Friction welding of tungsten heavy alloy with aluminium alloy

Journal of Materials Processing Technology 246 (2017) 42–55

Contents lists available at ScienceDirect

Journal of Materials Processing Technology journal homepage: www.elsevier.com/locate/jmatprotec

Friction welding of tungsten heavy alloy with aluminium alloy c ´ Radosław Winiczenko a,∗ , Olgierd Goroch b , Anna Krzynska , Mieczysław Kaczorowski d a

Warsaw University of Life Sciences, Faculty of Production Engineering, Nowoursynowska 166, 02-787 Warsaw, Poland Warsaw University of Technology, Institute of Mechanics and Printing, Narbutta 85, 02-524 Warsaw, Poland c Warsaw University of Technology, Institute of Manufacturing Technologies, Narbutta 85, 02-524 Warsaw, Poland d Warsaw University of Technology, Department of Mechanics and Weaponry Technology, Narbutta 85, 02-524 Warsaw, Poland b

a r t i c l e

i n f o

Article history: Received 11 August 2016 Received in revised form 10 March 2017 Accepted 11 March 2017 Available online 15 March 2017 Keywords: Friction welding Tungsten heavy alloy Aluminium alloy Microstructure Tensile strength TEM SEM

a b s t r a c t This paper is a study of mechanical properties and microstructure of friction welded coupe of weight heavy alloy (WHA) with aluminium alloy (AA). Scanning electron microscopy (SEM) was used for investigation of the fracture morphology and phase transformations taking place during friction welding process. Chemical compositions of the interfaces of the welded joints were determined by using energy dispersive spectroscopy (EDS). Effects of friction time (FT) and friction pressure (FP) on the ultimate tensile strength (UTS) were studied by plotting graphs. The maximum average strength of 234 MPa, which is 84.78% of the aluminium alloy base material, is achieved at a friction time of 3.5 s and friction pressure of 40 MPa, respectively. The UTS of joints increases with increasing of FP and FT and then decreases after reaching the maximum value, with increase of friction load and time. Microstructure of friction welds consisted of fine equiaxed grains (formed due to dynamic recrystallization) and coarse grains in the periphery region on AA side. A plastic deformation in the direction of burrs is visible mainly on AA side. EDS-SEM scan line analyses across the interface have not confirmed the diffusion of tungsten and nickel to AA side. The nature of the friction welding joints is rather adhesive than diffusive. The EDS point spectrometry showed some enrichment of Ni-Fe matrix with Al atoms close to the joint. Absence of intermetallic phases was found in the weld interface on SEM level observation. The fracture procedes mainly through the cleavage planes at the interface. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Tungsten heavy alloys (THAs) possess a characteristic microstructure consisting of spherical tungsten grains embedded in the matrix being usually Ni based solid solution containing tungsten, iron, cobalt and also other elements sometimes. These alloys are usually fabricated by liquid-phase sintering (LPS) in hydrogen atmosphere. THAs belong to the group of so called weight heavy alloys (WHAs). WHAs have high density (16–18 g/cm3 ), high ductility (10–30%), excellent strength (1000–1700 MPa) and offer good corrosion resistance. They also have a low coefficient of expansion and a high modulus of elasticity (German et al., 2009). Due to a unique combination of physical and mechanical properties, alloy is widely used in vibration dampers (Park et al., 2001), radiation shields (Sunwoo

∗ Corresponding author. E-mail addresses: [email protected] (R. Winiczenko), ´ [email protected] (O. Goroch), [email protected] (A. Krzynska), [email protected] (M. Kaczorowski). http://dx.doi.org/10.1016/j.jmatprotec.2017.03.009 0924-0136/© 2017 Elsevier B.V. All rights reserved.

et al., 2006), mass balance for aerospace (Ryu and Hong, 2003), rocket nozzles in space crafts (Wang et al., 2005) and kinetic energy penetrators – KEP (Cai et al., 1995). Recently, THAs have been specialised as kinetic energy penetrators, replacing conventional depleted uranium (DU) KEP, which is an extremely environmentally unsafe materials (Scapin, 2015). The penetrators are equipped with a soft aluminium alloy ballistic cup which protects the projectile against ricochet when it hits the armour plate, inclined usually at a very small angle with respect to projectile direction. Currently the ballistic cups are joined with main heavy alloy part of the projectile by a threat which is a time consuming and very expensive process. Thus natural is seeking more efficient method for joining the main part of projectile made from THAs with aluminium ballistic cup. One of such techniques is a rotary friction welding (FRW). However, there is a problem resulting from different physical and mechanical properties of the materials to be joined. Friction welding is a high efficiency welding technique applied for joining of similar materials. Furthermore, many friction welding joints, having various mechanical and metallurgical properties, such as: W with Cu, Al with Cu or stainless steel with pure Al, found practical application, as demonstrated

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

43

Fig. 1. Microstructures of base materials: (a) WHA and (b) AA.

