Estimation of Material Flow in Stir Zone during Friction ... of Material Flow in Stir Zone during Friction Stir Welding by Distribution Measurement of Si Particles Hidetoshi Fujii

Estimation of Material Flow in Stir Zone during Friction Stir Welding

by Distribution Measurement of Si Particles

Hidetoshi Fujii1, Young Gon Kim2, Takuya Tsumura1, Toru Komazaki3 and Kazuhiro Nakata1

1Joining and Welding Research Institute, Osaka University, Ibaraki 567-0047, Japan2Graduate School of Engineering, Osaka University, Suita 565-0871, Japan3Ryobi Limited, Tokyo 114-8518, Japan

The size, aspect ratio and direction angle of the Si particles in the stir zone were investigated for the ADC12 FSW joints to analyze thematerial flow in the stir zone by quantification of three elements (size, aspect ratio and direction) under various welding conditions. The numberof finer Si particles increases with the increasing welding speed, whereas, the change in the number of finer Si particles is not simple when therotation speed is varied. The stirring action during the FSW is appraised by the size of the Si particles. The aspect ratio of the particles increaseswith the increasing size of the particles. The welding conditions also influence the direction angle of the Si particles. Under the properconditions, the flow directions of the Si particles in the top and bottom are horizontally aligned, while they are longitudinally aligned on theretreating side and advancing side. However, under insufficient heat input conditions, the material flow becomes random in the bottom andretreating side regions. Under abnormal stirring conditions, on the other hand, it is random on the advancing side and top regions, even though adefect forms on the advancing side in both cases. This kind of identification of the material flow might be very useful to determine the optimalwelding conditions by avoiding any defect formation.

(Received September 30, 2005; Accepted November 25, 2005; Published January 15, 2006)

Keywords: material flow, silicon particle, stir zone, welding speed, rotation speed, defect formation

1. Introduction

Friction stir welding (FSW) was developed at TheWelding Institute (TWI) in 1991.1) The material flowbehavior in the stir zone is very important in this solid-statewelding process. As for the material flow, several inves-tigations have been carried out.2–13) In these studies, thematerial flow patterns were investigated by various methodssuch as dissimilar welding,2,3,10) the insert marker technique(IMT),4,5,9,11) the orientation imaging microscopic (OIM)analysis,6,8,13) and the finite element simulation.7,12)

Li et al.3) reported complex flow patterns in dissimilarwelds of 2024 Al to 6061 Al, which were visualized byoptical microscopy (OM) due to the different etchings of thetwo aluminum alloys. The complex patterns illustrate solid-state flow in intercalated vortices both parallel and perpen-dicular to the tool axis. By Reynolds,5) flow visualization wasperformed for the 2195-T8 Al alloy using an embeddedmaker material of a 5454-H32 sheet. Colligan4) demonstratedthe material flow of the 6061-T6 and 7075-T6 plate. Smallsteel balls were used as a tracer material embedded atdifferent positions within the butt joint welds of the twoaluminum alloys. It is allowed that the tracer line techniqueproduced a comprehensive depiction of material movementin friction stir welds. Sato et al.6) reported that the micro-textures in a friction stir weld of the 6063 Al alloy have beenanalyzed by OIM. They examined the texture in therecrystallized zone, and found that the texture evolution isassociated with the deviator stress states. From these variousmethods, the following characteristics were mainly foundabout the material flow during the FSW.(1) The material flow patterns are visualized complex

vortex, swirl features in the cross section of the FSWjoints.2,3,9,10,13) From the results of Seidel et al.,9)

material is pushed downward on the advancing sideand moved upward toward the top on the retreating side

within the probe diameter on the cross section perpen-dicular to the welding direction. Thus, the material flowis seen as asymmetric during the FSW.

(2) Colligan4) showed that the steel ball, which was insertedon the advancing side, moved along the rotationdirection of the probe, and finally diffused on theretreating side. On the contrary, the steel ball, whichwas inserted on the retreating side, moved backwardwithout the effect of the rotation of the probe. Apartfrom the rotating probe, the materials are transportedstraight along the welding direction in the regions,although the materials are transported along the probesurface near the probe.6) The materials pass behind therotating probe from the retreating side to the advancingside.5,7,11) Field et al.8) investigated the texture compo-nents of the weld at the top plane, the midplane, and thebottom plane. The dominant shear direction is alignedwith the tangent to the rotating tool at any position.

