WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2017, VOL. 96 63-s

Introduction Application of copper is quiteprominent in electronic and electricalindustries due to its excellent conduc-tivity, but it is limited by copper’shigh density and price. Recently, dis-similar materials welding betweenaluminum and copper has attractedplenty of attention due to the scarcityof copper and the relatively acceptableconductivity of aluminum. Alu-minum/copper welded structures alsooffer the benefits of cost saving andweight reduction. Veerkamp (Ref. 1) reported varioustypes of aluminum copper weldedstructures had been used in electricalapplications, especially in high direct-current bus systems. However, severalchallenges prevent sound aluminum/copper joints, such as oxidation, the

high conductivity of base materials,the great difference in thermal-physical properties between the twomaterials, and the inevitable forma-tion of brittle intermetallic com-pounds (IMCs) during welding. The early investigation on alu-minum/copper diffusion system un-dertaken by Funamizu and Watanabe(Ref. 2) showed the formation of Al-Cu IMCs was controlled by theprocess of volume diffusion mecha-nism, and the rate of layer growthobeyed the parabolic law. Peng et al.(Ref. 3) reported that as time and/ortemperature of heat treatment in-creased, Kirkendall voids formed andagglomerated, resulting in a weak lay-er in the roll-bonded aluminum/copper interface. Therefore, to avoid the massivegeneration of Al-Cu IMCs and Kirk-endall voids, low heat input welding

methods became the research focus injoining aluminum and copper. Maiand Spowage (Ref. 4) achieved laserwelding between oxygen-free copperand 4047 aluminum alloy at lowerwelding speeds, but they also claimedit was difficult to obtain a sound alu-minum/copper joint by fusion weld-ing due to the high reflectivity andhigh thermal conductivity of the twomaterials. A large number of recent papershave discussed friction stir welding ofaluminum alloy and copper plates.Xue et al. (Ref. 5) focused on the rela-tionship of welding parameters to themicrostructure and mechanical prop-erty of aluminum/copper friction stirwelded joints. Also studied by Xue etal. (Ref. 6) were the influences ofpostweld heat treatment (PWHT) onthe interfacial microstructure evolu-tion, tensile strength, and fracture be-havior of aluminum copper dissimilarfriction stir welded joints. Beygi et al.(Ref. 7) performed friction stir weld-ing of aluminum/copper bilayersheets and observed the anomalousstructure of Al-Cu IMCs in backscat-tered electron (BE) images. Friction welding (FW) relies on rel-ative motion between the parts beingjoined to generate friction heat whilepressure is applied. Meshram et al.(Ref. 8) documented that FW couldproduce sound dissimilar joints withacceptable mechanical property. Fig-ure 1 schematically shows the contin-uous drive friction welding process ofaluminum and copper bars applied inthis research. This welding processcan be briefly distinguished into fourmain stages: (a) rotating and ap-

Microstructure and Mechanical Property ofAluminum/Copper Dissimilar Friction

Welded JointsSound aluminum/copper dissimilar joints can be produced with continuous

drive friction welding after a 2­h postweld heat treatment at 100C

BY S. CHEN, H. DONG, P. LI, S. NIU, L. YU, AND H. JI

ABSTRACT Continuous drive friction welding of commercial pure aluminum to copper bars wasconducted. Sound joints were produced after postweld heat treatment (PWHT) at 100°Cfor 2 h with an ultimate tensile strength of 91 MPa and a bending angle of 90 deg with­out breaking. All heat­treated tensile samples failed at the aluminum/copper interfacewith a cleavage fracture feature. The microstructure, composition, and fractograph ofthe joints were analyzed with an electron probe microanalyzer and a scanning electronmicroscope. Copper was stirred into the aluminum matrix due to mechanical excavationduring welding. After PWHT at 100°C for 2 h, the minimized amount of Cu distributed atthe interface, which benefited the mechanical property of the joints. Massive formationsof brittle intermetallic compounds (IMCs) at the interface could deteriorate the mechani­cal property of joints. Fractographs revealed that Al­Cu IMCs initially formed at one­thirdof the radius of the faying surface due to comprehensive influences of rotation speedand friction pressure.

