were etched for 40–60 s in kroll reagent (1 ml HF+2 ml HNO3+50 ml H2O) in order to emerge a flesh condition for observation. According to DIN EN 6507 on cross-sections of additive direction with 100 g (HV0.1) on Tukon 1102(Wilson, BUEHLER), the Vickers hardness of different microstructures were measured in three locations. Microstructure and fracture surfaces of selected typical specimen were examined by a FEI Nova NanoSEM50 scanning electron microscope (SEM) equipped with an OXFORD XMax energy dispersive X-ray (EDS). Microstructural date were collected by Electron Back-Scattered Diffraction (EBSD) and analyzed by TSL OIM Collection5. In addition, sizes of prior beta grains were measured by the linear intercept method introduced in ASTM E11296: Standard Test Methods for Determining Average Grain Size, 2004. For the length of the beta grains, each grain were measured for three times along longitudinal direction, and taking the average of total beta grains.

Fig. 3. The profile of Ti-6Al-4V wall deposited by CPAM shows columnar prior-β grains and morphology of layer band region.

Table 3 Measurement for macrostructure of wall deposited by CPAM. Feature

Average ± Standard Deviation

Min

Max

Between distance the LBs Width of β grains Length of β grains

1.47 ± 0.35 mm

0.67 mm

1.93 mm

2.37 ± 0.91 mm 3.44 ± 1.11 mm

1.00 mm 1.78 mm

3.60 mm 4.81 mm

3.2. Microstructure evolution According to the sizes and arrangements of the two phases of hexagonal close-packed structure (hcp) (α) and body-centered cubic (bcc) (β), the microstructure of Ti-6Al-4V can be described as equiaxed structure, martensite α′, widmanstätten structure, and “basket-weave” structure of widmanstätten α plates in the β matrix (Stefanescu et al., 2004). Moreover, the shape of formed α phase are described as primary α, secondary α, acicular α, lamellar, massive α, and colony α. Different microstructures described as fine or coarse are mostly related to heat treatment process. Martensite α′ structure and fine basket-weave structure can be discerned by OM. Usually, martensite α′ structure is characterized by a rectangular grid structure (Brandl et al., 2012), while fine basket weave structure is described that a new α plates nucleate in α plate that tend to grow nearly perpendicularly to near plate (Baufeld et al., 2011). 3.2.1. Microstructure evolution of layer band region and EDS measurement Fig. 4 shows SEM of LB in the middle region of a deposited wall (Fig. 4a), Points 1–4 and Points 5–8 respectively illustrate two different microstructures, which are nominal martensite α′ structure and basket weave structure. Both microstructures have formed full α lamellar structure in SEM observation (Fig. 4b, c and d), including widmanstätten (0.27–0.84 μm) and coarse basket-weave (0.30– 1.15 μm). And in the transformed-β matrix, α rods and α dots are observed. Moreover, the parallel align acicular α of basket weave of widmanstätten are perpendicular to the prior β-grain boundary (GB), which transferred into colony α. Conversely, parallel align acicular α of nominal martensite α′ structure grows along the GB. As shown in Fig. 4, comparing to the prior β-GB with LB (Fig. 4b) and without LB region (Fig. 4c), we can conclude that the grain boundary of with LBs are coarser. Besides, α colonies with largerlength lamellar and bigger α parts (e.g. Fig. 4d) are observed. The EDS results of LBs in middle wall are shown in Table 4. The contents of main elements of Al and V of Ti-6Al-4V are in the ASTM-B 367-13 ranges (ASTM B367-13., 2013). The average volume of Al element in nominal martensite α′ region exceeds that in basket weave

3. Results 3.1. Macrostructure Fig. 3 shows the profile of Ti-6Al-4V wall in 16-layer height and one-bead width deposited by CPAM. The prior β grains with different columnar sizes grow perpendicularly from substrate to the top region (Fig. 3), acrossing multiple deposition layers. Meanwhile, the epitaxial growth of prior columnar β-grain has different growth directions. The average length of columnar gain is about 3.44 mm, and the maximum length is about 4.81 mm in the top region while the minimum is about 1.78 mm in the bottom region, as shown in Table 3. The aspect ratio of β-grain is about 1.45, which is near equiaxed dendrite. As revealed in Fig. 3, layer bands (LBs) are parallel to deposited layer. Moreover, compared to other regions, the LBs are made up of fine white strips and back straps under optical microscope (OM) observation. The distance of two nearby parallel LBs is about 1.36 ± 0.45 mm (Table 3). However, LBs are not observed in the top of wall cross 8.1 mm height. Similar structures of LBs have been reported in researches of ALM by laser (Kelly and Kampe, 2004). 21

