Joining of AZ91 Mg alloy and Al6063 alloy sheets by friction stir welding

B.Lkshmn Prsd,G.Neelih,M.Gopi Krishn,S.V.V.Rmn,K.Si Prksh,G.Srik,G.Prdeep Kumr Reddy,Rvikumr Dumpl,B.Rtn Sunil,∗

aDepartment of Mechanical Engineering,Rajiv Gandhi University of Knowledge Technologies(AP-IIIT),Nuzvid 520201,India

bDepartment of Mechanical Engineering,Vignana Bharathi Institute of Technology,Hyderabad 501301,India

cDepartment of Mechanical Engineering,Visvesvaraya National Institute of Technology,Nagpur 440010,India

1.Introduction

Obtaining permanent joint between two dissimilar metals is essential in developing low weight high strength structures with improved fuel economy in automobile applications.However,joining of dissimilar metals by fusion welding techniques is technically challenging in the welding industry[1].Aluminum(Al)alloys and magnesium(Mg)alloys are well known non-ferrous metallic systems which are being used widely in structural applications.The necessity of joining dissimilar metals is gradually being increased in modern designing and manufacturing industry.In this context,joining of two dissimilar non-ferrous metals can open wide scope for several applications in the manufacturing industry.Joining of dissimilar alloys of the same base material is difficult[2].If the base materials are chemically different for example Al-Mg,Al-steels and Mg-steels,then the welding of corresponding alloys is further complex[3,4].The melting point,rate of heat conduction,temperature dependent plastic deformation and crystal structure of base materials are a few influencing factors which make joining of Al alloys and Mg alloys as complex in nature by liquid state methods[1,3].On the other hand,friction stir welding(FSW),a solid state method developed to join similar and dissimilar metals without melting the material is gaining wide attention in solid state welding of dissimilar metals[5,6].

In Mg based dissimilar welding research,most of the works were done by using one of the AZ series Mg alloys as base material.Chen and Nakata[7]successfully joined AZ31 Mg alloy and zinc coated low carbon steel and reported that the joint was formed due to the formation of a low temperature melting eutectic phase which improved thediffusion between the magnesium alloy and the steel at the joint.Sato et al.[8]joined AA5083 aluminum alloy with AZ31 magnesium alloy and formation of Al3Mg2as embedded in a eutectic ofα-Mg and Mg17Al12was noticed with good lap shear strength of the joint.Similarly,Yamamoto et al.2009[9]also reported the formation of intermetallic layer consisting of Mg17Al12and Al3Mg2in the dissimilar weld zone of Al5083 and AZ31 alloys.Simoncini and Forcellese[10]successfully joined AA5754 and AZ31 thin sheets by FSW and observed lower stress and ductility levels for the welded samples compared with the base materials at all process parameters.Yan et al.[11]joined AZ31 Mg alloy sheets with 1060 Al alloy by FSW and the interface was observed with a lamellae structure.Furthermore,centerline crack was observed in the weld joint due to the formation of Al3Mg2and Al12Mg17phases in the weld region.Similar brittle mode fracture was also reported by Masoudian et al.[12]in the weld joint of AZ31-O Mg alloy with 6063-T6 Al alloy developed by FSW due to the complex mixture of the base materials in the stir zone.Additionally,the tensile strength of the weld material was measured as decreased compared with both the base materials.Formation of intermetallics(between Mg and Al)was also observed by Liyanage et al.[13]during the welding of Al6111 alloy with low carbon steel and AM60 Mg alloy with DP600 dual phase steel.

Table 1 Chemical composition of the base materials used in the present work.

In our previous work,AZ31 and Al6063 Mg alloys were successfully joined by FSW and deteriorated corrosion performance was observed at the joint due to galvanic corrosion[14].As observed in the previous work,the important mechanism by which the joint was established in the dissimilar metals is mechanical mixing of the material with and without alternate layers of base materials[15,16].However,the information on welding of AZ91 Mg alloy and Al alloys is insufficient in the literature.Therefore in the present study,AZ91 magnesium alloy and Al6063 aluminum alloy were selected to join by FSW.Different combinations of tool rotational speeds and tool travel speeds were adopted to obtain a successful joint.The important issues which dictate the material flow in dissimilar metal welding were studied with the help of the microstructural observations and microhardness measurement at the weld joint.

2.Experimental details

Fig.1.Photographs showing(a)the work pieces fixed on the work table of the universal milling machine and(b)FSW tool used in the present work.