KEYWORDS • Continuous Drive Friction Welding • Aluminum/Copper Dissimilar Joints • Postweld Heat Treatment • Mechanical Property • Intermetallic Compounds

S. CHEN, H. DONG (donghg@dlut.edu.cn), P. LI, S. NIU, and L. YU are with the School of Materials Science and Engineering, Dalian University of Technology,Dalian, China. DONG is also with the State of Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, Harbin, China. LI is also withthe State Key Laboratory of Solidification Processing, Northwestern Polytechnical University, Xi’an, China. H. JI is with Shanghai Aerospace Equipment Manufacturer, Shanghai, China.

Chen2016439_Layout 1 1/12/17 3:55 PM Page 63

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2017, VOL. 9664-s

proaching, (b) frictional heating, (c)forging, and (d) retreating. During welding, friction heat isgenerated by the comprehensive ef-fect of rotational motion and frictionload (F1), and then forging is achievedby forging load (F2). Considering thelarge difference in hardness betweenaluminum (HB 18) and copper (HB65) in this research, to avoid severedeformation on the aluminum bar, a15-mm-length steel ring was set onthe aluminum bar with a 10-mm dis-tance from the faying surface beforewelding. These steel rings were madefrom plain carbon steel with innerand external diameters of 16 mm and25 mm, respectively. Braunović and Aleksandrov (Ref. 9)systematically studied the influences ofAl-Cu IMCs on the electric and mechan-ical property of aluminum copper dis-similar friction welded joints, and docu-mented that Al-Cu IMCs at the interfacecould seriously impair the mechanicalproperty when the total width of IMCsphases at the interface exceed 2 m. Thecoupled effect of the main parametersof the aluminum/copper dissimilar FWprocess on the mechanical property ofjoints was statistically studied by Yilbaset al. (Ref. 10). Lee et al. (Ref. 11) employed variousannealing conditions to research theelectric and mechanical property of alu-minum/copper dissimilar friction weld-ed joints, and experimentally deter-mined the growth kinetic of Al-Cu IMCsalong the aluminum/copper interface.The influences of the initial geometry of

prestrain alu-minum alloyand copperbars on FWwere investi-gated byManjubhar-gavi et al.(Ref. 12). Thelarge residualstress leavingin the joint after the FW process deteri-orated the bending property, which issignificant for the aluminum/copperjoints used in the electric powerindustry. Dong et al. (Ref. 13) reportedPWHT could enhance comprehensiveproperties of friction welded jointsbetween TC4 titanium alloys and40Cr steel bars. Therefore, consider-ing the residual stress leaving in thejoint, PWHT was also performed inthis research. In addition, although some papershave reported the effect of PWHT onthe mechanical property of alu-minum/copper joints, there are fewreports on low-temperature heattreatments (less than 300°C) and

short dwelling times (1–2 h) for alu-minum/copper friction welded joints. This paper will focus on the effect ofPWHT with low temperature and shortdwelling time on the mechanical prop-erty and microstructure of aluminum/copper friction welded joints.

Experimental Procedure

Continuous drive friction weldingof commercially pure aluminum barsto copper bars was conducted using aHSMZ-20 friction welding machineproduced by Harbin Welding Insti-tute, China. The nominal compositionof the base materials is listed in Table1. The ultimate tensile strength (UTS)

Fig. 1 — Schematic illustration of the continuous drive frictionwelding process.

Fig. 2 — The appearance of the as­welded sample.

Fig. 3 — Dimensions: A — Tensile sample; B — three­point bending testsample.

B

Table 1 — Nominal Composition of Base Materials

Base Materials Composition (wt­%)

Aluminum — — Fe Si Ti Al — — 0.2 0.17 0.13 Bal.

Copper S Bi Sb As Fe Cu 0.005 0.001 0.002 0.002 0.005 Bal.

A

Chen2016439_Layout 1 1/12/17 3:55 PM Page 64

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2017, VOL. 96 65-s

and elongation of copper bars were295 MPa and 3%, and those of alu-minum bars were 90 MPa and 80%,respectively. The copper bars were machined to100 mm in length and 12 mm in di-ameter. The dimensions of the alu-minum bars were 120 mm in lengthand 16 mm in diameter. Both the alu-minum and copper bars had 60 degchamfers with 1-mm thickness at thefaying surfaces on two ends after pro-cessing. Before welding, the fayingsurfaces of the aluminum and copperbars were polished by SiC paper up togrit 1000 for removal of oxide film,and then cleaned by acetone. After trial welding, the optimizedwelding parameters were determined.The friction load (F1) and forging load

