Improved mechanical properties of cast Mg alloy welds via texture weakening by differential rotation refill friction stir spot welding [original]

Scripta Materialia 203 (2021) 114113

Contents lists available at ScienceDirect

Scripta Materialia

journal homepage: www.elsevier.com/locate/scriptamat

Improved mechanical properties of cast Mg alloy welds via texture

weakening by differential rotation reﬁll friction stir spot welding

Banglong Fu

a , ∗, Junjun Shen

a , ∗, Uceu F.H.R. Suhuddin

, Ting Chen

, Jorge F. dos Santos

Benjamin Klusemann

a , b

, Michael Rethmeier

c , d

Helmholtz-Zentrum Hereon, Institute of Material Mechanics, Solid State Materials Processing, Max-Planck-Str. 1, Geesthacht 21502, Germany

Leuphana University of Lüneburg, Institute of Product and Process Innovation, Universitätsallee 1, Lüneburg 21335, Germany

Technical University Berlin, Institute for Machine Tools and Factory Management (IWF), Pascalstr. 8-9, Berlin 10587, Germany

BAM Bundesanstalt für Materialforschung und -prüfung, Unter den Eichen 87, Berlin 12205, Germany

a r t i c l e i n f o

Article history:

Received 7 May 2021

Revised 18 June 2021

Accepted 22 June 2021

Available online 7 July 2021

Editor: Greg Rohrer

Keywords:

Reﬁll friction stir spot welding

Magnesium alloy

Texture

EBSD

Plastic deformation

a b s t r a c t

Cast magnesium alloys welds produced by reﬁll friction stir spot welding (reﬁll FSSW) show low lap

shear strength (LSS) and constantly fail in stirred zone (SZ) shear mode. The cause is most probably

related to the heavily textured microstructure. Here, to re-engineer the resulting microstructure, we pro-

pose a novel process variant, the differential rotation reﬁll FSSW (DR-reﬁll FSSW). DR-reﬁll FSSW stim-

ulates discontinuous dynamic recrystallization and produces a bimodal microstructure with weakened

texture. Therefore, the deformation incompatibility between SZ and thermal-mechanically affected zone

is avoided. The welds have 50% higher LSS than that of standard reﬁll FSSW welds, and fail in a different

failure mode, i.e., SZ pull-out mode. DR-reﬁll FSSW provides a new and effective strategy for improving

the performance of spot welds based on microstructural engineering.

This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

While the compelling need for lightweight, energy-eﬃcient, en-

vironmental friendly engineering systems has motivated the use of

magnesium (Mg) alloys [1] , its application in the automotive in-

dustry is still limited due to challenges associated with manufac-

turing, processing, and in-service performance [ 2 , 3 ]. In particular,

appropriate welding processes need to be developed to widen the

structural application range of Mg alloys. Solid-state welding tech-

niques are especially eﬃcient due to avoidance of serious metal-

lurgical problems as porosity and hot cracking, which are often

present in conventional fusion based welding processes [ 4 , 5 ].

During the past few years, a relatively new solid-state spot

welding process, reﬁll friction stir spot welding (reﬁll FSSW), has

attracted much attention due to the advantages of a keyhole-free

surface, sound mechanical properties and no need for additional

ﬁller material [ 6 , 7 ]. It has been successfully used mainly for weld-

ing aluminum (Al) based similar and dissimilar combinations [8–

11] . In contrast, studies concerning Mg alloys are limited [ 12 , 13 ].

Although Mg alloys produced by casting processes represent 98%

of Mg usage [14] , no work has been reported on reﬁll FSSW of cast

Mg alloy in the literature.

∗Corresponding author.

E-mail addresses: banglong.fu@hereon.de (B. Fu), [email protected] (J.

Shen).

Preliminary unpublished studies on reﬁll FSSW of cast Mg al-

loys consistently revealed low lap shear strength (LSS) of the

welds, failing in stirred zone (SZ) “shear” mode, regardless of

the process parameters used. It is well known that the texture

is strongly connected to the mechanical properties of Mg al-

loys due to its hexagonal close-packed (hcp) crystal structure,

which offers only limited independent slip systems and short-

age of accommodating deformation along the < c > -axis [ 15 , 16 ].