(3) The defect formation mechanism is dominated by thematerial flow in the stir zone. Liming et al.12) presentedthe cavity model which relates the material flow duringthe FSW. If the material does not flow fast enough to fillthe cavity, a hole may be formed at the weld center ornear the advancing side, when the travel speed is toofast or the probe rotation is too slow.

However, a complete understanding about the materialflow phenomena has not been achieved, because differentexplanations are still being made by different researchers.Accordingly, the aim of this study is to investigate in detailthe material flow in the stir zone by analyzing the distributionmeasurement of the Si particles in ADC12 alloy using theimage processing method. In order to clarify the defectformation mechanism in the stir zone, the effect of thewelding parameters, such as the welding speed and therotation speed, on the material flow phenomena during theFSW is also evaluated.

Materials Transactions, Vol. 47, No. 1 (2006) pp. 224 to 232#2006 The Japan Institute of Metals EXPRESS REGULAR ARTICLE

2. Experimental Procedures

Four mm thick plates of an ADC12 aluminum die castingalloy were used in this study. The gas content of the plate wasabout 2.6mL/100 g Al. Table 1 is the chemical compositionof the base material.

I-groove butt welding (50mm(w) � 300mm(l) � 2piece) was performed using a load control type of FSWmachine. The FSW tool is a columnar shape with a screwprobe. SKD61 was used as the tool material. Table 2 showsthe tool sizes and welding conditions. The tilt angle was 3degrees, and the tool plunge downforce was 14.2 kN. The toolrotation speed in a clockwise direction and the welding speedwere changed from 500 to 1750min�1 and from 250 to1000mm/min, respectively. The plate top surface andgroove surface were degreased with acetone just prior towelding. Mild steel was used as the backing plate.

After FSW, a metallurgical inspection was performed onthe cross-sections of the FSW joints. They were polished andetched with a 5% hydrofluoboric acid (HBF4) water solutionfor the scanning electron microscope (SEM) observations. In

order to examine the material flow in the stir zone, the size,the aspect ratio and the direction angle of the Si particleswere measured by the image processing method (Image-ProPlus ver. 4.0). In order to reduce the statistical error, fivephotographs were taken and quantified in each region. Theaspect ratio is a value of the major axis divided by the minoraxis of the closest ellipse shape. The direction angle isdefined as the angle between the major axis of the particleand the Y-axis (a vertical line). A 0 degree direction angledenotes an upward direction. When the direction angle is 90degrees, the major axis of the particle is perpendicular to theY-axis.

3. Results and Discussion

3.1 Distribution of Si particles in the stir zoneFigure 1 is a schematic illustration of the observation

position of the Si particle distribution in the stir zone.Figure 2 shows SEM images of the Si particle distributionsin the parts of (a) the top, (b) the retreating side, (c) themiddle, (d) the advancing side and (e) the bottom in the stirzone of a joint, obtained at 14.2 kN, 1250min�1 and 250mm/min. It is found that the shape of the Si particles in thetop is more elongated than that in the other regions. It isthought that the shape of the Si particles should be affectedby the temperature and the plastic flow. When the temper-

Table 1 Chemical compositions of base material.

AlloyChemical compositions (mass%)

Cu Si Mg Zn Fe Mn Ni Sn Pb Al

ADC12 2.35 11.82 0.17 0.56 0.81 0.18 0.04 0.02 0.06 Bal.

Table 2 FSW tool and welding parameters.

Tool size Welding parameters

Shoulder

diameter

(mm)

Probe

diameter

(mm)

Probe

length

(mm)

Tool tilt

(�)

Downforce

(kN)

Rotation

speed

(min�1)

Welding

speed

(mm/min)

15 5 3.9 3 14.2 500–1750 250–1000

Retreating side

3

5 Advancing side

2.4 2.4

1.5

1.5

(mm)

Fig. 1 Schematic illustration of observation positions of microstructure in

the stir zone.

5µm

(a) Top

(b) Retreating side (c) Middle (d) Advancing side

(e) Bottom

Fig. 2 SEM images of Si particles in each part of the stir zone (14.2 kN, 1250min�1–250mm/min).

Estimation of Material Flow in Stir Zone during Friction Stir Welding by Distribution Measurement of Si Particles 225

ature is higher, the Si particles should be more difficult tobreak down due to the lower resistance of the plastic flow. Inorder to clarify the effect of the temperature increase only, aheat treatment was conducted on the base metal at 723K for5min and 3 h. The heat treatment condition was determinedusing the measured welding temperature of ADC12 duringthe FSW.14) Figure 3 shows SEM images of the size and thedistribution of the Si particles in each part of the base metalsand that of the heat-treated base metal. There is no significanteffect of the heat treatment; it is almost the same in allregions of the base metal.

