Journal Pre-proofs

Structural morphology of Al-Mg-Si alloy friction stir welds through tool ec‐
centricity

L.H. Shah, N. Huda, S. Esmaeili, S. Walbridge, A.P. Gerlich

PII: S0167-577X(20)30803-X
DOI: https://doi.org/10.1016/j.matlet.2020.128098
Reference: MLBLUE 128098

To appear in: Materials Letters

Received Date: 10 April 2020

Revised Date: 1 June 2020
Accepted Date: 2 June 2020

Please cite this article as: L.H. Shah, N. Huda, S. Esmaeili, S. Walbridge, A.P. Gerlich, Structural morphology of
Al-Mg-Si alloy friction stir welds through tool eccentricity, Materials Letters (2020), doi: https://doi.org/10.1016/
j.matlet.2020.128098

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2020 Published by Elsevier B.V.

 Structural morphology of Al-Mg-Si alloy friction stir welds through tool
eccentricity
L. H. Shaha,b*, N. Hudaa, S. Esmaeilia, S. Walbridgec and A. P. Gerlicha
aDepartment of Mechanical and Mechatronics Engineering, University of Waterloo, 200

University Avenue West, Waterloo, Ontario N2L 3G1, Canada.

bFaculty of Mechanical and Automotive Engineering Technology, Universiti Malaysia

Pahang, 26600 Pekan, Pahang, Malaysia.

cDepartment of Civil and Environmental Engineering, University of Waterloo, 200

University Avenue West, Waterloo, Ontario N2L 3G1, Canada.

*Corresponding author: lhahmads@uwaterloo.ca

Abstract
In this work, the microstructure development in the stir zone of Al-Mg-Si alloy is
evaluated while employing tool eccentricity during friction stir welding. Low dislocation
density with dispersoids were observed in the inner band region of the stir zone produced
with aligned tooling. On the other hand, a high dislocation density with Mg2Si precipitates
can be observed in the same region of the stir zone when a tool eccentricity of 0.2 mm was
utilized. The discrepancy is attributed to the enhanced shearing activity imposed on the
material during the welding process.
Keywords: Friction stir welding; aluminum; tool eccentricity; dislocation; precipitates;
aluminum alloy.
1. Introduction
Friction stir welding (FSW) is a promising solid-state welding process due to its
various advantages such as ease of operation and avoidance of filler metal [1]. While the
technology has been applied in numerous industrial sectors which involve the application of
aluminum alloys, fundamental research to link the joint’s properties and its microstructural
morphology is still ongoing [2]–[5]. To date, several transmission electron microscopy
(TEM) analysis of AA6××× (Al-Mg-Si) alloy friction stir welds have been conducted [6]–
[11].
In our prior work on thick-plate AA6061 FSW [12], periodic bands were observed
in the stir zone (SZ) throughout the thickness of the samples. Hardness measurements
reveal softening occurring in these regions. With the utilization of tool eccentricity, i.e. the
effect of a slight tool offset from the spindle center that creates a ‘wobbling’ motion during
rapid spindle rotation, the periodic bands of softened material expanded horizontally,

1
enlarging the softened area of the SZ. To elucidate this phenomenon further, the effect of
tool eccentricity on the local SZ microstructure was investigated in this study.
2. Materials and Methods
In this study, a 9.5 mm thick plate AA6061-T651 (0.81 wt.% Mg, 0.53 wt.% Si,
0.19 wt.% Fe, 0.08 wt.% Mn, 0.06 wt.% Cr) was utilized. The FSW was done in a butt joint
configuration parallel to the rolling direction. The FSW tool used was an H13 steel tool
with a tapered, threaded and tri-flats pin profile. The diameter of the tool shoulder and pin
root is 15 mm and 6 mm, respectively, with a 9.3 mm pin length. Due to the thick plate
material, the rotational speed and weld speed in the FSW process was fixed to 1120 rpm
and 63 mm/min, respectively, based on preliminary experiments in prior work [12]. Apart
from the commonly used ‘aligned’ tool setup, tool eccentricity of 0.2 mm was employed to
observe structural morphology in the SZ. The tool modification and setup have been
outlined in [12]. The welded samples were left for at least 10 days to naturally age and
stabilize before metallurgical characterizations.
TEM samples from the aligned setup (aligned sample) and 0.2 mm eccentricity
setup (eccentric sample) were initially cut mid-thickness parallel to the top plane of the SZ
using a precision cutter machine into 15 mm × 10 mm × 1 mm coupons. Next, the
specimens’ thickness were carefully ground down to ~0.15 mm to ensure the inner band
region and outer band region (shown as thick red lines in Figure 1) of the periodic bands
can be extracted. The samples were then punched into 3 mm-diameter disks and underwent
jet electropolishing in nitric acid/methanol solution at -50 °C. TEM microscopy and energy
dispersive X-ray spectroscopy (EDS) analysis were done using a Philips CM12 with a LaB6
filament operating at 120 keV. The observed selected area electron diffraction (SAED)
patterns were analyzed using Crystbox software [13]–[15] and subsequently confirmed by
manual measurements.