by Aritoshi and Okita, (2003). Friction welding is also suitable for materials, which are welded with difficulty. Thus, in earlier works, dissimilar metals, such as stainless steel-titanium (Dey et al., 2009), aluminium-ceramic (Zimmerman et al., 2009), stainless steel-low alloy steel (Arivazhagan et al., 2011), maraging steel-low alloy steel (Reddy and Ramana, 2012), titanium-tungsten pseudoalloy (Ambroziak, 2010), aluminium-low carbon steel (Taban et al., 2010), niobum-tungsten alloy (Ambroziak et al., 2011), stainless steel-copper (Teker, 2013), ductile iron-stainless steel (Winiczenko and Kaczorowski, 2013), and ductile iron-low carbon steel (Winiczenko, 2016) were friction-welded by various researchers. Knowledge available from these works concentrated on structural and mechanical properties and metallurgical phase transformation. The main goal of these investigations was to verify the possibility of friction welding method for tungsten heavy alloy joining. In addition, there is not report in the literature on the application of a solid-state technique for joining of tungsten alloy with aluminium alloy bars. Moreover, we would like to take a closer look at metallurgical phenomena, accompanying the friction welding of these alloys. As such, an EDS line, points and the map spectrometry techniques were used additionally. 2. Experimental procedure 2.1. Materials selection A commercially available AA5454 type of non-heat treatable wrought aluminium alloy and conventional W-Ni-Fe type WHA, with typical 7:3 nickel to iron ratio, were used in experiments. They were machined to a bar of 20 mm in diameter and 100 mm in length. WHA was prepared by mixing an appropriate amount of powders, compacting and then liquid phase sintering (LPS) method. Details of WHA manufacturing are given by Das et al. (2010). Because of powder metallurgy method used for WHA preparation, its microstructure consists of 30–40 ␮m tungsten hard grains embedded in Ni-Fe matrix which, as sintered, is much softer, if compared to tungsten grains. The WHA specimens were ground and polished to get an even surface which was examined using an electron microscopy. The microstructures of base materials are shown in Fig. 1a and b. The tungsten grains are fairly rounded in the Ni-Fe matrix (Fig. 1a). There is also a microstructure visible, characterised by the direct welding between the tungsten grains. Additionally, the welding phase appears to be uniformly distributed. Porosity in the specimen is low and indicative of a successful sintering pro-

Fig. 2. Experimental setup for continuous drive friction welding.

cess, as reported by Das et al. (2011). The pores remaining in the specimen are mostly isolated in tungsten grains and clearly seen in the image. It can be seen in Fig. 1b that aluminium alloy contained a large number of undissolved second-phase intermetallic particles. The chemical compositions and mechanical properties of base materials were given in Tables 1 and 2, respectively. 2.2. Friction welding setup The process of joining was carried out using a continuous drive friction welding machine (ZT4-13, ASPA, Poland) (Fig. 2). The surface for friction welding was prepared on the abrasive cut-off machine. The joined specimens had 20 mm diameter and 100 mm length. As shown in Fig. 2, one workpiece is being rotated and the other is being held stationarily. When an appropriate rotational speed is reached, the workpieces are brought together under axial pressure. Abrasion at the weld interface heats the workpiece locally and axial shortening starts. Finally, the rotation of the workpiece stops and upset pressure is applied to consolidate the joint, as reviewed by American Welding Society (1989). The friction pressure (FP) and friction time (FT) varied within the following ranges: FP = 40–80 MPa and FT = 0.5–9.5 s, respectively,

44

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

Table 1 Chemical composition of base materials. Material

AA WHA

Alloying elements (wt.%) Al

W

Fe

Ni

Mg

Mn

Si

Sb

Zn

Sn

Ti

Bal. –

– Bal.

0.194 2.25

0.002 5.25

2.95

0.265

0.1 –

0.014 –

0.0117 –

0.05 –

0.018 –

Table 2 Mechanical properties of base materials. Material

Tensile strength (MPa)

Yield strength (MPa)

Elongation (%)

Hardness(HV)

Modulus of Elasticity (GPa)

AA WHA

276 960

207 680

22 27

83 285

70 390

Table 3 Welding parameters and tensile strength results. Number specimen No

Friction pressure FP (MPa)

Friction time FT (s)

Tensile strength TS (MPa)

S1 S2 S3 S4 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17

40 72 40 72 80 48 40 72 80 48 40 72 40 72 40 72

0.5 0.5 3.5 3.5 3.5 3.5 4.5 4.5 4.5 4.5 6.5 6.5 7.5 7.5 9.5 9.5

97 38 234 173 167 193 183 146 156 155 216 135 175 149 232 139

Fig. 3. Dimension (in mm) tensile test samples.