Figure 4 shows the optimum FSW conditions of ADC12 atthe tool plunge downforce of 14.2 kN.18) Within the range ofthe optimum conditions, the rotation speed and weldingspeed were changed in order to investigate the change in theflow pattern. Figure 5 shows cross sections for all theexperimental conditions. The effect of the welding param-eters on the size, the aspect ratio and the direction angle of the

Si particles for each FSW condition will be discussed indetail in sections 3.2, 3.3 and 3.4.

3.2 Effect of welding parameters on size of Si particles3.2.1 Effect of welding speed

Figure 6 shows the size distribution of the Si particles atvarious welding speeds. In order to diminish the statisticalerror, five photographs were taken and quantified in eachregion. The number of finer Si particles decreases with thelower welding speed. Particularly, in the top region, thenumber of finer Si particles decreases at the lower weldingspeed of 250mm/min compared to the higher welding speedof 750mm/min. This indicates that the Si particles are noteasily reduced in size because the heat input of the topincreases and the flow resistance decreases due to the lowerwelding speed. However, at 1000mm/min, the number of thefiner Si particles decreases in the top region. This is because acavity is formed in the bottom region of the advancing sideand the heat is accumulated in the top region. In this case, theSi size is also very small on the advancing side, which islocated near the defect.

Figure 7 shows the probabilities of the diameter distribu-tion of the Si particles at various welding speeds. Theprobability is defined as the number divided by the totalnumber of Si particles for each position. It is surprising thatthe probabilities of the finer Si particles are not significantlychanged in all regions at a welding speed from 250 to750mm/min, though the numbers themselves are signifi-cantly changed. These results indicate that an identicalphenomenon occurs for all conditions. Only at the higherwelding speed of 1000mm/min, however, the probability ofthe finer Si particles decreases in the top region and increasesin the retreating and bottom regions, which is due to thedefect formation. Figure 8 shows the size probability of theSi particles on a logarithmic scale at various welding speeds.

Base metal(Without heat treatment)

Top

Middle

Bottom

Base metal(With heat treatment: 723K, 3hours)

5µm5µm 5µm

Base metal(With heat treatment: 723K, 5minutes)

Fig. 3 SEM images of Si particles in each part of the base metal and that of the heat treated base metal.

250 500 750 1000 12500

250

500

750

1000

1250

1500

1750

2000FSW, ADC12, 4mmt, 14.2kN, I-Butt

Sound joints

Rot

atio

n sp

eed,

Rt/m

in-1

Welding speed, V/mm.min-1

Defect

Fig. 4 Optimum FSW conditions of ADC12 (4mm thickness).

226 H. Fujii, Y. G. Kim, T. Tsumura, T. Komazaki and K. Nakata

Note that these probabilities are almost symmetrical on thelogarithmic size scale. This result indicates that the crum-bling of the Si particles is the only main mechanism for

decreasing the Si size. When one large Si particle is brokeninto two pieces, two small particles are formed. In this case,the number of particles increases. Thus, it is natural, for

250mm/min 750mm/min 1000mm/min

ADC12, FSW, 4mm thickness, Tool plunge downforce 14.2kN

750m

in-1

1000

min

-112

50m

in-1

1500

min

-117

50m

in-1

1mm

Defect

Defect

Defect

500mm/min

500m

in-1

Fig. 5 Cross sections for all the experimental conditions.

Top

0

50

100

150

200

250

300

350

400

Num

ber

Diameter of Si particle, d/µm

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

(a) 250mm/min

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9 2.1~2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(b) 500mm/min

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(c) 750mm/min

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(d) 1000mm/min

Fig. 6 Size distribution of the Si particles at various welding speeds

(14.2 kN, 1250min�1).