AZ91 Mg alloy(Exclusive Magnesium,Hyderabad,India)and Al6063 alloy(Metro Metals,Vijayawada,India)sheets of size 100×50×4mm were prepared for the experiments.Chemical composition of the base materials are shown in Table 1.The workpieces were fixed on the work table of a universal milling machine(Bharat Fritz Werner Ltd.,India)and the process was carried out by using an FSW tool made of H13 tool steel with 15mm shoulder diameter consisting a tapered pin with 3-1mm taper for a distance of 3mm.Fig.1 shows the photograph of the equipment and FSW tool used in the present study.Welding was carried out at different tool rotational speeds(1100,1400,1600 and 1800rpm)and travel speeds(16,25 and 40mm/min).The process parameters were selected based on the literature and also considering the available speed and feed variations in the equipment used to carry out the experiments[17].For every combination of processing parameters,the weld joint was obtained and examined.Successful joint,without defects was obtained at 1100rpm with 16mm/min feed.

Specimens of size 30×10×4 mm3were cut across the weld direction and microstructural observations were carried out.Initially,the specimens were polished with different grades of emery papers up to 2000 grade followed by polishing with alumina paste.Then the specimens were polished with diamond paste using disc polishing equipment.After every step,the specimens were properly cleaned with ethanol to remove any residues resulted due to the previous polishing step.The polished specimens were chemically etched with picric acid reagent and optical microscope images were obtained using an inverted polarized optical microscope(Leica,Germany).For hardness measurements(Omnitech,India),the specimens of size 30×20×4 mm3were cut across the weld joint.Measurements were obtained for every 1mm in such a way that the base material,heat affected zone(HAZ),thermo mechanically affected zone(TMAZ)and nugget zones are measured.During measurement,100g load was applied for a dwell period of 15s.X-ray diffraction(XRD,Bruker advanced D8,USA)analysis was carried out for base materials and the weld joints.Scanning was carried out from 20°to 80°using CuKαradiation with 0.1°step size.

Fig.2.Photographs of the weld joints obtained at different tool rotational speed(rpm)and feed(mm/min):(a)1800rpm with 16mm/min,(b)1800rpm with 25mm/min,(c)1400rpm with 16mm/min,(d)1400rpm with 25mm/min,(e)1100rpm with 16mm/min and(f)1100rpm with 25mm/min,(g)photograph of cross section of 1100/16 weld joint and(h)photograph of cross section of 1100/25 weld joint.

Table 2 Weld joint status at different process parameters.

3.Results and discussion

The photographs of the weld joints obtained at different processing parameters are shown in Fig.2.The observations after welding are summarized in Table 2.From the results,the crucial role of tool rotational speed and travel speed on obtaining a successful dissimilar joint can be understood.When the tool rotational speed was 1800rpm,the generated heat was higher and excess material flush was noticed along with deposition of some base material to the FSW tool shoulder.When the feed is more(25mm/min),the material flow was observed as poor and hence,resulted in a deep groove at the joint as shown in Fig.2(b).By decreasing the tool rotational speed to 1400rpm,with 16mm/min feed,similar behavior was observed at the weld region and no joint was noticed.

Interestingly,with 25mm/min feed,marginally favorable outcome was observed as shown in Fig.2(d).During welding,it was observed the weld region as a perfect metallurgical joint.However,as the weld joint slowly cooled to the room temperature,a crack appeared in the weld joint which was classified as hot crack as indicated with an arrow in Fig.2(d).Similarly,at 1100rpm and 16mm/min,the joint appeared to be formed during the process.However,as soon the joint reached the room temperature,a thin hot crack appeared similar to that of 1400rpm/25mm/min weld joint.Hot crack resulted in fusion welding processes during the solidification and immediately after the solidification[2].As the liquid state metal at the weld joint starts to solidify,the shrinkage during the solidification leads to raise the residual stresses at the vicinity of the joint and cracks are produced if these stresses are higher or imbalanced[1].Whereas in the solid state welding such as FSW,the heat generated at the joint must dissipate through the workpiece,FSW tool and the ambient air[18].In this process major fraction of heat dissipation occurred through the workpiece and the tool.Therefore,there is a variation in temperatures at different regions of processed and unprocessed material;thermal conductivity nature of the base material certainly results in the development of hidden stresses.If the base materials are dissimilar as in the case of the present study,two different thermal conductivities at the weld region make the residual stresses more complex and lead to the development of hot crack during the cooling of the joint.This is the reason behind the development of hot cracks as observed for the joints which appeared as perfect during the process and defective after cooling to room temperature as shown in Fig.2(d)and(e).Whereas the resulted heat and the material flow was optimum which eliminated the formation of hot crack and led to formation of a defect free joint as shown in Fig.2(f)at 1100rpm with 25mm/min feed.The cross section of the weld joints obtained at 1100rpm with 16 and 25mm/min feed appeared without any defect or crack as shown in Fig.2(g)and(h).