while upset pressure (UP), upset time (UT), rotational speed (RS) and the braking time were constant and equalled: 159 MPa, 5 s, 1450 rpm and 0.1 s, respectively. A combination of friction pressure and time was estimated using design experiment (DOE) method and only the effects of friction pressure and friction time were considered in experiments, since the upset time showed less influence on the properties of the joints by preliminary experiments. Welding parameters used in experiments are listed in Table 3. 2.3. Methods Tensile specimens were prepared for different welding conditions to predict the tensile strength. Al least three repetitions for each condition were performed and the dimension of sample is shown in Fig. 3. During friction welding of dissimilar materials it was observed that many unwelded areas may be located both peripheral and central region of the weld interface (Winiczenko and Kaczorowski, 2012; Kimura et al., 2014). When preparing the specimens for a standard tensile test a significant portion of the outer surface is removed by machining. In this way, valuable information about participating these areas during the welding can be lost. Smaller diameter of specimen close to the weld interface

facilitates the fracture location. The no standard shape of the specimens for tensile test of dissimilar materials has been proved by the studies of Zhou et al. (1997) and Aritoshi et al. (1998). The tensile strength test was carried out on a universal testing machine (Instron 1115/PFZ100) using a crosshead speed of 5 mm/min at room temperature. The hardness measurements were conducted on Vickers microhardness testing machine (Zwick) using a load of 100 g and dwell time of 15 s. The microstructure and quantitative chemical analyses of friction joints were performed by an optical microscope (OM, Olympus IX-70) and scanning electron microscope (SEM, Zeiss 1530) working at 20 kV acceleration voltage, equipped with an energy dispersive spectroscope (EDS). After tensile tests, the fractures of the joint surfaces were examined by OM, SEM and EDS using a linear, point and maps analyses. The metallographic specimens were etched with Polution’s reagent which is the mixture of 12 ml HCl (conc.) + 6 ml. HNO3 (conc.) + 1 ml HF (48%) + 1 ml. H2O. Microstructure of joints was examined, using also transmission electron microscopy (TEM). The observations were performed on Philips EM300 transmission electron microscope, operating at 300 kV accelerating voltage. Thin-foil technique was applied for TEM study. First 3 mm rods were cut perpendicularly to the joining interface at the half-radius distance from the axis of joined specimens. Then, 0.1 mm disks were sliced from these rods, using a “load-less” IF-07A wire saw. Finally, thin foils were thinned electrochemically, using STRUERS automatic equipment. 3. Results and discussion 3.1. Tensile test Tensile test was applied after having removed the aluminium weld flashes which formed during the welding process. Tensile

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

45

Fig. 4. Example of appearance of joint tensile test specimen – (a) with weld interface fracture (b) on WHA side and (c) on AA side after tensile testing.

Fig. 5. Effect of friction welding parameters; (a) friction pressure on tensile strength, (b) friction time on tensile strength and (c) friction time on flash diameter.

strength results for various welding parameters are shown in Table 3. The most of WHA/AA samples were broken at the interface. Example of appearance of joint tensile test specimen with weld interface fracture after tensile testing is shown in Fig. 4. The macroscopic fractography showed a few scars across the whole fracture

surface that may be closely related to lower tensile strength of joint. The concentric rubbing marks were not observed at the interface on both sides of welds. However, tungsten transferred particles were observed on aluminium alloy side.

46

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

The effect of friction pressure and friction time on the ultimate tensile strength (UTS) of welded joints was presented in Fig. 5a. As can be seen in Fig. 5a and b, the maximum tensile strength could increase up to a certain value of 234 MPa, which was 84.78% efficiency of the aluminium alloy strength base material. In addition, it can be seen, that UTS decreased slightly with increasing the friction pressure for both friction times. The effect of friction time on tensile strength of WHA/AA joints is presented in Fig. 5b. The graph clearly shows that UTS joints increased with increasing the friction time and slightly decreased after having reached an optimum value for fiction time of 6.5 s. The tensile strength for FP = 40 MPa increased from 45 MPa at 0.5 s to a maximum of 165 MPa at 6.5 s and slight decreased to 128 MPa at 9.5 s. Similar trends were reported by Palanivel et al. (2017) and Sahin (2007). Probably, the frictional heat available at 0.5 s is not sufficient for effective plastic flow during the forging state. An increase friction time increased the volume of plasticised material at the plastic flow during forging, as stated by Palanivel et al. (2017). In addition, the higher values of UTS, the lower friction pressure for the same welding time. According to Kurt et al. (2011), the decrease in joints strength is probably due to flashing of the heated soft material from the interface in upsetting pressure. As expected, a larger diameter burrs were obtained using a larger frictional pressure.

The relationship between weld flashes on aluminium alloy side joints and friction time are given in Fig. 5c. It can be seen that initially, the amount of weld flash joints increased with increasing the friction time, next slightly decreased after having reached the optimum values for fiction time of 6.5 s. The relative velocity between the rotating bars generates frictional heat. The friction time determines the effective time exposure of this friction heat at the bondline. More heat is conducted in the traverse direction of the bondline with an increase in friction time, as reported by Li et al. (2014). This results in an increased volume of material being heated and an increase in the volume of the flash. In case of dissimilar materials joints by friction rotary welding, the formation of flashes depends on mechanical properties of the two parent materials, as informed Celik and Ersozlu, (2009). The properties of tungsten and aluminium alloys, especially the coefficient of thermal expansion, which are 4.6·10–6/K and 23.8·10–6/K respectively at room temperature, differ from each other. As a result, the residual thermal stresses and stress-induced distortions might be produced simultaneously during the heating and cooling process during friction welding.