(a) 250mm/min

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(b) 500mm/min

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9 1.1~1.3~1.1

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(c) 750mm/min

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(d) 1000mm/min

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

Fig. 7 Probability of Si size distribution at various welding speeds

(14.2 kN, 1250min�1).


example, that the number of particles with eight times thevolume is one eighth of the number of smaller particles.15,16)

Consequently, the distribution of the Si particles is sym-metrical on a logarithmic scale when the crumbling of the Siparticles is the only main mechanism for decreasing the Sisize. It is known that a similar distribution (Gaussiandistribution) is generally obtained on a logarithmic scalefor the size distribution of metal powder.17)

3.2.2 Effect of rotation speedFigure 9 shows the size distribution of the Si particles at

various rotation speeds. At the lower rotation speed of 500 or750min�1, the number of finer Si particles in the bottomregion is larger than that of the other regions. Under theseconditions, the Si particles easily break down in the bottomregion due to the lower temperature. The change in thenumber of finer Si particles is not simple, when the rotationspeed is varied from 750 to 1750min�1. It is considered thatthe rotation speed influences the plastic flow rate as well asthe heat input. The Si particle size should increase with theincreasing heat input, as discussed in the previous section(Effect of welding speed). However, the size should decreasewith the increasing flow rate, because the stirring effectincreases. Based on the two different effects, the number of Siparticles decreases when the rotation speed is changed from750 to 1250min�1, but it increases when the rotation speed ischanged from 1250 to 1500min�1. At the higher rotationspeed of 1750min�1, the number of finer Si particles in thetop region decreased, because the welding conditions aresimilar to the conditions under which abnormal stirringoccurs near the sample surface.18) Also, the Si particles in themiddle region are smaller than that of the other regions,because the temperature is higher even in the middle regionunder this condition. On the contrary, at the lower rotationspeed of 500min�1, a long crack is formed in SZ. The size ofthe Si particles in the bottom and middle regions are muchsmaller than the other regions, because the temperature islower even in the middle region under this condition.

Figure 10 shows the probabilities of the diameter distri-

bution of the Si particles at various rotation speeds. It issurprising that the probabilities of the finer Si particles are notsignificantly changed in all regions, when the rotation speedis varied from 750 to 1750min�1, similar to the weldingspeed. This result indicates that an identical phenomenon,namely break down of the Si particles, occurs under allconditions. When a logarithmic scale is used, the distributionis almost symmetrical, as in changing the welding speed.

3.3 Effect of welding parameters on aspect ratio of Siparticles

Figure 11 shows the aspect ratio distribution of the Siparticles at various welding speeds. The number of smalleraspect ratio particles decreases in all regions with the lowerwelding speed. The lower aspect ratio indicates that theparticles are significantly crumbled. By comparing Figs. 11with 6, it is easily realized that when the Si particles aresignificantly crumbled by a strong stirring action, both thesize and aspect ratio are decreased. At the lower weldingspeed of 250mm/min, the smaller aspect ratio is decreased inthe top and middle regions. At the higher welding speed of1000mm/min, the elongated Si particles are increased onlyin the top region due to the defect formation.

Figure 12 shows the aspect ratio distribution of the Siparticle at various rotation speeds. A similar tendency can beseen when the heat input is higher, that is, the particles aremore elongated and larger. When the heat input is muchhigher, namely, when the rotation speed is at 1500min�1 orhigher, the ratio of the particle with the aspect ratio of 2.0 is


0.55

0.61

0.50

0.67

0.74

0.90

0.81

0.99

1.09

1.32

1.45

1.20

1.60

1.76

2.13

1.93

2.34

2.57

2.83

3.11

3.77

3.42

3.77

~

Top

Retreating side

Middle

Advancing side

Bottom

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

(a) 250mm/min

Pro

babi

lity

(%)


0.55

0.61

0.50

0.67

0.74

0.90

0.81

0.99

1.09

1.32

1.45

1.20

1.60

1.76

2.13

1.93

2.34

2.57

2.83

3.11

3.77

3.42

3.77

~

Top

Retreating side

Middle

Advancing side

Bottom

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

(b) 500mm/minP

roba

bilit

y (%

)


0.55

0.61

0.50

0.67

0.74

0.90

0.99

0.81

1.09

1.32

1.45

1.20

1.60

1.76

2.13

1.93

2.34

2.57

2.83

3.11

3.77

3.42

3.77

~

Top

Retreating side

Middle

Advancing side

Bottom

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

(c) 750mm/min

Pro

babi

lity

(%)


0.55

0.61

0.50

0.67

0.74

0.90

0.81

0.99

1.09

1.32

1.45

1.20

1.60

1.76

2.13

1.93

2.34

2.57

2.83

3.11

3.77

3.42

3.77

~

Top

Retreating side

Middle

Advancing side

Bottom

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

Pro

babi

lity

(%)

(d) 1000mm/min

Fig. 8 Size probability of the Si particles on logarithmic scale at various

welding speeds (14.2 kN, 1250min�1).