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

Fig. 4. Microstructure of layer band region in the middle region of the deposited wall, (a) optical microscopy, the points (1–8), dots and rods indicate different place of the EDS microanalysis; (b), (c) and (d) SEM photos showing two different microstructure, respectively indicated prior β-grain boundary without layer band and within layer band region.

observed. The different microstructures are divided by prior β-grain boundaries (GBs). Furthermore, acicular α and colony α are displayed in different morphologies in internal grains or at prior β-GBs. With additive layers deposited by CPAM, martensite α′ structure turns into dim which is generally termed as acicular α in a matrix of transformed β (Gammon et al., 2004), and coarse basket weave can be observed in middle region (Fig. 5b). All of microstructures have formed full α lamellar structure, primary α with nano-dispersolds are observed by SEM (Fig. 5), and more α dots and rods are observed in the middle and top wall. Additionally, colony α is presented at prior β-GB or in matrix of transformed β. Furthermore, prior β-GBs in the middle or top regions are observed slightly coarser than those in bottom region. Orientation map shows the orientations relation between different α lamellae with different colors and different orientations. Generally, parallel neighbor lamellae seem to fuse together and form a larger region (or α colonies), which are called relatively large grains in the orientation map. As shown in Fig. 6, the Y–Z plane in the middle region represents prior β grains at triple point, while Fig. 7 shows the internal prior β gain area in the middle region. Prior β grains in different orientation are made up of martensite α′ structure and basket weave structure, which can be evaluated by different microstructures (Fig. 6a). The distribution of misorientation observed in Figs. 6a1 and 7a1 tends to be about 60°. The internal grains of martensite α′ with long streaks of lamellae are finer than the triple prior β grains (Figs. 6b1 and 7b1). Frequently, basket weave structures have larger fraction volume of β phase than the martensite α′ (Figs. 6c and 7c). As shown in Fig. 6d, the transformation of α and β phase fit the Burgers orientation relationship, i.e. < 0001 > // < 110 > .It is noteworthy that the main < 11–20 > poles split into triangular stars or rhombic patterns in Fig. 7d, while

Table 4 The EDS measurement of LBs in middle wall and summary of the width of α plates. Spectrums

Weight of elements /%

Microstructure

Width of α lamellar 0.27– 0.84 μm 0.31– 0.74 μm －－ －－ 0.36– 1.15 μm 0.30– 0.91 μm －－ －－ －－

Al

V

Ti, et al.

1

6.10

4.76(max)

Other

2

5.94

4.05

Widmanstätten (nominal martensite α′)

3 4 5

6.11 6.00 6.25(max)

3.85 3.85 4.34

Basket weave

6

5.90

3.94

7 8 ASTM-B 367

5.73(min) 5.94 5.5–6.75%

3.85 4.14 3.5–4.5%

region, and the area containing the maximum amount of Al is in the upper LB (point 5 in Fig. 4a). The average volume of V element in norminal martensite α′ region are lower than that in basket weave region, and the area with the maximum V amount is in the upper LB (shown the point 1 in Fig. 4a) as well.

3.2.2. Evolution of normal microstructure compared to layer band and EBSD analysis Fig. 5 shows optical microstructure of deposited wall, both martensite α′ structure and basket weave of widmanstätten structure are 22

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

Fig. 5. Optical micrograph of normal microstructure compared to layer band of the wall from (a) the bottom region; (b) the middle region; (c) the top region. (a1), (a2), (b1), (b2), (c1) and (c2) SEM photos showing two different microstructure of the wall from bottom to top under same scale.

Fig. 8b shows that in the as-built condition, the average hardness values in different regions of LBs of the whole wall are not obviously different, except that the average hardness value of the first LB is lower, of which the reason is the dilution of pure Ti due to high heat input in initial stage when deposited by CPAM. The average microhardness value almost matches the hardness level of ASTM B367-13. Specially, the microhardness value within LB (Fig. 8c, marked by single arrows) is slightly lower than that in the upper LB and the lower LB.

the < 0001 > pole directions are not observed to be the same patterns. 3.3. Mechanical properties 3.3.1. Microhardness Fig. 8 shows schematic illustration of hardness measurement on the profile of wall. Three typical regions are selected for hardness measurement, which are the top region, the middle region and the bottom region. All the selected regions have different characteristics of microstructure, including the LB microstructure, the upper LB microstructure, and the lower LB microstructure. Moreover, three test points were measured in each region.

