Microstructure characterization of the stir zone of submerged friction stir processed aluminum alloy 2219

M A TE RI A L S CH A R A CT ER IZ A TI O N 8 2 (2 0 1 3 ) 9 7–1 0 2

Available online at www.sciencedirect.com

www.elsevier.com/locate/matchar

Microstructure characterization of the stir zone of submerged friction stir processed aluminum alloy 2219 Xiuli Fenga,b,⁎, Huijie Liub , John C. Lippolda a

Welding Engineering Program, Department of Materials Science and Engineering, The Ohio State University, Columbus, OH 43221, USA State Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, Harbin 150001, China

b

AR TIC LE D ATA

ABSTR ACT

Article history:

Aluminum alloy 2219-T6 was friction stir processed using a novel submerged processing

Received 9 March 2013

technique to facilitate cooling. Processing was conducted at a constant tool traverse speed

Received in revised form

of 200 mm/min and spindle rotation speeds in the range from 600 to 800 rpm. The

12 May 2013

microstructural characteristics of the base metal and processed zone, including grain

Accepted 19 May 2013

structure and precipitation behavior, were studied using optical microscopy (OM), scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Microhardness maps were constructed on polished cross sections of as-processed samples. The effect of

Keywords:

tool rotation speed on the microstructure and hardness of the stir zone was investigated.

Microstructure

The average grain size of the stir zone was much smaller than that of the base metal, but

Submerged friction stir processing

the hardness was also lower due to the formation of equilibrium θ precipitates from the

Al–Cu alloy

base metal θ′ precipitates. Stir zone hardness was found to decrease with increasing

Microhardness

rotation speed (heat input). The effect of processing conditions on strength (hardness) was rationalized based on the competition between grain refinement strengthening and softening due to precipitate overaging. © 2013 Elsevier Inc. All rights reserved.

1.

Introduction

Friction stir welding (FSW) has proven to be an efficient, reliable, and environment friendly joining method for aluminum alloys [1,2]. During FSW, a rotational tool travels along the length of the abutting plates to be welded, and produces a high plastically deformed region through the associated stirring action produced by the tool. Friction stir processing (FSP), developed based on FSW, is emerging as an effective thermal–mechanical process to refine grains, achieve superplasticity, modify microstructures, and synthesize in-situ composite and intermetallic compounds [3–12]. As opposed to FSW, FSP is conducted on a single workpiece and is not designed as a joining method. During FSW/FSP, localized heating is generated by friction at the tool/workpiece interface, as well as from the plastic deformation of the workpiece. The combination of severe plastic deformation and high

temperature exposure, promotes dynamic recrystallization in the stir zone and resultant grain refinement. The grain size in the stir zone of FSW/FSP of aluminum alloys without external cooling has been reported in the range of 2–10 μm [13–16]. Since fine-grained microstructures are generally beneficial with respect to mechanical properties, there have been many attempts to achieve even finer grain size. Several researchers have reported that the grain size can be further refined by applying in-process or external cooling during FSW/FSP. Benavides et al. [17] performed FSW trials of 2024 aluminum alloy using liquid nitrogen to lower the initial temperature of the samples to −30 °C. They reported that the size of the grains was reduced to less than 0.8 μm when the starting workpiece temperature was reduced to 173 K (− 100 °C). Rhodes et al. [18] used a mixture of dry ice and isopropyl alcohol on the surface of the processed 7050 Al plate in an

⁎ Corresponding author at: 1248 Arthur E Adams Drive, Columbus, OH 43221, USA. Tel.: +1 614 390 2358; fax: + 1 614 688 3333. E-mail addresses: [email protected], [email protected] (X. Feng), [email protected] (H. Liu), [email protected] (J.C. Lippold). 1044-5803/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.matchar.2013.05.010