0

50

100

150

200

250

300

350

400

450

500

550

600

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3 ~1.5

1.3 1.5~1.7

2.5~2.7

1.7~1.9 ~2.1

1.9 2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Num

ber


(a) 500min-1

Top

Retreating side

Middle

Advancing side

Bottom

0

100

200

300

400

500

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(b) 750min-1

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(c) 1000min-1

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7~1.9

2.5~2.7

1.7 1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(d) 1250min-1

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

BottomN

umbe

r


(e) 1500min-1

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


(f) 1750min-1

Fig. 9 Size distribution of the Si particles at various rotation speeds

(14.2 kN, 500mm/min).


higher. This is related to the fact that the number of finerparticles increases at 1500min�1. At the higher rotationspeed of 1750min�1, the smaller aspect ratio as well as thenumber of smaller particles decrease in the middle region,because the heat input is large enough and the weldingconditions are similar to the conditions under which theabnormal stirring occurs near the sample surface.18)

3.4 Effect of welding parameters on direction angle of Siparticles

3.4.1 Effect of welding speedFigure 13 shows the direction angle probabilities of the Si

particles at various welding speeds. The direction angle isdefined as the angle between the major axis of the particleand the Y-axis (a vertical line). When the direction angle is 90degrees, the major axis of the particle is perpendicular to theY-axis. The welding speed influences the direction angle ofthe Si particles in the bottom region, whereas, it significantlydoes not in the top region. The most probable direction angleis constant at about 90 degrees in the top region at anywelding speed. In the bottom region, on the other hand, themost probable direction angle changes from about 80 degreesin the bottom region at 250mm/min, which moves to thelower angle with the increasing welding speed.

The directions of the Si particles in the advancing andretreating sides are also almost constant at 0 or 180 degrees atany welding speed. However, when the cavity is formed inthe advancing side of the stir zone at the higher weldingspeed of 1000mm/min, the direction angle of the Si particlesis random on the advancing side. Due to the defect formation,the material flow on the advancing side is separated from thematerial flow of the entire stir zone. Figure 14 is a schematicillustration of the material flow in the stir zone at variouswelding speeds. The flow directions in the stir zone areshown by the arrows. In fact, the material flow should bethree-dimensional during FSW. However, in this study, Siparticle direction of the stir zone is observed on the cross

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

(a) 500min-1

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3 ~1.5

1.3 1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1~3.3

3.3~3.5

3.9~3.1 3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(b) 750min-1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(c) 1000min-1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(d) 1250min-1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9 3.1~3.1

3.3~3.5

3.9~~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

~2.1

(e) 1500min-1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

(f) 1750min-1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9


Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)

Fig. 10 Probability of Si size distribution at various rotation speeds

(14.2 kN, 500mm/min).

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber

Aspect ratio of Si particle

Top

Retreating side

Middle

Advancing side

Bottom

(a) 250mm/min

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(b) 500mm/min

3.60

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(c) 750mm/min

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(d) 1000mm/min

Fig. 11 Aspect ratio distribution of the Si particles at various welding

speeds (14.2 kN, 1250min�1).

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8 4.0~3.6~4.8 ~4.0

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(a) 500min-1

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(b) 750min-1

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(c) 1000min-1

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(d) 1250min-1

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(e) 1500min-1

0

50

100

150

200

250

300

350

400

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8

Num

ber


Top

Retreating side

Middle

Advancing side

Bottom

(f) 1750min-1

Fig. 12 Aspect ratio distribution of the Si particles at various rotation

speeds (14.2 kN, 500mm/min).


section of the specimen. Accordingly, the flow shown inFig. 14 indicates a flow vector on the cross section plane. Thematerial flow is attenuated on the advancing side of the crosssection at the higher welding speed of 1000mm/min. Thiskind of identification of the material flow might be veryuseful to determine the optimal welding conditions byavoiding the defect formation.3.4.2 Effect of rotation speed

Figure 15 shows the direction angle probabilities of the Siparticles at various rotation speeds, and Figure 16 shows theestimated schematic illustration of the material flow in thestir zone from the results.

At the lower rotation speed of 750min�1, the directionangle was random in the bottom region. This result indicatesthat the particular material flow does not easily occur due tothe lower heat input at the lower rotation speed, although the

Si particles have significantly crumbled, as mentioned in theprevious section. At 500min�1, the direction angle wasrandom both in the bottom and retreating regions, whichindicated a discrete flow in these regions. Under suchconditions, a defect is formed due to the insufficient heatinput.