3.3.2. Tensile properties Typical tensile properties of the wall are presented in Table 5. The average yield strength (YS) of tested coupons is 877 ± 18.5 MPa, and

Fig. 6. The EBSD of grain boundary at triple microstructure from the middle region of the wall.(a)orientation map with inverse pole figure; (a1) the distribution of misorientation angle; (b)unique grain color map; (b1) grain size distribution; (c) distributed phase map; (d) pole figure.

23

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

Fig. 7. The EBSD of grain boundary at triple microstructure from the middle region of the wall.(a)orientation map with inverse pole figure; (a1) the distribution of misorientation angle; (b)unique grain color map; (b1) grain size distribution; (c) distributed phase map; (d) pole figure.

4. Discussion

ultimate tensile strength (UTS) is 968 ± 12.6 MPa, and elongation is 11.5% ± 0.5, respectively. The mechanical properties are in a higher level than that for wrought in ASTM B381-13 (ASTM B381-13., 2013). Without heat processing on condition, the wall deposited by CPAM have good tensile properties compared to typical AM technologies (Table 5). Moreover, the strain of failure can match some others typical AM technologies via stress-relieved or annealed conditions. Fig. 9 shows the SEM fractograph of typical tensile test at the middle region. The coupons present typical fractograph of dimple rupture shown in Fig. 9b. The arrows without filled black are used to note the crack growth orientation, and the original fractograph is located from the gas hole noted by arrow filled black (Fig. 9a). It can be seen that the crack growth in the middle is vertical to the fracture surface(i.e. Y direction). Respecting to the shape and depth of the dimples, the factograph of the coupon shows elongated dimples and tear ridges (Fig. 9b, noted with white arrow).

4.1. Evolution of macrostructure and microstructure AM process fabricates a component by means of direct energy of which the heat input will melt the previous deposited layer. Heat input has a significant effect on thermal gradient, nucleation and the growth rate of grains. In the present case, the thin wall deposited by CPAM consists of various morphologies, including epitaxial growth of prior β grains, horizontal layer bands, martensite and basket-weave microstructure of widmanstätten. 4.1.1. Macrostructure Prior β-grain varies in shape, perpendicularly growing up across multiple deposited layers. And the aspect ratio of β-grain is about 1.45, which is near equiaxed. In contrast, the perpendicular strips of columnar prior β grains observed in the Laser additive layer manu-

Fig. 8. Schematic illustration of hardness measurement on the profile of the wall (a); microhardness of layer bands (bottom, middle and top region) of the wall (b): the zone within LB were marked by single arrows; ASTM B367-13 was marked by dashed line.

24

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

Table 5 Mechanical properties of the walls deposited by CPAM vs. other typical AM technologies. Part

Condition

Deposited direction

UTS(MPa)

0.2%YS(MPa)

%EL

Average ± Standard deviation CPAM

As-built

H

877 ± 18.5

968 ± 12.6

11.5 ± 0.5

ASTM B381-13 Pulsed-PAM (Lin et al., 2016) EBM (Edwards et al., 2013) EBM (Edwards et al., 2013) EBM (Murr et al., 2009) LBM (Brandl et al., 2010) LBM (Brandl et al., 2011) TIG (Brandl et al., 2010)

Wrought As-built As-built As-built 620–690 ℃ As-built/600 ℃/840 ℃ As-built As-built 600 ℃/840 ℃

－ H H V － H or V H or V H or V

895 988 ± 19.2 833 ± 22 851 ± 19 1180 872–940 790 960 930–940

828 909 ± 13.6 783 ± 15 812 ± 12 －－ 791–874 697–884 791–874

10 7 ± 0.5 2.7 ± 0.4 3.6 ± 0.9 16–25% 4.2–12.5 5–12 6.6–20.5

H means horizontal;V means vertical.