Fig. 6. Microhardness profiles along the centreline in different welding parameters: (a) friction pressure and (b) friction time.

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

47

Fig. 7. Effect of friction welding parameters on the flash and shortening of WHA/AA friction welds.

3.2. Microhardness distribution The Vickers micro hardness profiles along the centreline in different welding parameters, such as friction pressure and friction time are shown in Fig. 6a and b. As can be seen in Fig. 5, the microhardness profiles for all welding parameters are almost similar. The hardness specimen reaches the maximum about 2.5 mm from the interface on WHA side and decreases in direction of parent material. Additionally, the microhardness almost remains unchanged on AA side in comparison to that on WHA side. As can be seen in Table 2, the hardness of WHA alloy is 285 HV but the weld interface hardness is around 502 HV. Similarly, the hardness value of aluminium alloy is 83 HV but the welding interface is around 98 HV. Various metallurgical phenomena, such as diffusion of elements, work hardening, dislocation density, grain refinement and precipitation at the interface, may have caused hardness increase, as suggested by Lancaster, (1987). It is obvious that the most intense processes will proceed close to the interface. The heat generated which achieved maximum in the interface is quickly transferred into the aluminum alloy side. It is known, that these alloys possessing one of the highest values of thermal conductivity coefficient, as reported by Chung, (2001). Besides, diffusion process which is limited to region very close to the interface, some processes like: dislocation formation and its redistribution, recovery and recrystallisation can take a place. These processes can proceed even at the distance several millimeters from the interface, as informed by Kaczorowski and Winiczenko (2013). As follows from TEM observations (see Section 3.5), the aluminum alloy specimen is characterised by high dislocation density, which sometimes forms dislocation cell structure. The high dislocation density is most responsible for strain hardening which is the highest close to the interface and decreases with distance from the weld interface.

Fig. 8. Relationships between temperature and yield stress of WHA and AA base materials (www.matweb.com).

3.3. Macrostructure observation Fig. 7 shows the appearances of the as-welded joints welded using various welding parameters. It can be observed that the material having lower strength in dissimilar metal combination showed more deformation resulting in higher flash formation. The weld flashes formed only on aluminium alloy side. According to Li et al. (2014), the reason of the above phenomena is the decrease of the tensile yield of aluminium alloy with temperature (see Fig. 8). Furthermore, the thermal conductivity of WHA (169 W/m K) is longer than that of AA (134 W/m K). Thus, the heat formed through friction is mainly generated on aluminium side because of its lower ther-

48

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

Fig. 9. Effect of friction welding parameters on microstructure of WHA/AA friction welds.

mal conductivity, which results in an unsymmetrical temperature field. The shape of welded flashes resembled cones (Fig. 7a–d) and cups (Fig. 7e–h), respectively. As can be seen in Fig. 7, the size of weld flashes depends on friction welding parameters. Consequently, the flashes undergo greater deformation resulting from a high axial force (Fig. 7e–h), when compared to flashes welded by less axial force (Fig. 7a–d). Moreover, it is evident that the flash scale for 7.5 s is much bigger than that for 0.5 s, which is consistent with the longer burn-off length. Similar phenomenon has already been observed by Li et al. (2014). 3.4. Microstructure investigation by optical microscope Fig. 9 shows the effect of welding parameters on the weld interface and microstructure at the central region of joints. It can be noticed that the interface is narrow and almost straight in high

friction time. The high-temperature flow strength of WHA is much greater than in aluminium alloy (Liu et al., 2008). Therefore, almost all the axial deformation during dissimilar friction joining operation occurs on aluminium alloy side. This explains the planar shape of interface in WHA/AA friction joint (see Fig. 9c and f). A continuous and homogeneous welding line during joining of tungsten carbide to nickel was also reported by Lemus-Ruiz et al. (2008). It can be seen that for all the joints a thermomechanical effect and dynamic recrystallisation have been proceeded. An effect of severe plastic deformation on the grain morphology can be seen clearly. These processes are predominant in aluminium alloy as a low strength material. This observation is similar to that described by Ahmad Fauzi et al. (2010). Thin layers formed in aluminium alloy consist of very fine grains which appear along the interface. The grains size had about 5–8 ␮m in diameter or even less. Adjacent to the interface, bent and elongated grains formed as a result of extreme deformation when a

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

49

Fig. 10. Optical micrograph showing microstructure formed on the aluminium alloy side adjacent to the joint interface.