Figure 1. Cross-sectional macrograph of the SZ for (a) aligned sample and (b) eccentric
sample.

2
3. Results and discussion
Figure 2 presents the TEM micrographs of the inner band region and the outer band
region of both samples. In both regions of the aligned sample (Figure 2(a) and (b)), coarse
rod-shaped dispersoids (shown in red arrows) averaging 1.31 ± 0.62 µm were present on
the grain boundaries as well as within the grains. The dislocation density for both regions
are also relatively low owing to the thermal cycle undergone during FSW [6]. In addition,
much finer rod-shaped particles (blue arrows) are seen dispersed throughout the aligned
sample’s inner band region (Figure 2(a)) while the plate-shaped particles (yellow arrows)
are more predominant in the outer region (Figure 2(b)).
However, higher dislocation density in the inner band region of the eccentric sample
was observed (Figure 2(c)). It is postulated that the region undergoes enhanced deformation
and shearing activity, amplifying the formation of dislocations. On the other hand, the outer
band region show abundance of the plate-shaped particles, likely to be Si-rich stable 𝛽
precipitates [16], which is observed throughout the area similar to the outer band region of
the aligned sample, but with the absence of the coarse dispersoids. Even so, it is probable
that the particle highlighted in the red circle in Figure 2(d) is a remnant of a dispersoid,
presumably being sheared due to the rigorous eccentric motion. It may also be because of
this reason that these dispersoids are largely absent in the eccentric sample.

Figure 2. TEM micrographs of (a) inner band region and (b) outer band region of aligned
sample, as well as (c) inner band region and (d) outer band region of eccentric sample.

3
 Figure 3 presents TEM micrographs comparison of the inner band region for
aligned and eccentric sample, taken from the [0 0 1]Al zone axis. The insets of Figure 3(a)
and (d) are the SAED patterns of the aluminum matrix extracted from the red circles,
respectively. The difference in dislocation density can be clearly seen. While only coarse
dispersoids (red arrows) and rod-shaped particles (blue arrows) are present in the aligned
sample, higher magnification of the eccentric samples reveals that, in addition to the rod-
shaped particles, finer needle-shaped 𝛽” precipitates (green arrows) are also present.

Figure 3. TEM micrographs of inner band region for aligned sample ((a) to (c)) and
eccentric sample ((d) to (f)).

Table 1 summarizes the observed dominant precipitates, dispersoids and dislocation

density of the inner and outer band region for both samples. The possible particle
compounds were suggested based on the EDS peaks observed (not shown). The precipitates
and dispersoids dimensions listed are an average measured from several observed locations.
The precipitation sequence in AA6061-T651 alloy is given as [17], [18],
Super-saturated solid solution → Guinier-Preston zone → 𝛽” (needle) → 𝛽’ (rod) → 𝛽
(plate)
where maximum strength can be obtained at the 𝛽” stage. However, due to the high heat
input of the process and the natural aging after FSW, the SZ region show a tendency to
form the more stable plate-shaped 𝛽 precipitates (Figure 2(b) and (d)) which softens the SZ
as compared to the base metal [19]. In addition, the coarse dispersoids remains in the
structure due to high thermal stability at elevated temperatures [6].

4
 Table 1. Dominant precipitate and dislocation density of inner and outer band region of
samples.

Dominant precipitate Dispersoids Dislocation

(average diameter/length in (average length in nm) density
nm)
Aligned sample
Large, rod-shaped
Al12(Fe,Cr,Mn)3Si
Inner band - 1308 ± 619 Low
region Small, rod-shaped
Al12(Fe,Cr,Mn)3Si
292 ± 53
Plate-shaped Large, rod-shaped
Outer band 𝛽 (Si) Al12(Fe,Cr,Mn)3Si Low
region 137 ± 44 1643 ± 755
Eccentric sample
Needle-shaped Small, rod-shaped
Inner band 𝛽” and 𝛽’ (Mg2Si) Al12(Fe,Cr,Mn)3Si High
region 233 ± 28 289 ± 75
Plate-shaped
Outer band 𝛽 (Si) - Medium
region 115 ± 45

Previous work by the authors have shown that the periodic bands have the lowest
hardness values in the SZ, while expansion of this ‘soft zone’ was observed when 0.2 mm
tool eccentricity setup was utilized [20]. This is despite the high dislocation densities
observed in the eccentric setup inner band region images (Figure 2(c) and Figure 3(d) to
(f)), which seem to show minimal effect in increasing the local hardness of the SZ, contrary
to conventional wisdom.
Owing to the enhanced shearing in the eccentric setup, it is postulated that the fine
grains and high dislocation density which is present mainly in the inner band region of this
sample may have acted as a catalyst for precipitate growth. Works by Andersen show that
the dimension of the semi-coherent strengthening 𝛽” precipitates is typically about 50 nm
in length [21]. These precipitates act as shearable obstacles against dislocation movement
[22]. However, if coarsening occurs, the precipitates become non-shearable and the
dislocations by-pass the particles instead, decreasing the retarding effect against dislocation
movements [23], [24]. Based on the average value in Table 1 (233 ± 28 nm), the 𝛽”