98

M A TE RI A L S CH A R A CT ER IZ A TI O N 8 2 (2 0 1 3) 9 7–1 0 2

attempt to “freeze in” the newly recrystallized microstructures and found extremely fine grains (50–100 nm). Hofmann and Vecchio performed submerged FSP on 6061-T6 Al alloy and obtained grains smaller than 200 nm [19]. Su et al. [20] obtained fine grains in the range of 100–400 nm by using a mixture of water, methanol, and dry ice to quench the friction stir processed 7075 Al plates immediately behind the tool. Similar to Su et al., Liu et al. [21] used room temperature water to quench the processed 7075 Al and obtained grains with an average diameter of 800 nm. Although fine grains are beneficial to the strength of the stir zone, a loss of strength is often reported in the stir zone and HAZ by FSW/P of precipitation strengthened aluminum alloys. This could be caused by precipitate dissolution and transformation to an equilibrium state and coarsening. Woo and Choo found that softening in the stir zone of FSW joints of 6061-T6 Al alloy was primarily caused by precipitate dissolution [22]. It was reported by Dong et al. that the softening in the stir zone of 6005-T6 Al alloy was caused by the dissolution of β″ during welding and limited re-precipitation during cooling [23]. They also reported that the softening in the HAZ was a result of the coarsening of the β′ and Q′ precipitates. Different precipitate evolution mechanisms were related to the thermal cycles during FSW. Therefore, in order to control the precipitate evolution, fast cooling was applied to limit the thermal effect during FSW. Fratini et al. [24] conducted in-process water-cooling on the top surfaces of weld samples during FSW and found that material softening in the thermo-mechanically affected zone (TMAZ) was reduced by 10–40 HV. Zhang et al. [25] performed underwater FSW on 2219-T6 Al alloys and found that the tensile strength of the samples was higher than that of the joint cooled in air, confirming the feasibility of improving the joint properties by water cooling. Furthermore, it was also reported that low tool rotational speed which leads to lower heat input, will result in lower stir zone temperature during FSW/FSP [25]. This can reduce the rate of precipitate dissolution or coarsening. However, the effect of the tool rotational speed and associated cooling rate on the microstructure and hardness in the stir zone of FSPed Al–Cu alloys has not been systematically investigated. In the present study, submerged friction stir processing was conducted on a precipitation strengthened Al–Cu alloy at a constant travel speed over a range of tool rotational speeds. Microstructure evolution in the stir zone, including grain structure and precipitate morphology, has

been analyzed. Hardness of the stir zone was also measured to quantify the effect of tool rotational speed and heat input on softening kinetics.

2.

Experimental Procedures

The submerged friction stir processing setup and the tool are shown in Fig. 1. Fig. 1a defines the processing coordinates. The processed plate was fitted into a tank, which was filled with room temperature water (Fig. 1b). A backing plate made of steel was used. The tool was a standard tool steel, consisting of a shoulder with a diameter of 22 mm and no pin feature, as shown in Fig. 1c. The base material used was 300 × 300 × 2.5 mm 2219-T6 aluminum alloy. The chemical composition (wt.%) of the base metal was Al–6.48Cu–0.32Mn– 0.23Fe–0.06Ti–0.08 V–0.04Zn–0.49Si–0.2Zr. The tool rotational speed (ω) was varied in the range of 600–1000 rpm at a constant processing speed (v) of 200 mm/min. Displacement control was used for tool plunge with a plunge depth of 0.5 mm. The tilt angle of the rotational tool with respect to the Z-axis of the FSW machine was 2.5°. Microstructure characterization was performed in two orthogonal directions of the friction stir processed plates, transverse cross section and plan view. The cross-section (YZ plane in Fig. 1a) of the plates was analyzed using Optical Microscopy (OM). Polished samples were etched with Keller's reagent (a mixture of 2.5 ml nitric acid, 1.5 ml hydrochloric acid, 1 ml hydrofluoric acid and 95 ml water) for 20 s. Vickers microhardness was measured on the polished cross-section using an automated tester under a load of 50 g for a dwelling time of 10 s. The center-to-center distance between the adjacent indents was 120 μm. The plan view (XY plane in Fig. 1a) of the samples was characterized using transmission electron microscopy (TEM). The processed surface was lightly ground to remove the periodic surface bands from tool rotation. Samples were then mechanically ground on the unprocessed side (back side) until the sample thickness was reduced to 80 μm. Disks with a diameter of 3 mm were punched from the center position of the thin foils and subjected to twin-jet electropolishing in a solution of 30% nitric acid and methanol at − 20 °C. The TEM samples were examined using the Philips CM200 microscope operating at 200 kV to analyze the grain structure and precipitate morphology and distribution.