Fig. 11. Optical micrograph showing microstructure and shape of layers formed on AA side.

high friction force was loaded. Moreover, close to this region grain coarser was observed. The upsetting pressure aids grain refinement, while friction pressure aids grain coarsening, as reported by Satyanarayana et al. (2005). Comparing to original morphology of WHA, it is evidently seen that plastic deformations at the weld are produced after welding. The shape of tungsten particles changes from the original ellipsoid to strip or directly into smaller grains. Moreover, tungsten grains are elongated along the bondline (Fig. 9g). It can be observed that with an increase of friction pressure, the level of deformed tungsten grains (DWG) is increased (Lee

et al., 2000). At this level of microstructure observation, there were no cracks in the welding alloys. Both, in the axis and joint periphery, the discontinuous transition layers preferentially formed on the aluminium alloy side adjacent to the joint interface (see Fig. 10a). An optical micrograph of WHA/AA joint interface shows a region of about 210–260 ␮m consisting of fine dynamic recrystallised grains (DRX) on the aluminium alloy side adjacent to the joint interface (Fig. 10a). It can be seen in Fig. 10a that the width of DRX region increases toward periphery because the rate of heat generation is bigger at the periphery than in the centre. The sliding

50

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

velocity occurring during friction welding process increases from the centre to periphery, since this region experiences more severe plastic deformation (see Fig. 10b) and higher temperature, when compared to the centre (Meshram et al., 2007). Fig. 11a–d shows shape and thickness of layer formed for friction pressure of 40 MPa and friction pressure of 72 MPa, respectively. These figures also show the effect of friction time on shape and thickness of the layers. These irregular layers are the areas of localised plastic flow and mechanical mixing in the aluminium alloy substrate, as reported by Pan et al. (1996). The plastic flow on aluminium alloy side is evidently visible in Fig. 11a. Additionally, the coarser grains were observed during longer heating time (see Fig. 11b and d). An effect of welding parameters on the average thickness of DRX layers of WHA/AA friction welds are shown in Fig. 12. In the graph it is clearly visible that the thickness of layers increases with increasing the friction pressure. After applying friction pressure of 72 MPa the average thickness layer increased about 1.5 mm. In addition, Figs. 11 and 12 show a slight increase in thickness layer with increasing friction time. However, in some cases, the thickness of the layers is irregular, very thin and discontinuous (see Fig. 11c). The difference in the thickness layers was due to different heat generations and plastic deformations occurred in various positions of interface, as reported by Akbarimousavi and GohariKia (2011).

Fig. 12. Effect of welding parameters on the average thickness of WHA/AA friction welds.

3.5. Microstructure investigation by transmission electron microscope Thin foils TEM observations showed (Fig. 13) some effect of heavy plastic working at higher temperature, assisting the friction welding process. As a result of plastic deformation of stacking faults, dislocation loops in the materials were formed (Fukumoto et al., 2002).

Fig. 13. TEM micrograph showing the microstructure in aluminium alloy area at different distances from the WHA/AA interface: (a) small sub-grain located close to grain boundary observed at distance of 0.12 mm, (b) dislocation arrangement into low-angle sub-grain boundary configuration at distance of 10 mm, (c) very fine weakly defined sub-grain cell structure at distance of 20 mm.

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

51

Fig. 14. (a) SEM micrograph of WHA/AA welds interface and (b) corresponding EDS scanning line across the interface.

Fig. 15. X-ray elemental mapping of WHA/AA friction weld interface.

It can be seen in Fig. 13a and b that some dislocation trapped at small precipitates and/or at subgrain boundary. These low-angle grains rotate to form high-angle strain free grains creating zone of very fine equiaxed grains (“DRX” grains), if compared to the base materials (Mishra and Ma, 2005). As the density of these dislocations increases, they tend to form sub-grain cell structure (see Fig. 13c). 3.6. Microstructure investigation by scanning microscope In order to identify the elemental diffusion, the EDS line scanning and spot analyses were applied at the weld interface. Fig. 14 shows EDS line analysis throughout the interface obtained in a WHA/AA. The variation of Al, Mg, W, Ni, Fe and Co was detected through the line scanning. As can be seen in Fig. 14b, there is no significant change in the elemental distribution after friction welding process. Moreover, Fig. 14 reveals that intermetallic phases and precipitates were not present at the interface because of low heat input and fast cooling after welding process (Asif et al., 2015). An

intermetallic compounds zone was not observed at the weld interface with applied friction pressure of 72 MPa, which was based on SEM observation level. Similar phenomenon between titanium and low carbon steel welded joints had been observed by Kimura et al. (2014). In order to get a good idea of the distribution of various elements in WHA sample, X-ray mapping was carried out and the results are presented in Fig. 15. The results of EDS chemical composition confirm the presence of main alloy additions of Mg, Al, Mn in the aluminium alloys and W, Ni, Fe and Co in WHA. This confirms the presence of Fe and Ni mostly in the ductile phase in between the tungsten grains. However, in Fig. 15e (red arrow) and Fig. 16, trace amounts of tungsten on aluminium side close to the interface have been identified. As can be seen in Table 4, the results of EDS point analysis in Fig. 16a indicate 0.38 of W and 0.78 of Ni in (wt.%), respectively. Fig. 17a and b shows the secondary electron image (SEI) of WHA/AA friction weld with six spots in the Ni-Fe-matrix. Since WHA is a binary phase alloy consisting of W-grains in the Ni-

52

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

Fig. 16. (a) Micrograph showing aluminium alloy side close to the interface and (b) EDS-SEM spectrum at the places marked with green cross. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 4 Results of EDS point analysis on aluminium alloy side.

