XIII Portuguese Conference on Fracture (2012)

Fatigue strength improvement of AA6082-T651 MIG butt welds submitted to friction stir processing J.S. Jesus1, A.R. Loureiro 1, J. M. Ferreira 1, L.P. Borrego 1,2 and J.M. Costa11

CEMUC, Departamento de Engenharia Mecnica, Faculdade de Cincias e Tecnologia da Universidade de Coimbra, Rua Lus Reis Santos, Pinhal de Marrocos, 3030-788 Coimbra, Portugal. E-mail: cemuc@dem.uc.pt2

Departamento de Engenharia Mecnica, Instituto Superior de Engenharia de Coimbra, IPC, Rua Pedro Nunes, 3030-199 Coimbra, Portugal. E-mail: borrego@isec.pt ABSTRACT

Friction Stir Processing (FSP) was based on the principles of the Friction Stir Welding (FSW), a solid-state joining process originally developed for aluminium alloys. It is an emerging metalworking technique that can provide localized modification and control of microstructures in near-surface layers of processed metallic components. In this research the FSP will be applied on MIG butt welds with reinforcement and without reinforcement. FSP effect was studied through microstructural analysis, hardness, tensile strength and fatigue tests of welded specimens with butt joint geometry. Fatigue tests were carried out under constant amplitude loading with the stress ratio R set to 0. The completion of post processing by friction stir welding does not change significantly the hardness and mechanical strength of the weld, but slightly improves its ductility, Also, an improvement of fatigue strength was observed due to a refinement of microstructure and an eventual residual stresses decrease. KEYWORDS: Fatigue strength improvement , Friction stir processing, MIG welding, Aluminium alloy. 1. INTRODUCTION Friction stir welding (FSW) is a solid-state joining technique invented at the Welding Institute (TWI) (Cambridge, United Kingdom) in 1991 [1], for welding soft materials such as aluminum alloys. From this technique another process called Friction Stir Processing (FSP) was develop, Mishra et al. [2], for microstructural modification based on the principles of FSW. This technology involves plunging a rapidly rotating nonconsumable tool, comprising a profiled pin and larger diameter shoulder, into the surface and then traversing the tool across the surface for localized microstructural modification for specific property enhancement. Furthermore, the FSP technique has been used to create composite surfaces. FSP is a solid-state processing technique that with one step processing achieves microstructure refinement, densification and homogeneity. Besides, the microstructure and mechanical properties of the processed zone can be accurately controlled by optimizing the tool design, FSP parameters, and active cooling/heating [3]. In previous work [4], the fatigue resistance of AA6082 friction stir welds was analysed. Detailed examination revealed a hardness decrease in the thermo-mechanically affected zone and the nugget zone average hardness was found to be lower than the base alloy hardness. The comparison with data collected from the literature shows that FSW specimens present higher fatigue resistance than specimens welded by MIG and TIG processes, but have significantly lower lives than the base material. 1 Furthermore, the characteristic curve obtained for friction stir welds is higher than the International Institute of Welding (IIW) fatigue class for fusion welds with full-penetration both-sided butt joints. Another performed work by the same authors [5] analysed the validity of Minerss and Manson-Halfords damage rules applied to FSW welded specimens under constant and variable amplitude loadings with stress ratios R of 0 and -1. As expected, a significant mean stress influence was observed. The comparison of experimental fatigue lives with predictions calculated with both Minerss and Manson-Halfords Damage Rules, revealed a good agreement for R=0. Under R=-1 both damage predictions methods were in general unconservative. The principal objective of this study is to apply the FSP technique to obtain microstructural refinement, densification homogeneity and eliminate defects (like porosity) of the MIG butt welds in order to improve the fatigue strength. Moreover, other properties that are intrinsically tied to fatigue live such as hardness, yield stress and tensile strength, will be analysed. 2. EXPRIMENTAL DETAILS 2.1. Material used The material used for this research was the heat treated AA6082-T651 aluminium alloy. T6 agehardened condition is obtained by solution treatment, quenching and artificial age-hardening. Table 1 presents the alloy nominal chemical composition while Table 2 presents the main mechanical properties [4]. MIG welds were

XIII Portuguese Conference on Fracture (2012)

performed in plates of 333x80x6 mm. Table 1. Chemical composition of AA6082-T651 aluminium alloy (wt %)Si 1.05 Mg 0.8 Mn 0.68 Fe 0.26 Cr 0.01 Cu 0.04 Zn 0.02 Ti 0.01 Other 0.05

Welded specimens were post processed in two conditions: MIG butt welds with reinforcement (MIG_R) and welds without reinforcement, (excess of material deposited in MIG welding was removed by machining, MIG_NR), as illustrated in Fig. 3. a) b) Figure 3. MIG butt welds (front view): a) with reinforcement and b) without reinforcement. 2.3 Specimens post-processed by FSP In this section a description of the procedure applied during the post-processing using the technique of Friction Stir Processing is presented. A Cincinnati milling machine was used, which allows control of both rotation speed and forward speed of the table but did not allow control of the shoulder load. Two different procedures and non-consumable tools for post-processing were used for each type of specimens. Figures 4 and 5 present the geometries of the tools used for the postprocessing of MIG_NR and MIG_R specimens, respectively.