Fig. 1 – Submerged friction stir processing and the tool: a) schematic drawing of FSP, b) set up of submerged friction stir processing, c) the tool.

M A TE RI A L S CH A R A CT ER IZ A TI O N 8 2 (2 0 1 3 ) 9 7–1 0 2

3.

Results and Discussion

3.1.

Microstructure

Optical and scanning electron microscopy images of the base metal are shown in Fig. 2. The average grain size of the base metal was determined to be 17 μm in an earlier publication [26]. The large white particles in the base metal were determined to be Al2Cu and had no significant contribution to the strengthening of the alloy [27]. Fig. 3 shows the macrostructure of the processed zone under different tool rotational speeds. All stir zones were defect free. The microstructure in the stir zones was homogeneous and the grain size was highly refined relative to the base metal. Grain size and precipitation behavior in the stir zone could not be quantified using optical microscopy. Using TEM, a more detailed characterization is shown in Fig. 4. As expected, the size of the stir zone decreases as the tool rotational speed decreases. This is caused by the reduction of the heat input with the decrement of the tool rotational speed. As the heat input decreases, the plasticized material region is narrowed, which depresses the material flow and results in a reduced stir zone. Fig. 4 shows the TEM images of the base metal and stir zones obtained at tool rotational speeds of 600, 800, and 1000 rpm. High dislocation density is absent within the grains, indicating that dynamic recrystallization occurs in the stir zone. Texture analysis shows that the possible recrystallization mechanisms include continuous dynamic recrystallization, geometric dynamic recrystallization and particle stimulate nucleation [26]. From Fig. 4a–c, it can be seen that the grain size decreases as the tool rotational speed decreases. When the speed is 1000 rpm, most grains have a size of 1 μm and above, while a small fraction is in the range from 500 to 800 nm. The average grain size was found to be 1.3 μm using TEM images, which indicates an order of magnitude reduction relative to the base metal. With the decrease of rotational speed, the fraction of small grains increases. When the rotational speed is 800 rpm, grains with the size of 1 μm and above are still present, but the fraction of small grains increases. The average grain size is estimated to be 1.1 μm. When the speed is 600 rpm, most of the grains are less than 1 μm, and the average grain size is estimated to be

99

0.8 μm. This degree of grain refinement is a result of dynamic recrystallization [28]. The precipitate characteristics in the base metal and stir zone can also be observed in Fig. 4. From Fig. 4d, the precipitates in the base metal have a plate shape. By indexing the selected area diffraction pattern, these precipitates were identified as meta-stable precipitates θ′. This precipitate is semi-coherent with the aluminum matrix. However, in the stir zones of the three as-processed samples (Fig. 4a–c), precipitates of this morphology are not observed. Instead, the stir zone precipitates have a disc shape and are identified as the equilibrium precipitate θ. This identification was based on indexing the selected area diffraction pattern and dark field imaging in the previous paper [26]. The θ precipitate is non-coherent with the aluminum matrix and does not have a specific orientation relationship with the matrix. It is shown that the size of the equilibrium precipitate θ decreases with decreasing tool rotational speed, and as the speed decreases, the spacing between the precipitates decreases. The phenomenon of equilibrium precipitates substituting for meta-stable precipitates has been reported by others as well [29–31]. The main precipitate evolution during FSW/FSP includes precipitate dissolution and overaging. In this study, the majority of the precipitates overage to equilibrium precipitates, not dissolved into the matrix, as evidenced by the presence of the large amount of equilibrium precipitates.

3.2.

Microhardness

Fig. 5 shows the microhardness maps in the stir zone under different tool rotational speeds. Softened stir zones relative to the base metal are observed for all three conditions. This is because the precipitate strengthening effect is dramatically reduced due to the loss of semi-coherent relationship between the precipitate and the matrix. Compared with the unprocessed base metal, which has a hardness of 138 HV, the average microhardness in the stir zone is 90 HV, 95 HV and 100 HV, respectively, corresponding to the tool rotational speed of 1000 rpm, 800 rpm and 600 rpm. The increment in hardness with the decrease in tool rotational speed is mainly caused by the difference in precipitate strengthening effect, which will be quantitatively calculated in this paper.