Fig.3.Optical microscope images at the joint:(a)weld zone showing interface with AZ91 Mg alloy,(b)corresponding magnified image,(c)interface at Al6063 alloy,(d)stir zone showing mixing of material and(e)corresponding magnified image.

Fig.4.Optical microscope images:(a)Al6063 base material,(b)stir zone at Al6063 side,(c)AZ91 base material and(d)stir zone at AZ91 side.

Fig.3 shows the optical microscope images of the weld joint obtained at 1100rpm and 25mm/min feed.From the microstructures,it can be seen that the joint is not from a perfect metallurgical bonding.Instead,it can be understood that the mechanical mixing of two dissimilar metals with lower level of metallurgical bonding resulted in the joint between AZ91 Mg alloy and Al6063 alloy.The stir zone of the weld joint indicates the mixing of both the materials with a sharp interface on both sides of the base materials.The network shaped secondary phase in AZ91 Mg alloy as indicated by arrows in Fig.3(b)was observed as greatly affected,in the stir zone.Close to the Al6063 base material,the plastically deformed base material with minimum dissolution can be observed along with the nugget zone.The microstructural observations carried out at another location of the weld joint(Fig.3(d))shows the material flow where the mixture of both the base materials and distorted secondary phase with elongated grains can be seen as shown in the magnified image(Fig.3(e)).The material from Al6063 was observed as mechanically mixed with AZ91 at the stir zone and established a joint.Fig.4 shows optical microscope images of base Al6063 and base AZ91 and corresponding base and stir zone interfaces.

Fig.5.Optical microscope images of AZ91 Mg alloy:(a)base material and(b)stir zone close to AZ91 Mg alloy sheet.

Fig.6.XRD patterns of the samples collected from base materials and the weld joint.

Fig.5 compares the microstructure of base and stir zone at the side of AZ91 Mg alloy.Before FSW(Fig.5(a)),theα-grains(as indicated by black arrows)contain less than 1%of aluminum.The excess amount of aluminum results in formation of network shaped Mg17Al12as indicated by white arrows.It is interesting to observe the reduced amount of secondary phase in the stir zone(Fig.5(b))with smaller grain size(11.4μm)compared with AZ91 Mg alloy base material grain size(160.2μm).Similarly,the grain size was measured as 17μm in the stir zone of Al6063 which is smaller compared with the starting grain size of Al6063(94.4μm).

Fig.6 shows the XRD patterns of both the base materials and the weld joint.From the XRD plots of base materials,all peaks were identified and indexed by comparing with the standard data(for Mg-JCPDS No 35-0821 and for Al-JCPDS No 04-0787).The XRD of weld joint has shown the presence of peaks corresponding to both the base materials.No new peaks were identified in the weld zone but the peak intensities were observed as increased after FSW.Particularly,intensity of(002)peak corresponding to Mg was observed as significantly increased which is a clear indication of texture effect induced due to FSW in the stir zone.Mg ishcpcrystal structure metal and slip occurs in(002)basal plane during the plastic deformation and hence,in the FSW region,higher intensity of(002)was observed.Similarly,intensity of(111)plane corresponding to aluminum(fcccrystal structure)was relatively increased in the weld zone which suggests the texture effect in aluminum after FSW.Close packed planes which are(002)inhcpmetals and(111)infccmetals are the highest density planes that play a crucial role in material properties[19].In the present study,weld zone contains higher fraction of these close packed planes compared with base materials which significantly improves the mechanical and corrosion behavior.Overall,XRD analysis confirms the presence of material from both the base materials in the weld zone and demonstrates the occurrence of preferred orientation due to FSW.

The distribution of microhardness across the joint is shown in Fig.7.In AZ91 Mg alloy,the hardness variations are usually observed as more due to the presence of two regions(i)solid solution of Mg and Al and(ii)compound of Mg and Al(Mg17Al12).Therefore,the range of hardness varied from lower at the grains to higher at the secondary phase present at the grain boundary as reported in our earlier study[20].Interestingly in the present work,the stir zone has lower variation in the hardness values.From the microstructural studies,the amount of secondary phase was observed as severely decreased in the stir zone.Therefore,the variation in hardness was decreased and the overall hardness was increased due to the grain refinement in the stir zone.A sudden drop in the hardness can be seen at the Al6063 side which is obvious as a sharp interface was observed at the region of nugget zone-Al6063 base material as shown in Fig.3(c).Hence,from the preliminary results,it can be understood that AZ91 Mg alloy and Al6063 alloy sheets can be joined by FSW and further investigation on mechanical performance and corrosion behavior subjected to different load conditions needs to be investigated.