Reﬁll FSSW is characterized by a layered material ﬂow behav-

ior driven by simple shear and extrusion [8] , thus the develop-

ment of strong crystallographic texture is expected. Indeed, in

friction stir welding/processing (FSW/P) of Mg alloys, the forma-

tion of sharp basal texture with inhomogeneous distribution has

been reported [17] . This favors activation of basal slip and exten-

sion twinning adjacent to the SZ edge while slip/twinning in the

SZ center during transverse tensile tests is inhibited [18] , leading

to severe strain localization and deteriorated mechanical proper-

ties. Through post-deformation/aging [19] , multi-pass FSW/P [20] ,

asymmetrical double-sided FSW [21] , the basal texture could be

weakened/randomized, and the mechanical performance was im-

proved. The beneﬁts of texture weakening of Mg alloys for the

overall mechanical response, i.e. improved formability/ductility, de-

creased anisotropy, have also been reported in rolling of Mg alloys,

by microalloying with other elements [ 22 , 23 ] and imposing severe

https://doi.org/10.1016/j.scriptamat.2021.114113

( http://creativecommons.org/licenses/by/4.0/ )

B. Fu, J. Shen, U.F.H.R. Suhuddin et al. Scripta Materialia 203 (2021) 114113

Fig. 1. Schematic illustration of reﬁll FSSW and DR-reﬁll FSSW process.

plastic deformation through asymmetric rolling [24] or high strain-

rate rolling [25] . Therefore, to achieve a sound reﬁll FSSW Mg al-

loys weld with superior mechanical properties, the texture inten-

sity needs to be reduced.

The present study aims at introducing a novel reﬁll FSSW

process variant, named “differential rotation reﬁll FSSW (DR-reﬁll

FSSW) ”, which produces spot welds in Mg alloys with improved

LSS compared to conventional reﬁll FSSW, through texture weak-

ening. Fig. 1 presents a schematic view of both processes and high-

lights the differences between reﬁll FSSW and DR-reﬁll FSSW. A

more detailed description of tool movement and rotation is pro-

vided in Supplementary Fig. S1 . A three-piece tool system, con-

sisting of a clamping ring (diameter of 15 mm), a shoulder (9 mm)

and a probe (6 mm), is used. Although the translational tool move-

ment between reﬁll FSSW and DR-reﬁll FSSW is the same, the ro-

tation state is different. To simplify the expression, the nomencla-

ture of rotation state is deﬁned as: ω

+ / −Speed

Shoulder/Probe

, which includes

the considered tool parts, i.e. shoulder or probe, rotation direction

(“+ ” denotes the clockwise direction) and speed in rpm. Unlike re-

ﬁll FSSW, in which probe and shoulder share the same rotation di-

rection and speed, in DR-reﬁll FSSW, these are different. Depending

on the rotation direction, two DR-reﬁll FSSW variants are possible:

(I) probe and shoulder rotate in the same direction, but the rota-

tion speed is different; (II) probe and shoulder rotate in opposite

directions with the same or different rotation speeds.

The base material (BM) selected for this study was Mg al-

loy AM50-F (Mg-5.2Al-0.5Mn-0.05Zn-0.03Si, in wt.%) with an av-

erage grain size of 1.8 mm, produced by permanent mold di-

rect chill casting [26] . The original cast ingot was sectioned into

target sheets measuring 100 (L) ×25.4 (W) ×3 (T) mm

. The

sheets were positioned in lap conﬁguration with an overlap dis-

tance of 25.4 mm before welding. The special-purpose welding

machine RPS 200, produced by Harms & Wende, was used to per-

form the welding. The employed welding parameters are summa-

rized in Table 1 . The cross-sections of the welds were polished and

color-etched with acetic-picral etchant. The etched samples were

examined by a LEICA DM-IRM optical microscope with polarized

light plus sensitive tint to reveal the colored grains. The global

macrotexture of the SZ was obtained by synchrotron radiation at

Deutsches Elektronen-Synchrotron (DESY). A monochromatic X-ray

beam with the energy of 87 keV and size of 1.0 ×1.0 mm

2 was

used and the extracted SZ was examined in transmission mode.