Fig. 2 – Microstructure of the base metal: a) optical, b) SEM.

100

M A TE RI A L S CH A R A CT ER IZ A TI O N 8 2 (2 0 1 3) 9 7–1 0 2

Fig. 3 – Macrostructure of the friction stir processed samples: a) 1000 rpm, b) 800 rpm, c) 600 rpm.

3.3. Strengthening Effect of Grain Refinement and Precipitation As stated above, softened regions were observed in the stir zones for all three conditions. The reason for softening is explained quantitatively in this paper. The main microstructure differences between the base metal and the stir zone in this study are grain size and precipitates. Therefore, the grain refinement strengthening and precipitation strengthening

effect, i.e., the yield strength contributions arising from grain refinement and precipitation are calculated. The yield strength increment due to grain refinement is calculated using the Hall–Petch relationship: k σ gb ¼ σ 0 þ pffiffiffi d

ð1Þ

where σgb is the increase in yield strength due to fine grain strengthening, σ0 is the friction stress (20 MPa), k is a constant

Fig. 4 – TEM images of stir zones processed at a tool rotational speed of: a) 1000 rpm, b) 800 rpm, c) 600 rpm, d) base metal.

101

M A TE RI A L S CH A R A CT ER IZ A TI O N 8 2 (2 0 1 3 ) 9 7–1 0 2

Fig. 5 – Microhardness maps of friction stir processed samples under different tool rotational speeds: a) 1000 rpm, b) 800 rpm, c) 600 rpm.

(0.04 MPa · m1/2) [32] and d is the average grain size. In this study, the average grain size for the stir zone is 1.3 μm, 1.1 μm and 0.8 μm corresponding to the tool rotational speed of 1000 rpm, 800 rpm and 600 rpm, respectively. Using Eq. (1), the strength contribution from grain boundary strengthening to the yield strength is calculated to be 55 MPa, 58 MPa and 65 MPa when the tool rotational speed is 1000 rpm, 800 rpm and 600 rpm, respectively. With the decrease of the tool rotational speed, the grain refinement strengthening effect increases. The grain refinement strengthening contribution to the base metal yield strength is 30 MPa. Therefore, compared to the base metal, the yield strength increase due to grain refinement effect after FSP is 25–35 MPa. The yield strength increase due to precipitation strengthening in the base metal was calculated to be 123 MPa in the previous paper [26]. As the precipitates in the stir zone are equilibrium θ, the precipitation strengthening effect is calculated using the Orowan–Ashby equation [33]: Δσ ppt ¼

0:13Gb r ln λ b

ð2Þ

where λ is the inter-particle spacing, r is the particle radius, G is the shear modulus and b is the Burgers vector (2.84 × 10−10 m for aluminum) [34]. Using the TEM images (Fig. 4a–c), the inter-particle spacing and radius of the precipitates are measured. The calculated yield strength increase due to precipitation strengthening is summarized in Table 1. As indicated in Table 1, the strengthening effect of precipitation in the stir zone is 16, 38 and 43 MPa corresponding to the

Table 1 – Yield stress increment due to precipitates strengthening in FSP regions. ω (rpm)

λ (nm)

r (nm)

Δσppt (MPa)

1000 800 600

337 140 115

100 75 50

16 38 43

tool rotational speed of 1000, 800 and 600 rpm. The summary of the grain refinement and precipitation effects, and the microhardness in the stir zone and the base metal is shown in Table 2. The yield strength contributions arising from grain refinement increase with the decrease of the tool rotational speed, while softening due to the transformation of the θ′ precipitates to θ increases at higher rotational speeds. This explains the reason for the higher SZ hardness with a decreasing tool rotational speed. Comparing with the base metal, the yield strength contribution from precipitation strengthening in the stir zone decreases 80–107 MPa. But the yield strength contribution from grain refinement in the stir zone only increases 25–35 MPa compared with the base metal. Thus, the increase in strength due to grain refinement cannot offset the decrease due to precipitate coarsening. For this reason, the stir zone cannot achieve the strength of the precipitation strengthened base metal. This suggests that a simple aging treatment following FSW/FSP will not be capable of restoring the strength of the stir zone to that of the base metal, as the θ present will only coarsen. Additionally, attempting to solutionize the θ so that it can be reprecipitated will cause recovery and coarsening in the SZ.