Table 5 Compositions of six spots shown in Fig. 17.

Element

Series

Norm. wt.%

Atom. wt.%

Al Fe Mn Mg Ni W

K-series K-series K-series K-series K-series K-series

66.92 23.17 5.32 1.05 1.41 2.13

80.77 13.57 3.15 1.41 0.78 0.38

Fe-matrix, hence, the welding of WHA to AA alloy could be considered via two ways: the welding of aluminium alloy to tungsten grains (Fig. 17a) and that of aluminium alloy to nickel-iron matrix (Fig. 17b), as suggested by Wang et al. (2005). The EDS spot analysis was employed to confirm the elemental content at the weld interface (Ma et al., 2016). The compositions of 1–6 spots have been analysed quantitatively and the results are presented in Table 5. It can be seen in Fig. 17a and b that 1, 2 and 3 are spots in the Ni-Fe-matrix for Alalloy to W-grains joint, while 4, 5 and 6 are spots in the Ni-Fe-matrix for aluminium alloy to Ni-Fe-matrix joint. Table 5 shows some elemental aluminium in the Ni-Fe matrix. It can be seen that aluminium content in Ni-Fe matrix increases a little along with approaching the interface for both joints. Greater amount of aluminum in Ni-Fe matrix near the interface was detected in case of aluminium alloy to the nickel-iron matrix joint interface. The appearance of aluminium in Ni-Fe-matrix may prove its little diffusion across the interface to WHA. 3.7. Fractography Fig. 18 shows SEM fractographs of WHA/AA tensile tested samples in as-welded conditions (FP = 40 MPa, FT = 7.5 s). Fig. 18a and b shows SEM fractographs of WHA side, while Fig. 18c and d shows SEM fractographs of aluminium alloy side, respectively. It is clearly seen in Fig. 18a and b that cleavage fractures through the planes of tungsten grains are detected on WHA side. Additionally, many tungsten particles can be clearly identified on aluminium alloy side, (Fig. 18c and d). It can be assumed that the tungsten particles strongly adhere to aluminium surface. Prior SEM-EDS point and map analyses on the welded surfaces showed a trace amount of tungsten in the immediate vicinity of the joint on aluminium alloy side. This confirms that the material adhering to aluminium alloy side are tungsten

Spots

1 2 3 4 5 6

Elements percentage (at.%) Ni

Fe

W

C

Co

Al

51.92 49.30 48.75 41.55 43.69 44.26

13.53 12.66 12.36 10.71 11.46 11.42

10.37 10.99 10.64 8.28 8.45 8.59

19.51 22.67 24.08 30.80 31.95 31.87

4.30 4.10 3.95 3.43 3.76 3.48

0.38 0.28 0.23 5.23 0.68 0.37

grains and some part of fracture during the tensile test propagated through tungsten grains. Moreover, the ductile fracture by micro/macro void coalescence (dimples) in the centre of fractograph (Fig. 18d) can be detected on aluminium alloy side. In this fractograph fracture mode was a combination of dimple fracture and cleavage fracture (Ma et al., 2016). 4. Conclusions The microstructure, tensile strength and microhardness of joints between tungsten and aluminium alloys after conventional rotary friction welding were investigated with a focus on changes in the chemical compositions at the interface. The following main results were obtained: 1. Friction welding of WHA to AA is successfully carried out without any interlayers and heat treatment. The maximum average strength of 234 MPa, which is 84.78% of the aluminium alloy base material, is achieved at a friction time of 3.5 s and friction pressure of 40 MPa, respectively. The tensile strength increases with friction time and slightly decreases after having reached the optimum value under the experimental conditions. 2. The microstructure of friction welds consisted of fine equiaxed grains due to dynamic recrystallisation and coarse grains in the periphery region in aluminium alloy side. Moreover, a plastic deformation in the direction of burrs is visible mainly on aluminium alloy side. 3. EDS-SEM maps and scan line analyses across the interface have not confirmed the diffusion of tungsten and nickel to aluminium alloy side. However, the EDS-SEM point analysis indicated a slight amount of aluminium in the Ni-Fe-matrix close to the

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

53

Fig. 17. SEM micrograph of two variants of weld interface with the marked spots: (a) aluminium alloy to tungsten grain, (b) aluminium alloy to nickel-iron matrix, (c) EDS spectrum from spot 1, (d) EDS spectrum from spot 2, (e) EDS spectrum from spot 3, (f) EDS spectrum from spot 4, (g) EDS spectrum from spot 5, (h) EDS spectrum from spot 6.