higher than that of gas tungsten arc welding. So CPAM process can not only provide a satisfying precision of large-scale structural components, but also can effectively weaken the aspect ratio of prior β columnar grain. On the other hand, the epitaxial growth of prior β-grains is in direction identical with additive height (ΔZ), because strong cooling gradient leads to the preferred growth or easy-growth direction (Karimzadeh and Ebnonnasir, 2006; Carroll et al., 2015). Especially, the grains of additive manufactured thin-wall components tend to grow perpendicularly to the prior layer. With additive layers deposited, the epitaxial growth of the prior β-grain is preferential in the direction of the maximum temperature gradient, so there is competitive relationship among the different growing orientations of prior β-grain. As

facturing (LAM) (Wu et al., 2004) are much larger. There are two primary reasons that it has reduced aspect ratio of β-grain compared with the morphology of thin component deposited by LAM and EBM (Wang et al., 2015; Galarraga et al., 2016). The first is that Ti-6Al-4V has a very narrow freezing range about 5 K (Lin et al., 2007), resulting in epitaxial nucleation from prior layers and freezes quickly, especially. As both of the energy density of electron beam and laser beam are higher than plasma beam, the stronger cooling rate can promote prior β columnar grain to grow, so it is more likely to form the morphology of narrow grain. The temperature zone of plasma arc transferred to the work piece in the concentrated beam can reach about 10,000–16,000 K (Aiyiti et al., 2006), and the energy density of plasma arc matches the lower limit of laser, and is much

Fig. 9. The SEM fractograph of typical tensile test at room temperature: (a) the middle region; (b) the crack growth region.(c) Typical tensile test curves: ASTM F1108-04 for surgical implants and ASTM B381 for forge were signed by dashed line.

25

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

in distance of about 5 mm from the deposited layer. Other settings are k=27 J/msK, and v=0.25 m/min, the β-transus of T=995 ℃, and Q=η*Iav*U, where η=0.8 (Kou, 2002). Results show that the cooling rate is 260℃/s at the bottom region, and 257 ℃/s at the top region. Cooling rate above 410 °C/s can generate a full martensitic microstructure, and the mixture of martensitic microstructure and massive transformation can be observed under the cooling rates between 410 and 20 °C/s. This transformation is gradually replaced by diffusion controlled widmanstätten α formation with decreasing cooling rate (Ahmed and Rack, 1998). The accumulation of heat input promotes to reduce the cooling rate. So this theoretical calculation can explain why that the reason that the microstructure consists of martensite and basket-weave microstructure of widmanstätten. With layers deposited, martensite α′ and matrix of transformed β decomposes to α rods and dots or α colony under the multiple thermal cycles (Fig. 4). The grain boundaries have high energy so that there tend to form large amount of secondary α compare to internal gains (Fig. 4b and c), moreover, the horizontal LBs observed are α colony (Fig. 4d). The whole wall observed in Fig. 5 includes the bottom region, middle region and top region, and there are α rods and dots, which are supposed to be different morphologies of secondary α. According to different locations, precipitated secondary α platelets with nanodispersoids sized 60–120 nm are observed in the whole wall. With continued multiple deposition by CPAM, nano-dispersoids grow up by lamellar, and the formation can be illustrated in Fig. 10. The horizontal LBs can be recognized in the wall deposited by CPAM, but they are not observed in the top three layers. Similar phenomena have been declared in researches of ALM by laser (Kelly and Kampe, 2004; Wu et al., 2004). So the LB is relevant to the effect of the later deposition. In essence, the cause of the occurrence of LBs is related to the multiple thermal cycles and gradual cooling. Firstly, the EDS measurement of LBs shows that the microstructure of components matches the standard of ASTM-B 367-13 (Table 4), i.e. the formed LBs are not related to components of segregation. Secondly, the LBs have no effect on the direction of prior β-GB resulted from inaction of heterogeneous nucleation and growth. On the other hand, plasma arc has higher heat input compared to LAM (Wu et al., 2004).The energy density of plasma arc is between the values of laser and gas tungsten arc, and the plasma arc is transferred to the work piece as a concentrated beam at the temperature of about 10,000– 16,000K, which is near the energy density of laser (Aiyiti et al., 2006), so plasma arc can increase the speed of atom diffusion. Respectively, the LBs do not coincide with deposited layers because of repeated heat input across the β/α+β-transus line (995℃) (Peters and Christoph, 2003). The martensite α′ may turn into α+β structure (Fig. 4) by subjecting these regions to elevated temperatures within the α+β phase field. The multiple thermal cycles promote to decomposition of martensite α′ and growth of α, and even to form α colony (Fig. 4d), moreover, the CPAM process holds the equilibrium of conductive and convective heat flow through decreasing the heat input step by step. Furthermore, the last layers cover about 8mm from the last LB, and it can be considered that insufficient thermal cycles lead to the absence of LBs in the top of thin wall.