4.

Conclusions

In this study, the microstructure of the stir zone in submerged friction stir processed 2219-T6 aluminum alloy was evaluated, including the grain structure and the nature of precipitates.

Table 2 – Summary of strengthening (softening) effects. ω (rpm)

σgb (MPa)

Δσppt (MPa)

Microhardness (HV)

600 800 1000 Base metal

65 58 55 30

43 38 16 123

100 95 90 138

102

M A TE RI A L S CH A R A CT ER IZ A TI O N 8 2 (2 0 1 3) 9 7–1 0 2

This was compared to the stir zone hardness. The main conclusions are as follows. 1) The stir zone grain size was an order of magnitude less than that of the base metal. Ultra-fine grains (less than 1 μm) are obtained in the stir zone and the area fraction of ultra-fine grains increases as the tool rotational speed (heat input) decreases. 2) The precipitates in the stir zone are primarily disc shaped equilibrium θ, which are different from the meta-stable plate shaped equilibrium θ′ in the base metal. The strengthening effect of the equilibrium precipitate θ is much smaller than that of meta-stable precipitate θ′ because of the loss of coherency and increased size and spacing between adjacent precipitates. 3) Softening of the stir zone relative to the 2219-T6 base metal was observed under all conditions. The degree of softening decreased as the tool rotational speed (heat input) decreased. 4) Quantitative analysis showed that the softening effect of precipitate coarsening (80–107 MPa) overwhelms the strengthening effect of grain refinement (25–35 MPa). 5) The dominant effect of stir zone softening during submerged (accelerated cooling) FSW/FSP of 2219-T6 is the formation of incoherent equilibrium θ precipitates from the base metal θ′ precipitates, rather than the dissolution of the θ′ precipitates.

REFERENCES

[1] Mishra RS, Ma ZY. Friction stir welding and processing. Mater Sci Eng R 2005;50:1–78. [2] Thomas WM, Nicholas ED, Needam JC, Murch MG, Templesmith P, Dawes CJ. GB Patent Application No. 9125978.8, December 1991 and US Patent No. 5460317;1995 October. [3] Mishra RS, Mahoney MW. Friction stir processing: a new grain refinement technique to achieve high strain rate superplasticity in commercial alloys. Mater Sci Forum 2001;357–359:507–14. [4] Kwon YJ, Saito N, Shigematsu I. Friction stir process as a new manufacturing technique of ultrafine grained aluminum alloy. J Mater Sci Lett 2002;21:1473–6. [5] Chang CI, Du XH, Huang JC. Achieving ultrafine grain size in Mg– Al–Zn alloy by friction stir processing. Scr Mater 2007;57:209–12. [6] Chang CI, Du XH, Huang JC. Producing nanograined microstructure in Mg–Al–Zn alloy by two-step friction stir processing. Scr Mater 2008;59:356–9. [7] Liu FC, Ma ZY. Achieving exceptionally high superplasticity at high strain rates in a micrograined Al–Mg–Sc alloy produced by friction stir processing. Scr Mater 2008;59:882–5. [8] Ma ZY, Mishra RS, Liu FC. Superplastic behavior of micro-regions in two-pass friction stir processed 7075Al alloy. Mater Sci Eng A 2009;505:70–8. [9] Alidokht SA, Zadeh AA, Soleymani S, Saeid T, Assadi H. Evaluation of microstructure and wear behavior of friction stir processed cast aluminum alloy. Mater Charact 2012;63:90–7. [10] Mishra RS, Ma ZY, Charit I. Friction stir processing: a novel technique for fabrication of surface composite. Mater Sci Eng A 2003;341:307–10. [11] Morisada Y, Fujii H, Nagaoka T, Fukusumi M. MWCNTs/AZ31 surface composites fabricated by friction stir processing. Mater Sci Eng A 2006;419:344–8.