54

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55

Fig. 18. SEM fractographs of WHA/AA friction welded joint: (a–b) on WHA side; (c–d) on AA side.

interface. The absence of intermetallic phases was indictaed in the weld interface on SEM level observation. 4. On the basis of EDS and fracture observation it can be concluded that the nature of friction welding joints is rather adhesive than diffusive. The fracture observation of WHA/AA interface shows brittle fracture of interface. The material which adheres to aluminium alloy side after rupture are tungsten particles. The fracture during the tensile test propagated through tungsten grains. Moreover, both map and point EDS-analyses indicate presence of trace amount of tungsten in the immediate vicinity of the joint on the aluminium alloy side.

References Ahmad Fauzi, M.N., Uday, M.B., Zuhailawati, H., Ismail, A.B., 2010. Microstructure and mechanical properties of alumina-6061 aluminum alloy joined by friction welding. Mater. Des. 31, 670–676. Akbarimousavi, S.A.A., GohariKia, M., 2011. Investigations on the mechanical properties and microstructure of dissimilar cp-titanium and AISI 316L austenitic stainless steel continuous friction welds. Mater. Des. 32, 3066–3075. ´ P., Winnicki, M., 2011. Friction welding Ambroziak, A., Korzeniowski, M., Kustron, of niobium and tungsten pseudoalloy joints. Int. J. Refract. Met. Hard Mater. 29, 499–504. Ambroziak, A., 2010. Friction welding of titanium–tungsten pseudoalloy joints. J. Alloys Compd. 506, 761–765. American Welding Society, 1989. Specifications and standards. In: Recommended Practice for Friction Welding. American Welding Society, Miami. Aritoshi, M., Okita, K., 2003. Friction welding of dissimilar metals. Weld. Int. 17 (4), 271–275. Aritoshi, M., Okita, K., Okamato, K., Igarashi, T., 1998. Friction welding of tungsten to copper with an Nb intermediate layer. Weld. Int. 12 (11), 845–850. Arivazhagan, N., Singh, S., Prakash, S., Reddy, G.M., 2011. Investigation on AISI 304 austenitic stainless steel to AISI 4140 low alloy steel dissimilar joints by gas tungsten arc electron beam and friction welding. Mater. Des. 32, 3036–3050. Asif, M.M., Shrikrishna, K.A., Sathiya, P., Goel, S., 2015. The impact of heat input on the strength, toughness, microhardness, microstructure and corrosion aspects of friction welded duplex stainless steel joints. J. Manuf. Processes 18, 92–106.

Cai, W.D., Li, Y., Dowding, R.J., Mohamed, F.A., Lavernia, E.J., 1995. A review of tungsten-based alloys as kinetic Energy penetrator materials. Rev. Part. Mater. 3, 71–131. Celik, S., Ersozlu, I., 2009. Investigation of the mechanical properties and microstructure of friction welded joints between AISI 4140 and AISI 1050 steels. Mater. Des. 30, 970–976. Chung, D.D.L., 2001. Materials for thermal conduction. Appl. Therm. Eng. 21, 1593–1605. Das, J., Appa Rao, G., Pabi, S.K., 2010. Microstructure and mechanical properties of tungsten heavy alloys. Mater. Sci. Eng. A 527, 7841–7847. Das, J., Appa Rao, G., Pabi, S.K., Sankaranarayana, M., Sarma, B., 2011. Deformation behaviour of a newer tungsten heavy alloy. Mater. Sci. Eng. A 528, 6235–6247. Dey, H.C., Ashfag, M., Bhaduri, A.K., Rao, K.P., 2009. Joining of titanium to 304L stainless steel by friction welding. J. Mater. Process. Technol. 209, 5862–5870. Fukumoto, S., Tsubakino, H., Aritoshi, M., Tomita, T., Okita, K., 2002. Dynamic recrystallization phenomena of commercial purity aluminium during friction welding. Mater. Sci. Technol. 18, 219–225. German, R.M., Suri, P., Park, S.J., 2009. Review: liquid phase sintering. J. Mater. Sci. 44 (1), 1–39. Kaczorowski, M., Winiczenko, R., 2013. The microstructure and mass transport during friction welding of ductile cast iron. Ind. Lubr. Tribol. 65 (4), 251–258. Kimura, M., Iijima, T., Kusaka, M., Kaizu, K., Fuji, A., 2014. Joining phenomena and tensile strength of friction welded joint between pure titanium and low carbon steel. Mater. Des. 55, 152–164. Kurt, A., Uygur, I., Paylasan, U., 2011. Effect of friction welding parameters on mechanical and microstructural properties of dissimilar AISI 1010-ASTM B22 joints. Weld. J. 90, 102–106. Lancaster, J., 1987. Metallurgy of Welding. Allen and Unwin, London. Lee, W.S., Lin, C.F., Chang, S.T., 2000. Plastic flow of tungsten-based composite under hot compression. J. Mater. Process. Technol. 100, 123–130. Lemus-Ruiz, José, Ceja-Cárdenas, L., Verduzco, J.A., Flores, O., 2008. Joining of tungsten carbide to nickel by direct diffusion bonding and using a Cu–Zn alloy. J. Mater. Sci. 43, 6296–6300. Li, P., Li, J., Salman, M., Liang, L., Xiong, J., Zhang, F., 2014. Effect friction time on mechanical and metallurgical properties of continuous drive friction welded Ti6Al4V/SUS321 joints. Mater. Des. 56, 649–656. Liu, J.X., Li, S.K., Zhou, X.Q., Zhang, Z.H., Zheng, H.Y., Wang, Y.C., 2008. Adiabatic shear banding in a tungsten heavy alloy processed by hot-hydrostatic extrusion and hot torsion. Scr. Mater. 59 (12), 1271–1274. Ma, H., Qin, H., Geng, P., Li, F., Meng, X., Fu, B., 2016. Effect of post-weld heat treatment on friction welded joint of carbon steel to stainless steel. J. Mater. Process. Technol. 237, 24–33.