5
precipitates undergone coarsening, making it less useful in strengthening the material.
Thus, formation of these overaged, large needle-shaped precipitates from the dislocation-
precipitate interactions may have effectively negate the strengthening effect of the high
dislocation densities, resulting in a comparable ‘soft zone’ with the aligned sample.
4. Conclusion
A high dislocation density and needle-shaped precipitates were observed in the
inner band region of AA6061 friction stir welds when tool eccentricity was employed. This
is in contrast with the commonly observed low dislocation density and dissolved
strengthening precipitates in the stir zone. The discrepancy is attributed to the enhanced
shearing activity imposed on the material during the welding process.
Acknowledgement
The authors would like to acknowledge Carmen Andrei from Canadian Centre for
Electron Microscopy (CCEM), McMaster University, Hamilton for the TEM observations.
This study was financially supported by the National Science and Engineering Research
Council (NSERC), Ministry of Education Malaysia, and Universiti Malaysia Pahang.

References
[1] D. Lohwasser, Z. Chen, Friction stir welding: From basics to applications,
Woodhead Publishing Limited, New York, 2010.
[2] A. P. Reynolds, Scr. Mater., 58 (2008) 338–342.
[3] X. Liu, C. Wu, G. K. Padhy, Scr. Mater., 102 (2015) 95–98.
[4] Y. Morisada, T. Imaizumi, H. Fujii, Scr. Mater., 106 (2015) 57–60.
[5] R. W. Fonda, J. F. Bingert, Scr. Mater., 57 (2007) 1052–1055.
[6] A. H. Feng, D. L. Chen, Z. Y. Ma, Metall. Mater. Trans. A, 41 (2010) 2626–2641.
[7] Y. S. Sato, H. Kokawa, M. Enomoto, S. Jogan, Metall. Mater. Trans. A, 30 (1999)
3125–3130.
[8] Y. S. Sato, H. Kokawa, M. Enomoto, S. Jogan, Metall. Mater. Trans. A, 30 (1999)
2429–2437.
[9] L.-E. Svensson, L. Karlsson, H. Larsson, B. Karlsson, M. Fazzini, J. Karlsson, Sci.
Technol. Weld. Join., 5 (2000) 285–296.
[10] W. B. Lee, Y. M. Yeon, S. B. Jung, Mater. Sci. Technol., 19 (2003) 1513–1518.
[11] C. A. W. Olea, L. Roldo, J. F. Santos, T. R. Strohaecker, Mater. Sci. Eng. A, 454–

6
 455 (2007) 52–62.
[12] L. H. Shah, S. Guo, S. Walbridge, A. Gerlich, Manuf. Lett., 15 (2018) 14–17.
[13] M. Klinger, CrysTBox - Crystallographic Toolbox, 48 (2015).
[14] M. Klinger, A. Jäger, J. Appl. Crystallogr., 48 (2015) 2012–2018.
[15] M. Klinger, J. Appl. Crystallogr., 50 (2017) 1226–1234.
[16] C. D. Marioara, S. J. Andersen, H. W. Zandbergen, R. Holmestad, Metall. Mater.
Trans. A, 36 (2005) 691–702.
[17] M. H. Jacobs, Philos. Mag., 26 (1972) 1–13.
[18] H. W. L. Phillips, J. Inst. Met., 72 (1946) 151.
[19] L. H. Ahmad Shah, S. Sonbolestan, A. R. H. Midawi, S. Walbridge, A. Gerlich, Int.
J. Adv. Manuf. Technol., 105 (2019) 889–904.
[20] L. H. Ahmad Shah, A. R. H. Midawi, S. Walbridge, A. Gerlich, J. Alloys Compd.,
826 (2020).
[21] S. J. Andersen, Metall. Mater. Trans. A, 26 (1995), 1931–1937.
[22] X. Wang, S. Esmaeili, D. J. Lloyd, Metall. Mater. Trans. A, 37 (2006) 2691–2699.
[23] P. Sepehrband, X. Wang, H. Jin, S. Esmaeili, Mater. Charact., 137 (2018) 212–221.
[24] S. Esmaeili, D. J. Lloyd, W. J. Poole, Acta Mater., 51 (2003) 3467–3481.
Highlights
 Friction stir welding tool was modified to provide enhanced eccentric motion.
 Change in dislocation density and secondary particles was observed in stir zone.
 It is attributed to the enhanced shearing activity imposed during welding process.