[12] Morisada Y, Fujii H, Nagaoka T, Fukusumi M. Effect of friction stir processing with SiC particles on microstructure and hardness of AZ31. Mater Sci Eng A 2006;433:50–4. [13] Murr LE, Liu G, McClure JC. Dynamic recrystallization in friction-stir welding of aluminium alloy 1100. J Mater Sci Lett 1997;16:1801–3. [14] Jata KV, Semiatin SL. Continuous dynamic recrystallization during friction stir welding of high strength aluminum alloys. Scr Mater 2000;43:743–9. [15] Su JQ, Nelson TW, Mishra RS, Mahoney M. Microstructural investigation friction stir welded 7050-T651 aluminium. Acta Mater 2003;51:713–29. [16] Cabibbo M, McQueen HJ, Evangelista E, Spigarelli S, Di Paola M, Falchero A. Microstructure and mechanical property studies of AA6056 friction stir welded plate. Mater Sci Eng A 2007;460–461:86–94. [17] Benavides S, Li Y, Murr LE, Brown D, McClure JC. Low-temperature friction-stir welding of 2024 aluminum. Scr Mater 1999;41:809–15. [18] Rhodes CG, Mahoney MW, Bingel WH, Calabrese M. Fine-grain evolution in friction-stir processed 7050 aluminum. Scr Mater 2003;48:1451–5. [19] Hofmann DC, Vecchio KS. Submerged friction stir processing (SFSP): an improved method for creating ultra-fine-grained bulk materials. Mater Sci Eng A 2005;402:234–41. [20] Su JQ, Nelson TW, Sterling CJ. Friction stir processing of largearea bulk UFG aluminum alloys. Scr Mater 2005;52:135–40. [21] Liu FC, Ma ZY. Low-temperature superplasticity of friction stir processed Al–Zn–Mg–Cu alloy. Scr Mater 2008;58:667–70. [22] Woo W, Choo H. Softening behavior of friction stir welded Al 6061-T6 and Mg AZ31B alloys. Sci Technol Weld Join 2011;16: 267–72. [23] Dong P, Li HM, Sun DQ, Gong WB, Liu J. Effects of welding speed on the microstructure and hardness in friction stir welding joints of 6005A-T6 aluminum alloy. Mater Des 2013;45:524–31. [24] Fratini L, Buffa G, Shivpuri R. Mechanical and metallurgical effects of in process cooling during friction stir welding of AA7075-T6 butt joints. Acta Mater 2010;58:2056–67. [25] Zhang HJ, Liu HJ, Yu L. Microstructure and mechanical properties as a function of rotation speed in underwater friction stir welded aluminum alloy joints. Mater Des 2011;32: 4402–7. [26] Feng XL, Liu HJ, Babu SS. Effect of grain size refinement and precipitation reactions on strengthening in friction stir processed Al–Cu alloys. Scr Mater 2011;65:1057–60. [27] Liu HJ, Feng XL. Effect of post-processing heat treatment on microstructure and microhardness of water-submerged friction stir processed 2219-T6 aluminum alloy. Mater Des 2013;47:101–5. [28] Su JQ, Nelson TW, Colin JS. Microstructure evolution during FSW/FSP of high strength aluminum alloys. Mater Sci Eng A 2005;405:277–86. [29] Arora KS, Pandey S, Schaper M, Kumar R. Microstructure evolution during friction stir welding of aluminum alloy AA2219. J Mater Sci Technol 2010;26:747–53. [30] Lee CY, Choi DH, Lee WB, Park SK, Yeon YM, Jung SB. Microstructures and mechanical properties of double-friction stir welded 2219 Al alloy. Mater Trans 2008;49:885–8. [31] Chen YC, Feng JC, Liu HJ. Precipitate evolution in friction stir welding of 2219-T6 aluminum alloys. Mater Charact 2009;60: 476–81. [32] Hansen N. Hall–Petch relation and boundary strengthening. Scr Mater 2004;51:801–6. [33] Dieter GE. Mechanical Metallurgy. 3rd ed. New York: McGraw-Hill; 1986. [34] Hornbogen E, Starke Jr EA. Theory assisted design of high strength low alloy aluminum. Acta Metallogr Mater 1993;41: 1–16.