(F2) displayed in Fig. 1 were set at 75and 100 MPa, respectively, the burn-offlength remained around 22 mm, theforging time was 5 s, and the total timeof the welding process was 12 s. The as-welded sample is shown in Fig. 2. As mentioned, the copper barcould be considered inconsumableduring welding due to the higherhardness of copper than that of alu-minum. The aluminum/copper inter-face location was approximately de-termined in Fig. 2, based on the origi-nal length of the copper bar. Afterwelding, the samples labeled from 2to 5 were heat treated under four con-ditions, as listed in Table 2. The sam-ples were subjected to heat treat-ments for a given time and thencooled to room temperature in a

forced air electrical furnace. Tensile and bending tests were car-ried out to evaluate the effect of PWHTon the mechanical property of the re-sultant joints. The dimensions of thetensile test sample is shown in Fig. 3A.According to China National StandardsGB/T 232-2010, selected samples withas-welded length and a diameter of 11mm (Fig. 3B) were employed for thethree-point bending test over a 60-mmspan with a crosshead speed of 0.5mm/s. Both tensile and bending testswere performed using the DNS-100universal test machine at ambient tem-perature. All mechanical results of thetensile and bending tests were obtainedby averaging the results of threeindividual samples. Metallographic samples preparedfrom the as-welded and PWHT condi-tions were subjected to mechanicalgrinding and final polishing with a solu-tion of 0.5-m silica suspension, andchemically etched with Keller's reagent(4 mL HF, 6 mL HCl, 10 mL HNO3, and180 mL H2O). The microstructural features were obtained using a LeicaMEF4A optical microscope. Microstructural morphology andcomposition analysis of joints were

carried out by electron probe micro-analyzer (EPMA) (Shimadzu, Japan)complemented with a wavelength dis-persive spectrometer (WDS). The frac-tographs and composition detectionwere presented by EPMA and JSM-5600LV scanning electron microscope(SEM) (Zeiss, Germany) with energydispersive spectrometer (EDS) under anaccelerating voltage of 15 kV.

Results

Mechanical Property andMicrostructure

Braunović and Aleksandrov (Ref. 9)and Xue et al. (Ref. 6) reported PWHThad remarkable influences on the evo-lution of Al-Cu IMCs, leading to avariation in the mechanical propertyof aluminum/copper joints. The re-sults of tensile and bending tests ofaluminum copper dissimilar frictionwelded joints are listed in Table 3. Asshown, the UTS and bending angle ofjoints initially increased with the helpof PWHT, and reached an optimal me-chanical property of 91 MPa in UTS

Fig. 4 — Fractured samples: A — Tensile test; B — bending test and the details on the bottom of the bending samples; C — sample 4; D — sample 5.

Table 2 — PWHT Parameters for Aluminum/Copper Friction Welded Samples

Sample Temperature (oC) Dwelling Time (h)

1 — — 2 100 1 3 100 2 4 200 2 5 300 2

A B

C D

Chen2016439_Layout 1 1/12/17 3:55 PM Page 65

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2017, VOL. 9666-s

and 90 deg in bending angle underPWHT at 100°C for 2 h, as shown inthe results for sample 3. Then, UTS deteriorated from 91 to51 MPa with the increasing tempera-ture of PWHT. Most tensile samplesfailed at the alminum/copper inter-face, as shown in Fig. 4A, except sam-ple 1, which necked and fractured onthe aluminum side. This indicatedthat the aluminum/copper interface isthe weakest part of the heat-treatedjoint. The fracture analysis will bepresented later. Figure 4B presents fractured sam-ples after the bending test. Sample 1fractured entirely at the interface,leaving some extent of deformationon the aluminum side. As mentionedpreviously, the satisfactory bendingangle was obtained in sample 3, whichcould reach 90 deg without breaking.Afterward, the bending angle droppedsharply, just like the variation tenden-cy of UTS. Samples 2, 4, and 5 failedduring the bending tests becausecracks formed on the bottom of thebending samples, as presented in Fig.4C, D. Because of the large difference inhardness, flash formed during the FWprocess only from aluminum, whereascopper didn’t participate in flash for-mation. This phenomenon was alsoreported by Manjubhargavi et al. (Ref.12). The metallograph of sample 1 isdisplayed in Fig. 5. It is obvious thatsevere mechanical stir action and me-chanical excavation occurred, leavinga concave interface on the copperside, where perhaps there was copperthat detached from the base material.Because of the difference in corrosionproperty between base materials andAl-Cu IMCs, there is an obvious black-mixed zone consisting of aluminum,copper, and Al-Cu IMCs. It is worth noting the diameter ofthe tensile sample at the interface is 8mm, which is less than the mixedzone diameter. According to the frac-