Table 6 Summary of heat input of each layer deposited by CPAM. Layers

L1 L2 L3 L4-6 L7-10 L11-L16

Parameters of PPAM Iav(A)

U(V)

Pa(KW)

P/vb (KJ/cm)

250 230 210 190 170 150

21.7 21.5 20.3 19.3 19.1 18.2

5.42 4.94 4.26 3.66 3.24 2.73

13.02 11.86 10.23 8.80 7.79 6.55

v means deposited speed. a Power of Plasma Arc. b Rate of heat input.

shown in Table 6, there is a summary of heat input of each layer deposited by CPAM, several Ti-6Al-4V thin walls were deposited by optimized weld wire-feed CPAM process. The heat input was about 1.2 KJ/cm gradually decreased layer by layer. The heat input decreases step by step can effectively reduce the energy accumulation of melting layer in deposition process, so that the nucleation can easily occur in different directions of dendrite growth. Hence the CPAM process can hold the equilibrium of conductive and convective heat flow. Based on this two factors, the aspect ratio of prior β grains was controlled in CPAM process (as shown in Fig. 3). That means the CPAM process without post heat treatment can substantially reduce the anisotropy in the deposited components. 4.1.2. Evolution of microstructure and layer band The thin wall deposited by CPAM have formed full α lamellar structure with nano-dispersolds in primary α observed in Fig. 5, and more α dots and rods are observed in the middle and top wall. Additionally, colony α presented slightly coarser and clearer at prior β-GBs and within LBs region (Figs. 4 and 5). Those microstructures depend on the heat input, multiple thermal cycles and gradual cooling rate in the deposition process, and the cooling rate of deposited layer can be calculated by using the Rosenthal (Rosenthal, 1946) Eq. (1).

2π (T −T0)kR − V (R − x ) = exp Q 2α

(1)

Where T is liquid line, T0 is substrate or prior deposited layer, Q is heat input, v is travel speed, k is conductivity, R is the radius of T, x is reference point, α is thermal diffusivity rate. In order to calculate the z of cooling rate, it is assumed that R equals X equals Z. The cooling rate is calculated by Eq. (2).

⎛ ∂T ⎞ ⎛ ∂T ⎞ ⎛ ∂T ⎞ (T −T0) 2 ⎜ ⎟ = ⎜ ⎟ ⎜ ⎟ = −2π kv ⎝ ∂t ⎠ z ⎝ ∂z ⎠t ⎝ ∂t ⎠ z Q

(2)

The bottom temperament (T0) is assumed at 25 ℃, the deposited process could be cut off until the temperament of the previous layer had fallen to T0 which equals 150 ℃, and the temperament were detected by thermocouple probes, which were inserted in the substrate

Fig. 10. Schematic illustration of the evolution of lamellar growth after multiple deposition by CPMA.

26

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

Fig. 11. Schematic showing the evolution of microstructures of thin coupons produced by CPAM: the heat flow direction are indicated by arrow, and compared to different region, (+) = more, (0) = neutral, and (－) less.

Fig. 12. α and β phase orientation map of pole figures and recalculated pole figures as in Fig. 7.

Fig. 13. Schematic stereographic projection in typical four cases of hexagonal system as in Fig. 7.

noteworthy that the main < 11–20 > poles split into triangular stars or rhombic patterns in Fig. 7d, while the < 0001 > pole directions are not observed to be the same patterns. The recalculated pole figures help to understand the general scope of pattern shown in Fig. 12, both < 0001 > and < 11–20 > show fourfold symmetry and slightly off-centered. According to Burgers orientation relationship, a bcc crystal can transform to 12 hexagonal variants, having different orientations with regard to the parent β crystal, i.e. any of the 6{110} planes can transfer into a {0001} basal plane. Moreover, {0001} basal plane of two variants orientations is possible. Frequently, only three different < 0001 > poles with one variant of the two possible < 11–20 > orientations dominate (Baufeld et al., 2010a). Therefore, The transformation of β → α follows the burgers orientation relationship (BOR). Naturally, the BOR also applies for the < 11–20 > pattern. i.e. one of three < 11–20 > directions in a basal plane{0001} would be parallel to a < 111 > direction of

Based on above analysis, the evolution of microstructure and layer band with nano-dispersoids observed in the whole wall deposited by CPAM are caused by thermal cycle and cooling rate, which can be shown in Fig. 11. 4.2. Spatial orientation distributions Deposited components by different AM process can occur different phase transformation, the transformation from β phase to α phase is very important in α–β Ti alloys, as it greatly influences the final microstructure. In this study, the phase transformation of the wall deposited by CPAM follows the burgers orientation relationship (Fig. 6d), i.e. {0001}//{110}and < 11–20 > // < 111 > . To avoid the interference between prior β-GB, the EBSD analysis of the internal grain is used for more accuracy, as is shown in Figs. 12 and 13. It is 27