Table 2. Mechanical proprieties of AA6082-T651 aluminium alloy Tensile strength, urs (MPa) Yield strength, ys (MPa) Elongation, r (%) Hardness, Hv0.2 2.2. MIG welding and specimens preparation MIG welds were performed using a SAFMIG TRI 480 welding machine, with the weld torch mounted on a automatic tracking car running on a table where the coupon plates are fixed, as depicted in Figure 1. The filler metal was the AWS A5.10-80:ER 5356, with 5 % of magnesium and pure argon was used to shield the welds. 330 307 10 115

Figure 1. Car for automatic MIG welding The joint preparation as well as the welding parameters is indicated in Fig. 2 and Table 3, respectively. One layer was deposited in each side of the welds and an argon purge was used in the root of the first layer. Table 3. MIG welding parametersCurre nt (A) 160 Voltage (V) 26.5 Speed (cm/mi n) 50 Torch dist (mm) 15

Figure 4. Geometry of the tool used for post-processing the MIG_NR specimens (dimensions in mm)

6m m 2,5-3 m m 1,5 m m

Figure 2. Joint preparation. 2

XIII Portuguese Conference on Fracture (2012)

applied load cycle, respectively. 3. Results and discussion 3.1 Microstructures Fig. 6 illustrates the microstructures observed in a MIG weld before any post-processing. The base material is composed by grains elongated in the rolling direction, as shown in Fig. 6 a). This microstructure is partially destroyed in the heat affected zone as is illustrated in Fig, 6 b) and d). Melted zone is formed by equiaxed grains in the centre of the layers, see Fig. 6 c), and grains elongated following heat flow direction close fusion line, see Fig. 6 d), except in the region affected by second weld pass, where grain refinement was observed. Figure 5. Geometry of the tool used for post-processing the MIG_R specimens (dimensions in mm) The procedure and parameters of the post-processing were chosen based in a study of Mjali [5] and are shown in Table 4. Different post-processing procedures were adopted for each type of specimen. Table 4. Post-processing parametersSeries MIG_NR MIG_R Rotation speed (rpm) 500 1500 Forward speed (mm/min) 120 500 Number of passages 1 2

a)

b)

c)

d)

Figure 6. Metallographic analysis of MIG welding, a) base material, (BM) b) heat affected zone, (HAZ) c) fused zone, (FZ) and d) fused line (FL) Figure 7 shows images of the cross section a postprocessed MIG weld. The effect of post-processing is clearly shown in upper image of Fig. 7, where four postprocessing passes are visible. By comparing with Fig. 6 a) and b) a significant refinement of the microstructure is visible with post-processing. Fig. 7 b) illustrates the central zone of MIG weld unprocessed by FSP passes. Even in the thermo mechanically affected region, where the grain was plastically deformed but not recrystallized, see the upper left side of Fig. 7 c), important microstructural change was obtained.

2.4 Tensile strength-testing and fatigue testing After welding and post-processing, the specimens (20x160 mm) for tensile strength-testing and fatigue testing were removed transversely to the welding direction. Tensile tests were carried out in an Instron screw tensile/compression testing machine, model 4206, using a testing speed of 2 mm/min. Fatigue tests were carried out in an Instron hydraulic machine, applying a sinusoidal load wave to the specimens, with: 30 Hz of frequency, 0 stress ratio and stress ranges between 75 and 180 MPa. As for the realization of the fatigue tests was necessary introduce both load amplitude and mean load, they were calculated by equations 1 to 3. P[kN ] = [ MPa ] B[mm] W [mm] 1000 1+ R Pm = P 2( 1 R ) P Pa = 2 (1) (2) (3)

a)

b)

c)

where, B is the thickness in the welding zone, W the width of the specimen, R the stress ratio, the stress range and P, Pmax, Pmin Pm and Pa are the range, maximum, minimum, mean and amplitude values of 3

Figure 7. Metallographic analysis of the cross section of a post processed MIG weld: a) nugget; b) MIG melted unprocessed zone; c) thermo-mechanically affected zone/heat affected zone. 3.2 Hardness The hardness profiles of the welds before and after