The pole ﬁgures (PFs) were determined using Fit2D [27] and in-

house code package SABO according to steps developed by Yi et al.

[28] . Detailed supporting information of measurement and calcu-

lation is provided in Supplementary Fig. S2 . The local microtexture

of selected positions in the SZ was determined by FEI Quanta 650

ﬁeld-emission scanning electron microscope equipped with EDAX

electron backscattered diffraction (EBSD) detector. EBSD data was

acquired at 15 kV with a step size of 0.2 μm and analyzed via TSL

OIM software. Room temperature LSS tests were performed via a

Zwick/Roell universal testing machine at a crosshead speed of 1

mm/min. For completeness, the dimension of the LSS test sample

and the used ﬁxture system are illustrated in Supplementary Fig.

S3 . Three samples per welding condition were tested.

Fig. 2 (a) summarizes the resulting LSS and fracture mode for

different probe rotation states ω

+ / −

Probe

. The surface morphology and

cross-section of failed specimens are presented in Fig. 2 (b) . Com-

pared to welds obtained by reﬁll FSSW ω

+ 1800

Probe

, the DR-reﬁll FSSW

welds according to variant I, ω

Probe

, reveal only a minor increase

of the LSS, and the difference in rotation speeds is negligible. All

these welds still fail in the shear mode. The cracks mainly propa-

gate along the boundary between SZ and thermo-mechanically af-

fected zone (TMAZ), then separating the upper and lower sheets

throughout the SZ. However, when the rotation directions of shoul-

der and probe are opposite, i.e. variant II as ω

−

Probe

, the LSS in-

creases signiﬁcantly, and nearly all produced welds fulﬁll the

strength requirement of referred standard [29] . For these welds,

when ω

−

Probe

is higher than ω

+ 1800

Shoulder

, i.e. ω

−2400

Probe

and ω

−30 0 0

Probe

, dur-

ing LSS test, the circular cracks mainly grow along the SZ edge and

ﬁnally close on the upper sheet surface, thus the SZ remains en-

tirely on the lower sheet, resulting in a different failure mode: SZ

“pull-out” mode.

The results of LSS illustrate that the DR-reﬁll FSSW with

opposite rotation directions of probe and shoulder has evident

advantages for the welding of cast Mg alloys. To reveal the un-

derlying mechanism of this performance improvement, welds

obtained by reﬁll FSSW with ω

+ 1800

Probe

(weld-A), DR-reﬁll FSSW

with ω

−1800

Probe

(weld-B) and ω

−30 0 0

Probe

(weld-C), were investigated

in detail. The cross-sectional micrographs are shown in Fig. 2 (c) .

All three welds display defect-free macrostructures, where no

voids or lack-of-reﬁll are observed. The coloration of grains under

polarized light is related to birefringence induced by anisotropic

ﬁlm deposition during etching, which varies according to crystal-

lographic orientation of the underlying grains [30] . In contrast to

the probe reﬁll region of the SZ in weld-A, which is characterized

by a nearly uniform region, the corresponding regions of weld-B

and C can be subdivided into two subregions. The newly formed

upper subregion is deduced to be related to the reverse rotation

of the probe. In the lower subregion, weld-B and C display the

heterogeneous color patterns, implying possible great difference

in orientation, which usually means the underlying dilute integral

texture development. Thus, the texture patterns were investigated

to verify the hypothesis that texture was weakened in weld-B and

C compared to weld-A.