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

the original {110} plane. Moreover, two different < 1–11 > directions lie in one original {110} plane and both can be the basis of the transformation. Consequently, for every basal plane, two variants of orientations are possible. e.g. two sets < 11–20 > reflections are expected, and each set will have three reflections, which lead to different < 11–20 > poles forming three dominating orientations. According to different color in orientation map (Fig. 13), the different < 11–20 > poles in a same original {110} plane forming triangular stars or rhombic patterns are shown in Fig. 13. It means that there are three dominating orientations out of twelve possibilities, as observed in Fig. 7d.

ship, and three orientations out of the twelve possibilities occupy dominating orientations, which finally form triangular stars or rhombic patterns. (4) The microstructure of α lamellar with nano-dispersoids significantly improves mechanical properties, moreover, the mechanical properties exceed the standard level of forging.

4.3. The effect of microstructure on mechanical properties

The research is performed with financial assistance from the Military Plan Projects of China, No. 613213.

Furthermore, the result of the components deposited by CPAM process can enrich AM database of materials processing. Acknowledgments

Deposited wall by CPAM process experiences multiple thermal cycles that include melting and solidifying, and the thermal cycles promote the decomposition of α′ → α + β . In different locations of specimen (Fig. 5), different morphologies of precipitated secondary α platelets with nano-dispersoids are observed in whole (Fig. 5). With continued multiple deposition by CPAM, near nano-dispersoids grow to lamellar. With the increase of amount, α lamellar grows along with grain boundary or band layers, becoming α colony which leads to the increase of size of actual slip under the effect of strain. Specifically, the width of α colony can be about 3μm within LB (Fig. 4d). The macrostructure, thus, possibly exits actual slip due to the larger α colony under tensile test, causing the elongated dimples and tear ridges in factograph shown in Fig. 9, because nano-dispersoids enhance hardness and strength, and larger colony α adverse to mechanical properties (Lujering and Williams, 2007; Palanivel et al., 2016).So the hardness of internal LB is lower than that in other locations. Moreover, the hardness tendency at the LB region is in similar value, which results from the equilibrium of conductive and convective heat flow through decreasing the heat input step by step. In bottom region, the hardness value declines most because the heat input of prior number layers in bottom is larger than that of the later deposited layers, which leads to the increase of the dilution rate. The wall deposited by CPAM normally has high tensile strength and good elongation shown in Table 5. There are two reasons: 1) The small deviation of the average heat input restrains the growth of prior β grain and α lamellar, and refined microstructure always exhibits ductile facture with a dimple fractograph according to the Hall-Petch relation. 2) The contribution of α lamellar with nano-dispersoids to the tensile strength is more significant than that of the prior β grains (Baufeld et al., 2010b; Lu et al., 2016), and the summary of α lamellar is shown in Table 4. Avoiding air pore in the wall deposited by CPAM is significant for further enhancing mechanical properties. Specially, the α colonies grow along with the prior β-grain boundaries or band layers, which leads to the occurrence of elongated dimples or tear ridges in factograph (Fig. 9b).