XIII Portuguese Conference on Fracture (2012)

processing, measured in the cross section, are shown in Fig. 8. Fig. 8 shows that all the welds, unprocessed and processed, display a significant decrease in hardness in the thermo-mechanically affected zone. This decrease is due to the dissolution of strengthening precipitates in the thermo-mechanically affected zone and also coarsening of precipitates in the heat affected zone as mentioned elsewhere [7]. The completion of post processing does not alter significantly the hardness in welding, because for this aluminium alloy the main mechanism of hardening is not the refinement of the structure or plastic deformation, but the presence of hardening precipitates. This aspect is evidenced by the increase in width of the region with lower hardness in welds post-processed. In fact, as the post processing was performed in the area of the weld leg of either side, the width of the zone where the mentioned phenomena occurs, dissolution or coarsening of precipitates, increased, as shown in Figure 7. The greatest width of the observed loss of hardness of the post processed weld without reinforcement can be related to the greater amount of heat input in post processing, as suggested by the ratio of rotational speed versus forward speed of the tool, 4.1 and 3 respectively for MIG_NR and MIG_R. However the difference in widths is very small and can only be due to the positioning of the post processing tool. The lowest hardness values were observed in each side of the welds, in the region where the precipitates coarsened. 110

significant reduction of the yield stress and strength, under match condition, consistent with the reduction of hardness shown in Fig 8. This reduction of mechanical strength in fusion welds or friction stir welds in heat-treatable aluminium alloys is usual and has been widely reported in the literature [8, 9]. Table 5. Tensile testing resultsSeries MIG_NR MIG_R MIG_NR+FSP MIG_R+FSP Base material ys (MPa) 137 153 135 154 307 urs (MPa) 217 221 203 225 330 (max. load) 0,047 0,037 0,062 0,055 6 0,141 Rupture zone HAZ HAZ Retreating side Retreating side -

The local reduction of mechanical strength leads to the concentration of plastic deformation in the area less resistant, zone of the weld, leading to an overall reduction in the elongation of the specimens, when compared with those taken from the base metal. However, the post processing has a favourable effect on elongation, as illustrated in Table 5, essentially due to the increase of the area less resistant (see figure 8). When comparing MIG_NR and MIG_R series, the MIG_NR properties are lower than MIG_R because MIG_NR suffered a material reduction due the elimination of the reinforcement, and therefore a reduction of thickness in the weld zone. A slightly increase of resistance was also obtained for MIG_R+FSP vs MIG_R: 0.6 % for yield stress and 1.6 % for the ultimate tensile stress.250

2 )

100 90 80 70

MIG_NR+FSP MIG MIG_R+FSP

200 150 MIG_R MIG_NR MIG_NR+FSP MIG_R+FSP 0 0,02 0,04 0,06 0,08 (mm/mm) 0,1 0,12

m / f g K ( V s e n d r a H

60

( ) a P M-20 -15 -10 -5 0 5 10 15 Distance to center of welding(mm) 20 4

50

100 50 0

Figure 8. Hardness profiles of unprocessed and processed welds.

3.3 Tensile strength-testing The results of tensile tests of the unprocessed and processed welds are shown in Table 5 and Fig 9. It is observed in both the processed and unprocessed welds, a

Figure 9. Stress/strain curves for unprocessed and processed welds. 3.4 Fatigue resistance Figure 10 presents the fatigue results for the four series analysed, plotting the nominal stress range against the

XIII Portuguese Conference on Fracture (2012)

number of cycles to failure. The MIG_NR series is placed above the MIG_R series. Therefore, as expected an increase of the fatigue resistance was obtained for the MIG_NR series relatively to the MIG_R series due the elimination of the reinforcement in the MIG_R specimens.MIG_R

200 ] a P 150 M [m o n

MIG_NR MIG_R+FSP MIG_NR+FSP

fatigue stress concentration factors of the reinforced specimens, respectively. These parameters were statistically analysed in order to characterize the geometry of the welds of the MIG_R and MIG_R+FSP series, using a Gauss distribution. An average angle of 22.7 and an average curvature radius of 0.43 mm were obtained, both measured in the foot of MIG_NR welds. For the post-processed MIG_R+FSP series an average curvature radius of 1.33 mm was obtained. Table 6. summarizes the values obtained for kt and kf factors for MIG_NR and MIG_R+FSP series.Table 6 Static and dynamic stress concentration factors

e 100 g n a R s s e r t S 601.E+04 1.E+05 1.E+06 1.E+07

SeriesMIG_R MIG_R+FSP

Radius, (mm) 0.43 1.33

kt 1.86 1.46

kf 1.35 1.31

The local stress range at the weld toe loc can be calculated from the nominal stress range nom using the following equation,Number of cycles, Nr

loc = nom k f

(6)