The initial texture of AM50 Mg alloy BM was ﬁrst examined,

and the obtained (0 0 02) basal and (10

1 0) prismatic plane PFs are

shown in Fig. 3 (a) . The as-cast BM exhibits a near-random texture

with the maximal basal pole intensity I

max of 6.2 in multiple ran-

dom distribution. The recalculated (0 0 02) and (10

1 0) PFs of the SZ

of welds A-C are shown in Fig. 3 (b) . After welding, according to the

(0 0 02) PF of weld-A, a “basal” texture with symmetrical spread of

basal plane normal ( < c > -axis) by ±35 °maximum from ND to-

ward TD-LD plane is formed. Since the {0 0 02} plane tends to ro-

tate around the < c > -axis randomly, the corresponding (10

1 0) PF

B. Fu, J. Shen, U.F.H.R. Suhuddin et al. Scripta Materialia 203 (2021) 114113

Table 1

DR-reﬁll FSSW parameters range (likewise, “+ ”means rotation direction is clockwise, see Fig. 1 ). The

rotation state of the probe varied, while that of shoulder was ﬁxed at

+ 1800

Shoulder

, i.e. clockwise at 1800

rpm.

Rotation state (rpm) Welding time (s) Plunge depth (mm)

Shoulder Probe

∗Plunge time Dwell time Retraction time

+ 1800

Shoulder

+ / −( 60 0 −30 0 0 )

Probe

2 1 2 3

∗Note when the probe rotates in ω

+ 1800

Probe

, the weld is obtained by conventional reﬁll-FSSW.

Fig. 2. (a) LSS and fracture mode for different probe rotation states, the shoulder rotation state is ﬁxed at

+ 1800

Shoulder

, thus

Probe

and

−

Probe

correspond to the variant I and

II of DR-reﬁll FSSW, respectively. (b) Illustrations of failure modes: SZ shear and pull-out mode. (c) Cross-sectional macrostructures of weld-A, B and C. The experimental

coordinate system is deﬁned by normal direction (ND), transverse direction (TD) and longitudinal direction (LD) as shown in (c).

Fig. 3. The (0 0 02) and (10

1 0) PFs showing macrotexture of (a) BM and (b) SZ of weld-A, B and C. (c) (0 0 02) PF of P1-4 positions illustrated in Fig. 2 (c). Note that the initial

macrotexture of BM was examined by X-ray diffraction. To

obtain grain statistics, an area of 100 ×50 mm

was scanned.

B. Fu, J. Shen, U.F.H.R. Suhuddin et al. Scripta Materialia 203 (2021) 114113

Fig. 4. EBSD analysis of microstructure at the position P3. (a) weld-A, (b) weld-C. The grain orientation spread (GOS) of 2 °is the threshold value for DRXed (colored by

green in GBs map) and unDRXed grains. The regions black arrowed illustrate DDRX, and those blue arrowed show CDRX. Some grains

are highlighted by blue hcp crystal

lattice to show the orientation intuitively.

indicates the development of a weak ﬁber texture without signiﬁ-

cant preferred orientation ( I

max

< 5). The texture patterns of weld-

B and weld-C are similar to weld-A. However, the dominant basal

texture is weakened, base pole intensity I

max decreases from 19.1

for weld-A to 13.0 of weld-B and 8.8 of weld-C.

To accurately reveal the texture difference, further insight into

the local microtexture is necessary. Fig. 3 (c) shows the (0 0 02)

PFs at positions P1-P4, indicated in Fig. 2 (c) , for both weld-A and

weld-C.

In case of weld-A, the evolved sharp basal texture ( I

max > 30)

is determined at all the investigated positions, which is consistent

with the measured macrotexture. It is important to notice that,

when approaching the SZ center, i.e. from P1 to P4, the basal plane

tilt is characterized with “transitional” features. The exact orienta-

tion of the basal pole peak deviates gradually from TD towards ND

with the decrease of < 0 0 02 > inclined angle from 35 °at P1 to 5 °

at P4, indicating the strengthening of the near- < 0 0 02 >  ND texture

component. The reorientation of < 0 0 02 > with positions P1-P4 re-

veals the development of typical {0 0 02} < uvtw > B-ﬁber shear tex-

ture [31] , in which the {0 0 02} basal plane tends to align parallel

to the macroscopic shear plane due to basal slip. The predominant

material ﬂow in reﬁll FSSW can be generally described as simple

shear, which is mainly driven by the rotation of the shoulder [8] .