R. Winiczenko et al. / Journal of Materials Processing Technology 246 (2017) 42–55 Meshram, S.D., Mohandas, T., Reddy, G.M., 2007. Friction welding of dissimilar pure metals. J. Mater. Process. Technol. 184, 330–337. Mishra, R.S., Ma, Z.Y., 2005. Friction stir welding and processing. Mater. Sci. Eng. R 50, 1–78. Palanivel, R., Laubscher, R.F., Dinaharan, I., 2017. An investigation into the effect of friction welding parameters on tensile strength of titanium tubes by utilizing an empirical relationship. Measurement 98, 77–91. Pan, C., Hu, L., Li, Z., North, T.H., 1996. Microstructural features of dissimilar MMC/AISI 304 stainless steel friction joints. J. Mater. Sci. 31, 3667–3674. Park, S., Kim, D.K., Lee, S., 2001. Dynamic deformation behavior of an oxide-dispersed tungsten heavy alloy fabricated by mechanical alloying. Metall. Mater. Trans. A 32 (8), 2011–2020. Reddy, G.M., Ramana, P.V., 2012. Role of Nickel as an interlayer in dissimilar metal friction welding of maraging steel to low alloy steel. J. Mater. Process. Technol. 212, 66–77. Ryu, H.J., Hong, S.H., 2003. Fabrication and properties of mechanically alloyed oxide-dispersed tungsten heavy alloys. Mater. Sci. Eng. A 363 (1–2), 179–184. Sahin, M., 2007. Evaluation of the joint-interface properties of austenitic-stainless steels (AISI 304) joined by friction welding. Mater. Des. 28, 2244–2250. Satyanarayana, V.V., Reddy, M.G., Mohandas, T., 2005. Dissimilar metal friction welding of austenitic-ferritic stainless steels. J. Mater. Process. Technol. 160, 128–137. Scapin, M., 2015. Mechanical characterization and modeling of the heavy tungsten alloy IT180. Int. J. Refract. Met. Hard Mater. 50, 258–268.

55

Sunwoo, A., Groves, S., Goto, D., Hopkins, H., 2006. Effect of matrix alloy and cold swaging on micro-tensile properties of tungsten heavy alloys. Mater. Lett. 60 (3), 321–325. Taban, E., Gould, J.E., Lippold, J.C., 2010. Dissimilar friction welding of 6061-T6 aluminum and AISI 1018 steel properties and microstructural characterization. Mater. Des. 10, 2305–2311. Teker, T., 2013. Evaluation of the metallurgical and mechanical properties of friction-welded joints of dissimilar metal combinations AISI2205/Cu. Int. J. Adv. Manuf. 66, 303–310. Wang, C.B., Shen, Q., Zhou, Z.G., Zhang, L.M., 2005. Diffusion welding of 93W alloy to OFC and structural control of 93W/OFC joint. J. Mater. Sci. 40, 2105–2107. Winiczenko, R., Kaczorowski, M., 2012. Friction welding of ductile cast iron using interlayers. Mater. Des. 34, 444–451. Winiczenko, R., Kaczorowski, M., 2013. Friction welding of ductile iron with stainless steel. J. Mater. Process. Technol. 213, 453–462. Winiczenko, R., 2016. Effect of friction welding parameters on the tensile strength and microstructural properties of dissimilar AISI 1020–ASTM A536 joints. Int. J. Adv. Manuf. 84, 941–955. Zhou, Y., Zhang, J., North, T.H., Wang, Z., 1997. The mechanical properties of friction welded aluminium-based metal–matrix composite materials. J. Mater. Sci. 32, 3883–3889. ´ Zimmerman, J., Włosinski, W., Lindemann, Z.R., 2009. Thermo-mechanical and diffusion modelling in the process of ceramic-metal friction welding. J. Mater. Process. Technol. 209, 1664–1653.