ture positionshown in the me-chanical test re-sults, this mixedzone may play apredominantrole in the me-chanical proper-ty of alu-minum/copperdissimilar fric-tion weldedjoints. To under-stand the effect of PWHT on the evo-lution of the microstructure at the in-terface of the aluminum/copperjoints, the composition analysis of allsamples was performed as shown inFig. 6. The BE images clearly showcopper was detached from the basematerial and stirred into the alu-minum matrix. The composition dis-tributions at the interface varied withthe PWHT condition. There is no big difference betweensamples 1 and 2 in the distribution ofCu on the aluminum side, except for arelatively large decrease or absence insample 3. The agglomeration of Cu oc-curred both in samples 4 and 5, whereasit was more serious in sample 5. The dis-tribution of Cu in the aluminum matrixis consistent with the fluctuation ofUTS, which implies the distribution ofCu influences the mechanical propertyof friction welded joints. Beygi et al. (Ref. 7) reported manylamellar structure phases formed in alu-minum/copper dissimilar friction stirwelded joints. This phenomenon is also

observed in Fig. 7A from the magnifiedview marked in the dotted rectangle inFig. 6. Composition analysis results list-ed in Table 4 reveal that the lamellarstructure phases may consist of copperand Al2Cu phases. There were variousphases in the magnified view of sample4, as shown in Fig. 7B. According to the nature of the BE im-age and the results of compositionanalysis, as the color deepened fromwhite to black, the Al content in thephases increased, which implied the ex-istance of various Al-Cu IMCs due to thePWHT. The same phenomenon was alsoobserved in the microstructure of sam-ple 5 with a higher PWHT temperaturethan sample 4.

Fracture Morphology In this research, the fracture surfacesfrom samples 2 to 5 on the copper siderevealed a pale grey aluminum layer,which Manjubhargavi et al. (Ref. 12) at-tributed to interatomic diffusion duringthe aluminum/copper FW process. Frac-

Fig. 6 — Composition analysis of all samples at the interface center.

Fig. 5 — Metallograph of sample 1 at the aluminum/copperinterface.

Table 3 — Results of Tensile and Bending Tests

Sample UTS (MPa) Fracture Position Bending Angle (deg)

1 68 Aluminum side 40 2 76 Interface 66 3 91 Interface >90 4 74 Interface 35 5 51 Interface 22

Chen2016439_Layout 1 1/12/17 3:55 PM Page 66

WELDING RESEARCH

FEBRUARY 2017 / WELDING JOURNAL 67-s

tographs of sample 2 are presented inFig. 8. Observable in the secondary elec-tron (SE) images are the existance ofsmall hemispherical “embossments” at aconcentric annulus area beyond 2 mmradius, which showed the anomalousstructure phases in the BE image. Thesephases were also reported by Beygi et al.(Ref. 7). Composition analysis at point B1

listed in Table 5 suggests anomalousstructure phases seem to be mainlyAl2Cu. According to the distribution ofanomalous structure phases, thesephases formed due to the synthetic ef-fect of friction heat and pressure. Notonly were anomalous structure phasesfound in the fracture surface of tensilesample 2, but these phases were alsoobserved on bending sample 1.

The morphology of the anomalousstructure phases evolved with PWHTconditions. Figure 9A is the BE imageof sample 3 on the fracture surface ofthe copper side. It shows that someunchanged anomalous structure phas-es remain, but others transformedinto strip structure phases. Addition-ally, many cracks generated at thestrip structure phases and propagatedthrough the strip direction. The same phenomenon was alsofound in sample 4, as shown in Fig.9B, but the size of the strip structurephases diminished with the disap-pearance of the anomalous structurephases. However, the BE image ofsample 5 was totally changed com-pared with that of former samples. InFig. 9C, white phases are distributedon the fracture surface and surround-ed by dark phases. According to thecomposition analysis results in Table5, these dark phases (B2) and whitephases (B3) seem to be mainly Al2Cuand Al2Cu3, respectively. According to the fractographs oftensile samples under differentPWHT conditions, it can be deter-mined that PWHT is responsible forthe formation and growth of IMCs atthe interface, resulting in influenceson the mechanical property of joints.