The plastically deformed material interacts simultaneously with

both the side and bottom surfaces of the shoulder. Since the edge

effect of the shoulder side surface decreases towards the SZ center,

the combined macroscopic shear plane inclines gradually from the

plane along shoulder side surface to that parallel to the bottom

surface, giving rise to the location-dependency of basal pole tilt

from P1 to P4 position. The P1 position of weld-C, which belongs

to the shoulder reﬁll region, shows a similar texture to weld-A.

However, within the probe reﬁll region (P2-P4), the basal texture is

diminished, and the texture is signiﬁcantly weakened/randomized

with complicated pattern. For example, at P3, I

max is reduced from

47.9 of weld-A to 14.7 in weld-C. Additionally, no apparent regu-

larity of texture variation is identiﬁed from P2 to P4 for weld-C,

indicating a turbulent material ﬂow within the probe reﬁll region.

The above results clearly show that DR-reﬁll FSSW at ω

−

Probe

i.e. probe and shoulder rotate in opposite direction, leads to sub-

stantial texture weakening compared to reﬁll FSSW. Considering

the LSS improvement and the different observed failure mode, see

Fig. 2 (a) and (b) , the beneﬁts of “weakened texture” on the me-

chanical properties of cast Mg weld are revealed. The characteristic

microstructural differences and their inﬂuence on the deformation

behavior between standard reﬁll FSSW and DR-reﬁll FSSW are fur-

ther discussed in the following to better understand the underlying

mechanisms.

The microstructure at the representative position P3, which is

within the probe reﬁll region, see Fig. 2 (c) , was analyzed via EBSD

for weld-A and weld-C, see Figs. 4 (a) and (b) . The extracted grain

and texture information are summarized in Table 2 .

For the weld-A, a continuous high-angle grain boundaries

(HAGBs) perimeter cannot be observed. The grain morphology fea-

tures an irregular mixture of low-angle grain boundaries (LAGBs)

and HAGBs with prevalence of LAGBs and high HAGBs distance.

Such microstructures have also been reported in FSW/P of Mg al-

loys [ 4 , 17 , 32 , 33 ], which are related to “grain convergence” [17] un-

der the effect of strong texture development during severe plastic

deformation. Additionally, the identiﬁed grain structure indicates

the occurrence of dynamic recrystallization (DRX). The black ar-

rowed positions of IPF and GBs maps show the DRXed grains orig-

inate from bulging of corrugated HAGBs. According to the embed-

ded Kernel Average Misorientation (KAM) map, the local regions

with low misorientation exist on the concave sides, thus strain-

induced boundary migration [ 34 , 35 ] occurs, implying the activation

of discontinuous dynamic recrystallization (DDRX). However, ﬁne

DRXed grains, blue arrowed, also form due to the gradual LAGB-

to-HAGB transformation through dislocation rearrangement, which

B. Fu, J. Shen, U.F.H.R. Suhuddin et al. Scripta Materialia 203 (2021) 114113

Fig. 5. EBSD analysis of microstructure at the position P3 and the bottom boundary between SZ and TMAZ after applying 4 kN force. (a) weld-A, (b) weld-C. The image

quality (IQ) map overlapped with IPF shows the existence of shear zone, which can be indexed by EBSD, while the shear band cannot be indexed due to severe lattice

distortion.

Table 2

Extracted grain and texture information according to EBSD for weld-A and weld-C at position P3 after welding (ﬁnal weld) as well

as after LSS test at 4 kN.

Weld

Grain information Basal texture intensity

Size (μm) HAGBs distance (μm) LAGBs fraction (%) DRX fraction (%) Whole unDRXed DRXed