References Ahmed, T., Rack, H.J., 1998. Phase transformations during cooling in α+β titanium alloys. Mater. Sci. Eng. A. A243, 206–211. Aiyiti, W., Zhao, W., Lu, B., et al., 2006. Investigation of the overlapping parameters of MPAW-based rapid prototyping. Rapid Prototyp. J. 12 (3), 165–172. Antonysamy, A.A., 2012. Microstructure, Texture and Mechanical Property Evolution During Additive Manufacturing of Ti6Al4V Alloy for Aerospace Applications. University of Manchester, Manchester, UK. ASTM B367-13., 2013. Standard Specification for Titanium and Titanium Alloy Castings, ASTM International. ASTM B381-13., 2013. Standard Specification for Titanium and Titanium Alloy Forgings, ASTM International. ASTM F1108-04., 2009. Standard Specification for Titanium-6Aluminum-4Vanadium Alloy Castings for Surgical Implants (UNS R56406), ASTM International. Banerjee, D., Williams, J.C., 2013. Perspectives on titanium science and technology. Acta Mater. 61, 844–879. Baufeld, B., Brandl, E., Van der Biest, O., 2011. Wire based additive layer manufacturing: comparison of microstructure and mechanical properties of Ti-6Al-4V components fabricated by laser-beam deposition and shaped metal deposition. J. Mater. Proc. Technol. 211, 1146–1158. Baufeld, B., Van der Biest, O., Gault, R., 2010b. Additive manufacturing of Ti–6Al–4V components by shaped metal deposition: microstructure and mechanical properties. Mater. Des. 31, 106–111. Baufeld, B., Van Der Biest, O., Dillien, Steven, 2010a. Texture and crystal orientation in Ti-6Al-4V builds fabricated by shaped metal deposition. Metall. Mater. Trans. A 41A, 1917–1927. Brandl, E., Baufeld, B., Leyens, C., et al., 2010. Additive manufactured Ti-6Al-4V using welding wire: comparison of laser and arc beam deposition and evaluation with respect to aerospace material specifications. Phys. Procedia 5, 595–606. Brandl, E., Palm, F., Michailov, V., et al., 2011. Mechanical properties of additive manufactured titanium (Ti–6Al–4V) blocks deposited by a solid-state laser and wire. Mater. Des. 32 (10), 4665–4675. Brandl, E., Schoberth, A., Leyens, C., 2012. Morphology, microstructure, and hardness of titanium (Ti-6Al-4V) blocks deposited by wire-feed additive layer manufacturing (ALM). Mater. Sci. Eng. A 532, 295–307. Brooks, H., Molony, S., 2016. Design and evaluation of additively manufactured parts with three dimensional continuous fibre reinforcement. Mater. Des. 90, 276–283. Carroll, B.E., Palmer, T.A., Beese, A.M., 2015. Anisotropic tensile behavior of Ti–6Al–4V components fabricated with directed energy deposition additive manufacturing. Acta Mater. 87, 307–320. Correa, E.O., Costa, S.C., Santos, J.N., et al., 2008. Weldability of iron-based powder metal materials using pulsed plasma arc welding process. J. Mater. Process. Technol. 198, 323–329. Davim, J.P., 2014. Machining of titanium. In: Niknam, S.A., Khettabi, R., Songmene, V. (Eds.), Machinability and Machining of Titanium Alloys: A Review. Springer, Berlin, Berlin. Ding, D., Pan, Z., Cuiuri, D., et al., 2015. Wire-feed additive manufacturing of metal components technologies,developments and future interests. Int. J. Adv. Manuf. Technol. 81, 465–481. Donachie, Matthew J., 2000. Titanium: A Technical Guide second ed.. ASM International, USA. Edwards, P., O'Conner, A., Ramulu, M., 2013. Electron beam additive manufacturing of titanium components: properties and performance. J. Manuf. Sci. Eng. 135 (6), 061016. Galarraga, H., Lados, D.A., Dehoff, R.R., et al., 2016. Effects of the microstructure and porosity on properties of Ti-6Al-4V ELI alloy fabricated by electron beam melting (EBM). Addit. Manuf. 10, 47–57. Gammon, L.M., Briggs, R.D., Packard, J.M., et al., 2004. Metallography and microstructures of titanium and its alloys. In: Voort, G.F., James, H., Mills, M.K. (Eds.), ASM Handbook, Metallography and Microstructures 9. ASM International, 899–917. Jiang, Y., Liu, M., Lu, Y.H., et al., 2013. Experimental analysis of welding parameters on variable polarity plasma arc pressure. Adv. Mater. Res. 651, 355–360. Karimzadeh, F., Ebnonnasir, A., 2006. A. Foroughi, Artificial neural network modeling

5. Conclusions The Ti-6Al-4V thin wall deposited by the advanced weld wire-feed CPAM process without cooling substrate exhibits excellent mechanical properties. The following conclusions can be obtained from this research. (1) The aspect ratio of prior β grains are affected by the gradually reducing heat input of the CPAM process. (2) The formation of horizontal LBs results from the influence of multiple thermal cycles in the CPAM process, and insufficient thermal cycles lead to the absence of LBs in the top of thin wall, and the low hardness of LBs is caused by the formation of α colonies. (3) The phase transformation follows the Burgers orientation relation28