SEM images revealed more detailsabout the tensile fracture behavior ofheat-treated joints. Shallow cleavagestep patterns were clearly observed inthe fracture surfaces of samples 3 and 5on the copper side, as shown in Figs.10A and 10C, respectively. The samepatterns also existed on the fracturesurfaces of samples 2 and 4. In the fur-

Fig. 7 — Magnified views: A — Sample 1; B — sample 4 marked in Fig. 6.

Fig. 8 — A, B — SE images; C, D — BE images of fracture of tensile sample 2 on the cop­per side.

A B

A

C D

B

Table 4 — Major Composition of Different Positions Marked in Fig. 7

Composition (at.­%) Position Possible Phase Al Cu O Al/Cu

A1 16.60 80.71 2.69 0.2 Copper A2 63.38 31.65 4.97 2.0 Al2Cu A3 12.28 82.25 5.47 0.1 Copper A4 29.46 58.64 11.90 0.5 Al4Cu9/Al2Cu3 A5 49.34 42.20 8.46 1.2 AlCu A6 61.99 32.49 5.51 1.9 Al2Cu A7 92.42 6.36 1.22 14.5 Aluminum

Chen2016439_Layout 1 1/12/17 3:55 PM Page 67

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2017, VOL. 9668-s

ther magnified views of Figs. 10B and10D, the intergranular fracture featureon sample 3 turned to the transgranularone on sample 5 by PWHT, which isconsistent with what Xue et al. (Ref. 6)reported.

According to composition analysisresults, equiaxed grains on the frac-ture surface of sample 3 seemed to bemainly Al2Cu, but Al2Cu3 was detectedon sample 5, which is consistent withformer analysis with EPMA.

Discussion As mentioned, this research focusedon the effect of PWHT on the mi-crostructural evolution and mechanicalproperty of aluminum/copper joints.The sound aluminum/copper jointswere obtained in sample 3, and had anoptimal mechanical property of 91 MPain UTS and 90 deg in bending anglewithout breaking. Heat-treated samplesfractured at the aluminum/copper in-terface in the tensile test, and mainlycracked at the bottom of samples duringthe bending test.

However, sample 1necked and fracturedon the aluminumside around 25 mmaway from the inter-face. As Fig. 2 shows,the root of flash onthe aluminum side isapproximately locat-ed at the fracture po-sition of sample 1 inFig. 4A. The resultsare mainly caused byresidual stress leav-

ing in deformed aluminum. After fric-tion welding, residual stress caused bysevere deformation remained on thealuminum side, which weakened themechanical property of the joints. Residual stress leaving in the jointalso resulted in the total fracture ofsample 1, rather than a crack at thebottom during the bending test. Dur-ing PWHT, the residual stress in thejoint was gradually released, and themechanical property was improved.Meanwhile, Al-Cu IMCs formed andgrew at the interface. Therefore, the

weakest position of the joints shiftedfrom the aluminum side to the alu-minum/copper interface. Composition analysis in Fig. 6 re-vealed the minimized amount of Cudistributed on the aluminum side un-der 100°C PWHT for 2 h. The soundaluminum/copper joints were also ob-tained under this PWHT condition.As the temperature of PWHT in-creased, Cu agglomerated on the alu-minum side in samples 4 and 5, whichwas consistent with the sharp dropsin the mechanical property of the alu-minum/copper joints. According to the study by Lee et al.(Ref. 11), aluminum/copper frictionwelded joints with excellent tensilestrength reaching 87 MPa were ob-tained under PWHT at 375 K (100°C)for 129.6 ks (36 h), and fracture oc-curred at the heat-affected zone onthe aluminum alloy side. With the in-creasing temperature of PWHT, thetensile strength tended to decrease,and the fracture location changedfrom the heat-affected zone to thealuminum/copper interface, which

Fig. 9 — BE images on the fracture surface of tensile: A — Sam­ple 3; B — sample 4; C — sample 5 on the copper side.