weld-A 7.6 16.2 36.9 58.6 47.9 53.2 44.1

weld-C 6.9 8.8 36.8 51.6 14.7 24.3 5.9

weld-A 4.6 18.2 43.8 58.3 69.6 75 65.7

weld-C 5.4 5.8 23.8 60 15.4 23.4 10.5

are related to continuous dynamic recrystallization (CDRX). The

DRXed grains show a strong basal texture close to that of the un-

DRXed grains with rotation mainly along < 0 0 02 > . Moreover, com-

pared to the misorientation distribution of unDRXed grains, the

fraction of LAGBs decreases while that of the range 15-30 °pref-

erentially increases in DRXed grains. Both results discussed above

indicate that CDRX is the main contributor to recrystallization. It

is important to point out that, for the SZ of weld-A, the misorien-

tation distribution is somewhat unusual, the maximum misorien-

tation is usually restricted below 30 °, and the misorientation axes

cluster near < 0 0 01 > in the interval of 5-30 °. This unique feature

can be interpreted by the development of a strong B-ﬁber texture,

which means that the grains are arbitrarily rotated along < 0 0 02 > .

The rotation angle above 30 °is considered as equivalent due to the

symmetry of the hcp crystal structure [ 4 , 17 ].

In weld-C, a bimodal grain morphology is identiﬁed. The neck-

laces of ﬁne DRXed grains generally decorate the GBs of coarse un-

DRXed grains. It is clear that most DRXed grains have nucleated

through bulging of GBs, and are completely surrounded by HAGBs

without sub-structures. The crystallographic orientations of the

DRXed grains are not close to the matrix as observed in weld-A but

random, see the inserted hcp unit cells, thus the texture of DRXed

grain is signiﬁcant weakened ( I

max = 5.9). These observations well

document the primary recrystallization mechanism in weld-C to be

DDRX [36–38] , although some traces of CDRX are observed as well.

Due to the weakened texture of weld-C, convergence of GBs dis-

appears, and the HAGBs distance is smaller than that in weld-A.

However, the fraction of LAGBs does not change, which is related

to the development of geometrically necessary boundaries [ 39 , 40 ]

within the unDRXed grain interiors. These boundaries are LAGBs in

natural and straight in morphology with the traces close to {10

1 0}

or {11

2 0} planes. The misorientation variation from unDRXed to

DRXed grains is characterized with a homogeneous increase in the

entire HAGBs range, which again indicates the dominant role of

DDRX in weld-C.

The differences in microstructure between weld-A and C inﬂu-

ence the deformation behavior during LSS, and ﬁnally result in var-

ied failure modes. To reveal the underlying relationship, both welds

were ﬁrstly loaded at 4 kN, i.e. near to the maximum LSS of weld-

A, and then unloaded to analyze the microstructure. The EBSD re-

sults are shown in Fig. 5 . The corresponding grain and texture in-

formation is also included in Table 2 .

For the weld-A, the microstructure with a mixture of LAGBs

and HAGBs still exists at position P3 of the SZ, see Fig. 5 (a) . Com-

pared to the original weld, after loading, both the LAGBs fraction

and HAGBs distance increase, see Table 2 , which are related to the

further developments of dislocation interaction and grain conver-

gence, indicating strain hardening within the SZ [33] . Thus, de-

B. Fu, J. Shen, U.F.H.R. Suhuddin et al. Scripta Materialia 203 (2021) 114113

formation incompatibility occurs along the boundary between SZ

and TMAZ, leading to the formation of shear bands with severe

lattice distortion and local strain, which are preferred sites of fail-

ure. While in weld-C, the deformation of the SZ is concentrated

mainly along the GBs of ﬁne and coarse grains, leading to further

development of DRX and the formation of micro- shear zones, see

Fig. 5 (d) . These shear zones are short and discontinuous, which can

accommodate strain by dislocation slip and repeated DRX [ 41 , 42 ].

The shear zones are distributed homogeneously within the whole

SZ, providing a stabilizing effect between the deformation of SZ

and TMAZ. Thus, the formation of shear bands is restrained, re-

sulting in the SZ pull-out failure mode.

In summary, a novel welding variant, DR-reﬁll FSSW, is re-

ported, which can introduce a bimodal microstructure with weak-

ened texture compared to conventional reﬁll FSSW, leading to a

signiﬁcant increase in LSS of the weld in cast Mg alloy. Addition-

ally, DR-reﬁll FSSW has also shown competitive advantages in the

welding of other materials combinations such as Al/Al, Mg/steel,

which will be reported in the future.