Figure10. Results of fatigue tests: nominal stress range versus number of cycles to failure. Samples of the series MIG_NR+FSP broke up within a few cycles for a range of stress exceeding 120 MPa and below this stress range presented a high variability, unlike the series MIG_R specimens, which showed superior behavior, as illustrated in Fig . 10. This behavior may be related to the presence of defects in welding MIG_NR+FSP. With regard to the series MIG_R+FSP it presents superior behavior than the MIG_R series, which shows the favorable effect of post processing This increase of fatigue resistance can be explained by the action of several factors: i) geometry, especially due to the alteration of curvature radius; ii) due to a microstructure modification or porosity elimination; iii) an eventual decrease of residuals stress (not measured). In order to evaluate the influence of the geometry in the fatigue resistance the Lawrence equation (eq. 4) for butt welds [10] and Peterson equation (eq. 5) [11] were used to calculate kf and kt factors, respectively, t k t = 1 + 0,27 (tan ) 1 / 4 p k f = 1+1/ 2

Figure 11 present the fatigue results for the MIG_R, MIG_NR and MIG_R+FSP series plotting the local stress range against the number of cycles to failure.MIG_R 200 ] a P 150 M [co l

MIG_NR MIG_R+FSP

e g 100 n a R s s e r t S l a c o 60 L1.E+04 1.E+05 1.E+06 1.E+07

Number of cycles, Nr

(4)

Figure 11. Results of fatigue tests: local stress range versus number of cycles to failure. It is worthwhile to mention that the geometric factor is absent of these curves as they are plotted in terms of local stresses. MIG_R and MIG_NR series are nearly superimposed which indicates that the use of equation 6 is adequate. MIG_R+FSP series is clearly above MIG_NR and MIG_R series. Therefore, it can be concluded that the increase of about 25% observed in fatigue resistance was especially due to the improvement of microstructure and/or decrease of residuals stress.

k t 1 a 1+ a= 0.635 for aluminium alloys

(5)

where is the weld angle and is the weld toe radius of curvature, t the thickness and kt and kf the theoretical and

5

XIII Portuguese Conference on Fracture (2012)

4. Conclusions The completion of post processing by friction stir welding does not alter significantly the hardness and mechanical strength of the weld, but slightly improves its ductility. An improvement in fatigue life was obtained by post-processing MIG welds with Friction Stir Processing. The refinement of the microstructure and an eventual decrease of residual stress seem to be the causes for this improvement, although the simple removal of the weld reinforcement highly improves the fatigue resistance of MIG welds. ACKNOWLEDGEMENTS The authors thank Portuguese Foundation to Science and Technology for funding this work, Project n PTDC/EME-PME/114605/2009 co-financed by FEDER, through the Operational Factors for Competitiveness Programme of the QREN with reference COMPETE: FCOMP-01-0124-FEDER015152. REFERENCES [1] W.M. Thomas, E.D. Nicholas, J.C. Needham, M.G. Murch, Templesmith, and C.J. Dawes: G.B. Patent Application No.9125978.8, Dec. 1991. [2] R.S. Mishra and Z.Y. Ma: Mater. Sci. Eng., R, 2005, vol. R50,pp. 1 78. [3] Z.Y.MA. Friction Stir Processing Technology: A Review, DOI: 10.1007/s11661-007-9459-0, Volume 39A, March 2008657. [4] J.D. Costa, J.A.M. Ferreira, L.P. Borrego, L.P. Abreu. Fatigue behaviour of AA6082 friction stir welds under variable loadings, International Journal of Structural Integrity, Vol. 2 No. 2, 2011, pp. 122-134. [5] J.D. Costa, J.A.M. Ferreira, L.P. Borrego and L.P. Abreu. Fatigue behaviour of AA6082 friction stir welds under variable loadings. Int J Fatigue 37 (2012) 8-16. [6] K.V. Mjali. Analyzing The Effect of Friction Stir Processing on Mig-Laser Hybrid Welded AA 6082-T6 Joints, Magister Thesis, Nelson Mandela Metropolitan University, 2007. [7] Leal RM, Loureiro A. Microstructure and mechanical properties of friction stir welds in aluminium alloys 2024-T3, 5083-O and 6063-T6. Materials Science Forum, Vols 514-516, 2006, pp. 697-701. [8] Jata K.V, Sankaran K.K and Ruschau J.J. Frictionstir welding effects on microstructure and fatigue of aluminum alloy 7050-T7451. Metallurgical and Materials Transactions A, Volume 31, Number 9, 21812192, DOI: 10.1007/s11661-000-0136-9 [9] Mroczka K, Pietras A. FSW characterization of 6082 6

aluminium alloys sheets. Archives of Materials Science and Engineering. Vol. 40(2) 2009, pp. 104-109. [10] Yung J.Y. and Lawrence F.V. Analytical and Graphical aids for the fatigue design of weldments. Fatigue Fract Eng Mater Struct. 8-3 (1985) 223-241. [11] Peterson R.E. Relation between stress analysis and fatigue of metals. Proc. SESA 11 (1950) 199-206.