Table 5 — Major Composition of Different Positions Marked from Figs. 8 to 10

Composition (at.­%) Position Possible Phase Al Cu O Al/Cu

B1 64.48 31.50 4.02 2.0 Al2Cu B2 55.49 32.55 11.95 1.7 Al2Cu B3 35.71 58.36 5.92 0.6 Al2Cu3

C1 63.12 34.33 2.55 1.8 Al2Cu C2 40.15 56.15 3.10 0.7 Al2Cu3

A

C

B

Chen2016439_Layout 1 1/12/17 3:55 PM Page 68

had the same tendency as the resultsobtained in this research. Xue et al. (Refs. 5, 14) reported cop-per particles friction stirred into thealuminum matrix would react with alu-minum, resulting in the formation ofIMCs. Severe mechanical stir action andmechanical excavation can be observedin Figs. 5 and 6, resulting in the detach-

ment of copper from the base material.According to the aluminum copper con-cave interface and lamellar structurephases in Fig. 7A, it is assumed copperwas stirred into the aluminum side spi-rally and pressed to the aluminum/copper interface, as schematicallyshown in Fig. 11. Composition analysisshows lamellar structure phases consistof copper and Al2Cu phases, which maybe attributed to short-circuit diffusionthrough the defects caused by the severe mechanical stir action. The evolution of Al-Cu IMCs at theinterface is related to the reaction tem-perature, which was mainly decided byPWHT conditions in this research. AsFig. 7 shows, the lamellar structurephases observed in sample 1 trans-formed into the combinational struc-ture phases, consisting of various kindsof IMCs in sample 4 due to PWHT. Anomalous structure phases wereobserved at a concentric annulus area

beyond 2 mm in radius on the frac-ture surface of sample 2. It can be explained by the nature of the FWprocess. At the beginning of welding,the faying surfaces of metal bars arein the elastic contact stage. The varia-tion of the temperature field alongthe radius was inhomogeneous, be-cause the relative rotation speed waszero at the center but increased to themaximum at the edge, and the fric-tion pressure unevenly distributed onthe contacted faying surfaces betweenthe aluminum and copper bars. Under the comprehensive influencesof rotation speed and friction pressure,generally, the heating starts at one-thirdor half of the radius of the faying sur-faces from the center, and leads to anannular heated zone, which is also sup-ported by Li et al. (Ref. 15). In our re-search, since the radius of the copperbar was 6 mm, it is reasonable that theAl-Cu IMCs formed at a concentric an-

WELDING RESEARCH

FEBRUARY 2017 / WELDING JOURNAL 69-s

Fig. 10 — Fractographs of tensile: A, B — Sample 3 ; C D — sample 5 on the copper sides.

A

C D

B

Fig. 11 — Schematic illustration of stir­ring action at the aluminum/copper interface.

Chen2016439_Layout 1 1/12/17 3:55 PM Page 69

nulus area beyond radius 2 mm. Thecomposition analysis indicated theseanomalous structure phases seem to bemainly Al2Cu. Additionally, it was found that themorphology and composition of IMCschanged on the fracture surface, andcracks formed and propagated aloneIMCs. All of these revealed that the me-chanical property of aluminum/copper dissimilar friction welded jointswas controlled by the formation andgrowth of Al-Cu IMCs in this research.Furthermore, the cleavage fracture fea-ture was observed in all heat-treatedsamples, and in further magnifiedviews, the intergranular fracture modeturned to transgranular fracture modedue to PWHT.

Conclusions In this paper, the effect of PWHTon the mechanical property, mi-crostructure, and fracture feature offriction welded joints between com-mercially pure aluminum and copperbars was investigated. The main con-clusions are drawn as follows: 1) Sound joints were produced un-der PWHT at 100°C for 2 h with UTSof 91 MPa and a bending angle reach-ing 90 deg without breaking. The me-chanical property was impaired byPWHT with higher temperatures.Heat-treated samples failed at the in-terface during the tensile test, but theas-welded sample fractured on thealuminum side. 2) Copper was stirred into the alu-minum matrix due to mechanical ex-cavation during welding. After PWHTat 100° C for 2 h, the minimizedamount of Cu distributed on the alu-minum side, which enhanced the me-chanical property of aluminum/copper joints. Various Al-Cu IMCsformed and overgrew under PWHT at200° or 300°C for 2 h. 3) Fractographs showed Al-CuIMCs initially formed at one-third ofthe radius of the faying surface fromthe center due to the comprehensive

influences of rotation speed and fric-tion pressure. Additionally, IMCscould generate and propagate cracks.Moreover, all heat-treated samplespresented the cleavage fracture fea-ture, and in further magnified views,the intergranular fracture modeturned to transgranular fracturemode under the effect of PWHT.