Declaration of Competing Interest

The authors declare that they have no known competing ﬁnan-

cial interests or personal relationships that could have appeared to

inﬂuence the work reported in this paper.

Acknowledgments

The authors are grateful to Prof. Dr. Norbert Hort at Helmholtz-

Zentrum Hereon (HEREON) for providing the Mg alloy ingots. We

acknowledge Deutsches Elektronen-Synchrotron DESY (Hamburg,

Germany), a member of the Helmholtz Association HGF, for the

provision of experimental facilities. Parts of the research were car-

ried out at the High Energy Materials Science (HEMS) of HEREON

and we would like to thank Dr. Xiaohua Zhou for assistance in

using P07B beamline. We would also like to acknowledge to Dr.

Sangbong Yi for training on SABO software. Banglong Fu grate-

fully acknowledges funding by China Scholarship Council (grant no.

201506220158 ).

Supplementary materials

Supplementary material associated with this article can be

found, in the online version, at doi: 10.1016/j.scriptamat.2021.

114113 .

References

[1] T.M. Pollock , Science 328 (2010) 986–987 .

[2] W.J. Joost , P.E. Krajewski , Scr. Mater. 128 (2017) 107–112 .

[3] B.C. Suh , M.S. Shim , K.S. Shin , N.J. Kim , Scr. Mater. 84–85 (2014) 1–6 .

[4] U. Suhuddin , S. Mironov , Y. Sato , H. Kokawa , C.-W. Lee , Acta Mater. 57 (2009)

5406–5418 .

[5] V.K. Patel , S.D. Bhole , D.L. Chen , Scr. Mater. 65 (2011) 911–914 .

[6] U.F.H. Suhuddin , V. Fischer , J.F. dos Santos , Scr. Mater. 68 (2013) 87–90 .

[7] J. Shen , U.F.H. Suhuddin , M.E.B. Cardillo , J.F. dos Santos , Appl. Phys. Lett. 104

(2014) 191901 .

[8] J. Shen , S.B.M. Lage , U.F.H. Suhuddin , C. Bolfarini , J.F. dos Santos , Metall. Mater.

Trans. Phys. Metall. Mater. Sci. 49 (2018) 241–254 .

[9] Z. Shen , Y. Chen , J.S.C. Hou , X. Yang , A.P. Gerlich , Sci. Technol. Weld. Join.

(2015) 48–57 .

[10] A.H. Plaine , U.F.H. Suhuddin , C.R.M. Afonso , N.G. Alcantara , J.F. dos Santos ,

Mater. Des. 93 (2016) 224–231 .

[11] Z. Shen , Y. Ding , J. Chen , B.S. Amirkhiz , J.Z. Wen , L. Fu , A.P. Gerlich , J. Mater.

Sci. Technol. 35 (2019) 1027–1038 .

[12] Z. Shen , Y. Ding , J. Chen , A.P. Gerlich , Int. J. Fatigue 92 (2016) 78–86 .

[13] L.C. Campanelli , U.F.H. Suhuddin , A.Í.S. Antonialli , J.F. dos Santos , N.G. de Al-

cantara , C. Bolfarini , J. Mater.

Process. Technol. 213 (2013) 515–521 .

[14] A .A . Luo , J. Magnes. Alloys 1 (2013) 2–22 .

[15] Z. Wu , R. Ahmad , B. Yin , S. Sandlöbes , W.A. Curtin , Science 359 (2018)

447–452 .

[16] J.H. Hwang , A. Zargaran , G. Park , O. Lee , B.-J. Lee , N.J. Kim , J. Magnes. Alloys 9

(2021) 4 89–4 98 .

[17] S. Mironov , T. Onuma , Y.S. Sato , H. Kokawa , Acta Mater. 100 (2015) 301–312 .

[18] R. Xin , B. Li , A. Liao , Z. Zhou , Q. Liu , Metall. Mater. Trans. A 43 (2012)

2500–2508 .

[19] A.H. Feng , Z.Y. Ma , Acta Mater. 57 (2009) 4248–4260 .