Journal of the mechanical behavior of biomedical materials 69 (2017) 19–29

J. Lin et al.

deposition of a β-titanium alloy. Mater. Des. 81, 21–30. Qiu, C., Ravi, G.A., Dance, C., et al., 2015b. Fabrication of large Ti–6Al–4V structures by direct laser deposition. J. Alloy. Compd. 629, 351–361. Rosenthal, D., 1946. The theory of moving sources of heat and its application to metal treatments. Trans. ASME. 68 (8), 849–866. Stavinoha, J.N., 2012. Investigation of Plasma Arc Welding as a Method for the Additive Manufacturing of Titanium-(6) Aluminum-(4) Vanadium Alloy Components. ProQuest LLC, Michigan, USA. Stefanescu, D.M., Ruxanda, R., 2004. Solidification structures of titanium alloys. In: Voort, G.F., James, H., Mills, M.K. (Eds.), ASM Handbook, Metallography and Microstructures 9. ASM International, Materials Park, OH, 116–126. Szost, B.A., Terzi, S., Martina, F., et al., 2016. A comparative study of additive manufacturing techniques: residual stress and microstructural analysis of CLAD and WAAM printed Ti-6Al-4V components. Mater. Des. 89, 559–567. Tan, X., Kok, Y., Tan, Y.J., et al., 2015. An experimental and simulation study on build thickness dependent microstructure for electron beam melted Ti–6Al–4V. J. Alloy. Compd. 646, 303–309. Wang, T., Zhu, Y.Y., Zhang, S.Q., et al., 2015. Grain morphology evolution behavior of titanium alloy components during laser melting deposition additive manufacturing. J. Alloy. Compd. 632, 505–513. Wu, X., Liang, J., Mei, J., et al., 2004. Microstructures of laser-deposited Ti-6Al-4V. Mater. Des. 25, 137–144. Xu, F.J., Lv, Y.H., Xu, B.S., et al., 2013. Effect of deposition strategy on the microstructure and mechanical properties of Inconel 625 superalloy fabricated by pulsed plasma arc deposition. Mater. Des. 45, 446–455. Zhao, X., Li, S., Zhang, M., et al., 2016. Comparison of the microstructures and mechanical properties of Ti–6Al–4V fabricated by selective laser melting and electron beam melting. Mater. Des. 95, 21–31. Zhu, Y., Tian, X., Li, J., et al., 2015. The anisotropy of laser melting deposition additive manufacturing Ti-6-5Al-3-5Mo-1-5Zr-0-3Si-titanium-alloy. Mater. Des. 67, 538–542.

for evaluating of epitaxial growth of Ti6Al4V weldment. Mater. Sci. Eng. A 432 (1), 184–190. Kelly, S.M., Kampe, S.L., 2004. Microstructural evolution in laser-deposited multilayer Ti-6Al-4V builds: part II. Thermal modeling. Metall. Mater. Trans. A. 35 (6), 1869–1879. Kou, S., 2002. Welding Metallurgy second ed.. John Wiley & Sons, New Jersey, USA. Lin, J.J., lv, Y.H., Liu, Y.X., et al., 2016. Microstructural evolution and mechanical properties of Ti-6Al-4V wall deposited by pulsed plasma arc additive manufacturing. Mater. Des. 102, 30–40. Lin, X., Yue, T.M., Yang, H.O., et al., 2007. Solidification behavior and evolution of phase in laser rapid forming of graded Ti6Al4V-Rene88DT alloy. Matall. Mater. Tran. A 38, 127–137. Lu, S.L., Qian, M., Tang, H.P., et al., 2016. Massive transformation in Ti-6Al-4V additively manufactured by selective electron beam melting. Acta Mater. 104, 303–311. Lujering, G., Williams, J.C., 2007. Titanium second ed. Springer, New York. Martina, F., Mehnen, J., Williams, S.W., Colegrove, P., Wang, F., 2012. Investigation of the benefits of plasma deposition for the additive layer manufacture of Ti–6Al–4V. Mater. Process. Technol. 212, 1377–1386. Miranda, R.M., Lopes, G., Quintino, L., et al., 2008. Rapid prototyping with high power fiber lasers. Mater. Des. 29 (10), 2072–2075. Murr, L.E., Esquivel, E.V., Quinones, S.A., et al., 2009. Microstructures and mechanical properties of electron beam-rapid manufactured Ti–6Al–4V biomedical prototypes compared to wrought Ti–6Al–4V. Mater. Charact. 0 (2), 96–105. Palanivel, S., Dutt, A.K., Faienson, E.J., et al., 2016. Spatially dependent properties in a laser additive manufactured Ti-6Al-4V component. Mater. Sci. Eng. A. A645, 39–52. Paydas, H., Mertens, A., Carrus, R., et al., 2016. Laser cladding as repair technology for Ti–6Al–4V alloy—influence of building strategy on microstructure and hardness. Mater. Des. 85, 497–510. Peters, M., Christoph, L., 2003. Titanium and Titanium Alloys: fundamentals and Applications. Wiley. Qiu, C., Ravi, G.A., Attallah, M.M., 2015a. Microstructural control during direct laser

29