This work was financially support-ed by the National Natural ScienceFoundation of China (Grant No.51374048 and 50904012); the Funda-mental Research Funds for the Cen-tral Universities (No. DUT13ZD209and DUT16RC(3)009); CollaborativeInnovation Center of Major MachineManufacturing in Liaoning; State KeyLaboratory of Advanced Welding &Joining, Harbin Institute of Technol-ogy (AWJ-M14-06); and State KeyLaboratory of Solidification Process-ing in NWPU (SKLSP201614).

1. Veerkamp, W. 1995. Copper-to-alu-minum transitions in high direct-currentbus systems. Petroleum and Chemical Indus-try Conference. Industry Applications Socie-ty, pp. 187–195. doi: 10.1109/PCI-CON.1995.523953. 2. Funamizu, Y., and Watanabe, K.1970. Interdiffusion in the Al-Cu system.Autumn Meeting of the Japan Institute ofMetal, Hiroshima, Japan, pp. 147–152. 3. Peng, X., Wuhrer, R., Heness, G., andYeung, W. 1999. On the interface develop-ment and fracture behaviour of roll bondedcopper/aluminium metal laminates. J.Mater. Sci. 34(9): 2029–2038. 4. Mai, T., and Spowage, A. 2004. Char-acterisation of dissimilar joints in laserwelding of steel-kovar, copper-steel andcopper-aluminium. Mater. Sci. Eng. A 374:224–233. 5. Xue, P., Ni, D., Wang, D., Xiao, B., andMa, Z. 2011. Effect of friction stir weldingparameters on the microstructure and me-

chanical properties of the dissimilar Al-Cujoints. Mater. Sci. Eng. A 528: 4683–4689. 6. Xue, P., Xiao, B., and Ma, Z. 2015. Ef-fect of interfacial microstructure evolutionon mechanical properties and fracture be-havior of friction stir-welded Al-Cu joints.Metall. Mater. Trans. A 46(7). 3091–3103. 7. Beygi, R., Kazeminezhad, M., Kokabi,A., and Loureiro, A. 2015. Friction stirwelding of Al-Cu bilayer sheet by taperedthreaded pin: Microstructure, materialflow, and fracture behavior. Metall. Mater.Trans. A 46(6): 2544–2553. 8. Meshram, S., Mohandas, T., and Red-dy, G. 2007. Friction welding of dissimilarpure metals. Journal of Materials ProcessingTechnology 184: 330–337. 9. Braunović, M., and Aleksandrov, N.1994. Intermetallic compounds at alu-minum-to-copper electrical interfaces: Ef-fect of temperature and electric current.IEEE Transactions on Components, Packag-ing, and Manufacturing Techology — Part A17(1): 78–85. 10. Yilbas, B., Sahin, A., Kahraman, N.,and Al-Garni, A. 1995. Friction welding ofSt-A1 and A1-Cu materials. Journal of Ma-terials Processing Technology 49: 431–443. 11. Lee, W., Bang, K., and Jung, S.2005. Effects of intermetallic compoundon the electrical and mechanical propertiesof friction welded Cu/Al bimetallic jointsduring annealing. Journal of Alloys andCompounds 390(1): 212–219. 12. Manjubhargavi, V., Chakravarthy, P.,Venugopal, P., and Achar, D. Influence ofrubbing profile on friction welding of pre-strained aluminum to copper. Second Inter-national Conference on Emerging Trends inEngineering and Technology, ICETET-092009, pp. 510–514. 13. Dong, H., Yu, L., Deng, D., Zhou,W., and Dong, C. 2015. Effect of post-weldheat treatment on properties of frictionwelded joint between TC4 titanium alloyand 40Cr steel rods. Journal of MaterialsScience & Technology 31(9): 962–968. 14. Xue, P., Xiao, B., Ni, D., and Ma, Z.2010. Enhanced mechanical properties offriction stir welded dissimilar Al-Cu jointby intermetallic compounds. Mater. Sci.Eng. A 527: 5723–5727. 15. Li, P., Li, J., Li, X., Xiong, J., Zhang,F., and Li, L. 2015. A study of the mecha-nisms involved in initial friction processof continuous drive friction welding. Jour-nal of Adhesion Science and Technology29(12): 1246–1257.

WELDING RESEARCH

WELDING JOURNAL / FEBRUARY 2017, VOL. 9670-s

References

Acknowledgments

Chen2016439_Layout 1 1/12/17 3:55 PM Page 70