Fig.7.Microhardness distribution across the weld joint.

4.Conclusions

In the present work,AZ91 Mg alloy and Al6063 alloy sheets were successfully joined by friction stir welding.It was demonstrated that the tool rotational speed and travel speed play a crucial role in obtaining a good weld joint between two dissimilar metals.1100 r.p.m.of tool rotational speed with 25mm/min tool travel speed were found as optimum parameters for the FSW tool with 15mm shoulder diameter with 3 to 1mm tapered pin of 3mm length.The joint between AZ91 Mg alloy and Al6063 alloy was observed as mainly due to the mechanical mixing with a poor metallurgical bonding.Microstructural observations and microhardness measurements revealed a sharp interface at the region of stir zone and Al6063 base material.Additionally,secondary phase from AZ91 Mg alloy was observed as decreased with smaller grains in the stir zone.No significant new phases were observed after FSW from the XRD studies.Further investigations on assessing the effect of presence of dissimilar metals in a weld joint on mechanical behavior and corrosion behavior of the joints are suggested.

Acknowledgment

The authors would like to thank Mr.Khan and Mr.Chandrasekhar,Lab assistants,Department of Mechanical Engineering for helping in the experiments.

[1]R.W.Messler Jr.,Principles of Welding:Processes,Physics,Chemistry,and Metallurgy,Wiley,USA,1999.

[2]B.Mvola,P.Kah,J.Martikainen,Int.J.Mech.Mater.Eng.9(2014)21.

[3]Y.V.Budkin,Weld.Int.25(7)(2011)523-525.

[4]M.Aonuma,K.Nakata,Mater.Sci.Eng.B 173(1)(2010)135-138.

[5]R.S.Mishra,Z.Y.Ma,Mater.Sci.Eng.R 50(2005)1-78.

[6]T.DebRoy,H.K.D.H.Bhadeshia,Sci.Technol.Weld.Joint.15(4)(2010)266-270.

[7]Y.C.Chen,K.Nakata,Mater.Trans.50(11)(2009)2598-2603.

[8]Y.S.Sato,A.Shiota,H.Kokawa,K.Okamoto,Q.Yang,C.Kim,Sci.Technol.Weld.Joint 15(4)(2010)319-324.

[9]N.Yamamoto,J.Liao,S.Watanabe,K.Nakata,Mater.Trans.50(2009)2833-2838.

[10]M.Simoncini,A.Forcellese,Mater.Des.41(2012)50-60.

[11]J.Yan,Z.Xu,Z.Li,L.Li,S.Yang,Scr.Mater.53(2005)585-589.

[12]A.Masoudian,A.Tahaei,A.Shakiba,F.Sharifianjazi,J.Aghazadeh Mohandesi,Trans.Nonferrous Met.Soc.China 24(2014)1317-1322.

[13]T.Liyanage,J.Kilbourne,A.P.Gerlich,T.H.North,Sci.Technol.Weld.Joint.14(6)(2009)500-508.

[14]B.Ratna Sunil,G.Pradeep Kumar Reddy,Cogent Eng 3(1)(2016)1145565.

[15]B.Ratna Sunil,G.Pradeep Kumar Reddy,A.S.N.Mounika,P.Navya Sree,P.Rama Pinneswari,I.Ambica,R.Ajay Babu,P.Amarnadh,J.Magnesium Alloys 3(4)(2015)330-334.

[16]N.Kumar,R.S.Mishra,W.Yuan,Friction Stir Welding of Dissimilar Alloys and Materials,Elsevier,USA,2015.

[17]V.Firouzdor,S.Kou,Weld.J.88(2009)213s-223s.

[18]R.S.Mishra,M.W.Mahoney,Friction Stir Welding and Processing,ASM International,USA,2007.

[19]G.E.Dieter,Mechanical Metallurgy,McGraw Hill Education,New York,USA,1989.

[20]B.Ratna Sunil,K.V.Ganesh,P.Pavan,G.Vadapalli,Ch.Swarnalatha,P.Swapna,P.Bindukumar,G.Pradeep Kumar Reddy,J.Magnesium Alloys 41(1)(2016)15-21.