[20] A . Tripathi , A . Tewari , A .K. Kanjarla , N. Srinivasan , G.M. Reddy , S.M. Zhu ,

J.F. Nie , R.D. Doherty , I. Samajdar , Metall. Mater. Trans. Phys. Metall. Mater. Sci.

47 (2016) 2201–2216 .

[21] J. Chen , H. Fujii , Y. Sun , Y. Morisada , K. Kondoh , J. Mater , Process. Technol. 242

(2017) 117–125 .

[22] X. Zeng , P. Minárik , P. Dobro

n , D. Letzig , K.U. Kainer , S. Yi ,Scr. Mater. 166

(2019) 53–57 .

[23] B.C. Suh , J.H. Kim , J.H. Bae , J.H. Hwang , M.S. Shim , N.J. Kim , Acta Mater. 124

(2017) 268–279 .

[24] B. Beausir , S. Biswas , D.I. Kim , L.S. Tóth , S. Suwas , Acta Mater. 57 (2009)

5061–5077 .

[25] S.Q. Zhu , H.G. Yan , X.Z. Liao , S.J. Moody , G. Sha , Y.Z. Wu , S.P. Ringer , Acta Mater.

82 (2015) 344–355 .

[26] F.R. Elsayed , N. Hort , M.A. Salgado Ordorica , K.U. Kainer , Mater. Sci. Forum 690

(2011) 65–68 .

[27] A.P. Hammersley , J. Appl. Crystallogr. 49 (2016) 646–652 .

[28] S.B. Yi , C.H.J. Davies , H.G. Brokmeier , R.E. Bolmaro , K.U. Kainer , J. Homeyer , Acta

Mater. 54 (2006) 549–562 .

[29] American Welding Society, AWS

D17.2/D17.2M:2013 speciﬁcation for resistance

welding for aerospace applications, 2013.

[30] N. Stanford , R.K.W. Marceau , M.R. Barnett , Acta Mater 82 (2015) 447–456 .

[31] B. Beausir , L.S. Tóth , K.W. Neale , Acta Mater. 55 (2007) 2695–2705 .

[32] S. Mironov , Q. Yang , H. Takahashi , I. Takahashi , K. Okamoto , Y.S. Sato ,

H. Kokawa , Metall. Mater. Trans. A 41 (2010) 1016–1024 .

[33] F.C. Liu , Z.Y. Ma , M.J. Tan , Metall. Mater. Trans. Phys. Metall. Mater. Sci. 44

(2013) 3947–3960 .

[34] H. Liu , K. Ushioda , H. Fujii

, Acta Mater 166 (2019) 324–334 .

[35] Z. Mengran , S. Yufeng , M. Yoshiaki , U. Kohsaku , F. Hidetoshi , J. Magnes. Alloys

8 (2020) 1071–1083 .

[36] C.D. Barrett , A. Imandoust , A.L. Oppedal , K. Inal , M.A. Tschopp , H. El Kadiri ,

Acta Mater. 128 (2017) 270–283 .

[37] M.G. Jiang , C. Xu , H. Yan , G.H. Fan , T. Nakata , C.S. Lao , R.S. Chen , S. Kamado ,

E.H. Han , B.H. Lu , Acta Mater. 157 (2018) 53–71 .

[38] A. Heidarzadeh , S. Mironov , R.

Kaibyshev , G. Çam , A. Simar , A. Gerlich , F. Khod-

abakhshi , A. Mostafaei , D.P. Field , J.D. Robson , A. Deschamps , P.J. Withers , Prog.

Mater. Sci. 117 (2021) 100752 .

[39] É. Martin , J.J. Jonas , Acta Mater. 58 (2010) 4253–4266 .

[40] S. Biswas , B. Beausir , L.S. Toth , S. Suwas , Acta Mater. 61 (2013) 5263–5277 .

[41] E. Dogan , M.W. Vaughan , S.J. Wang , I. Karaman , G. Proust , Acta Mater. 89

(2015) 408–422 .

[42] A. Beer , M. Barnett , Metall. Mater. Trans.

A 38 (2007) 1856–1867 .