At the higher rotation speed of 1750min�1, the directionangle is random on the advancing side. This condition shouldbe near the conditions where abnormal stirring occurs nearthe sample surface, as mentioned in the next section. With asomewhat higher heat input, the symmetrical characteristic ofmaterial flow becomes better.

3.5 Effect of defect formation by abnormal stirring on Siparticle distribution

Because of the rotation speed limit of 1750min�1 for theFSW machine, FSW was performed at 14.2 kN, 1500min�1

and 750mm/min. Figure 17 shows the size and aspect ratiodistribution of Si particles in the stir zone. Under thiscondition, the cavity was formed in the stir zone due to theabnormal stirring,8) as shown in Fig. 5. The number of finerSi particles is smaller than that under the defect formationconditions due to an insufficient heat input. This means thatthe Si particles are not easily crumbled due to the abnormalstirring action.

Figure 18 shows the direction angle probability and aschematic illustration of the material flow in the stir zoneunder the same condition as in Fig. 17. The direction angle issignificantly random in the top region. This result indicates

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom

Direction angle of Si particle (° )

(d) 1000mm/min

Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom


(c) 750mm/min

Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom


(a) 250mm/min

Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom


(b) 500mm/minP

roba

bilit

y (%

)

Fig. 13 Direction angle probabilities of the Si particles at various welding

speeds (14.2 kN, 1250min�1).

Cavity

(a) 250mm/min

(c) 750mm/min (d) 1000mm/min

(b) 500mm/min

RS AS RS AS

RS AS RS AS

Fig. 14 Schematic illustration of the material flow in the stir zone at

various welding speeds (14.2 kN, 1250min�1).

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom

(a) 500min-1


Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom


(b) 750min-1

Pro

babi

lity

(%)

Top

Retreating side

Middle

Advancing side

Bottom

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30


0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

(c) 1000min-1

Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom


(d) 1250min-1

Pro

babi

lity

(%)

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170

Top

Retreating side

Middle

Advancing side

Bottom

Pro

babi

lity

(%)


(e) 1500min-1

Top

Retreating side

Middle

Advancing side

Bottom

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170


(f) 1750min-1

Pro

babi

lity

(%)

Fig. 15 Direction angle probabilities of the Si particles at various rotation

speeds (14.2 kN, 500mm/min).


that an abnormal stirring action occurs. Although at thedistance of 2.4mm from the center on the advancing side, thedirections of the Si particles are almost constant at 0 or 180degrees, the direction angle is random on the advancing sideat the distance of 1.8mm from center near the cavity. Thismaterial flow is very similar to that under the conditions of1250min�1–1000mm/min and 1750min�1–500mm/min.On the other hand, under insufficient heat input conditions,such as 500min�1–500mm/min and 750min�1–500mm/min, shown in Fig. 16, the material flow in the bottom andretreating side regions becomes random.

4. Conclusions

An aluminum die casting alloy ADC12 was joined under

various welding conditions by FSW. The effect of weldingparameters on the material flow in the stir zone for the FSWjoints was investigated in detail. As a result, the followingconclusions were reached.(1) The size and aspect ratio of the Si particles decrease

with the increasing welding speed, because Si particleseasily break down at a lower temperature. They are notsimply changed when the rotation speed is changedbecause both the heat input and the plastic flow rate aresimultaneously changed.

(2) The probabilities of the Si particle size distribution arealmost the same in all regions, even when the weldingspeed or the rotation speed is varied. This resultindicates that an identical phenomenon occurs todecrease the Si particles. The main phenomenon shouldbe crumbling Si particles due to stirring by the toolbecause the size distribution of the Si particles issymmetrical on a logarithmic scale for all conditions.

(3) Under the proper conditions, the flow directions of theSi particles in the top and bottom are horizontallyaligned, while they are aligned longitudinally on theretreating side and advancing side. The directionsreflect the material flow during the FSW.

(4) Under insufficient heat input conditions, the materialflow becomes random in the bottom and retreating sideregions. Under the abnormal stirring conditions, on theother hand, it is random in the advancing side and topregions, even though a defect forms in the advancingside in both cases. This kind of identification of thematerial flow might be very useful to determine theoptimal welding conditions by avoiding defect forma-tion.

Acknowledgments

This work was supported by the Toray Science Founda-tion, the 21st century COE program, the Three ResearchInstitute Project: Development Base of Joining Technologyfor New Metallic Glasses and Inorganic Materials and aGrant-in-Aid for Science Research from Japan Society forPromotion of Science.

REFERENCES

1) W. M. Thomas: International Patent Appl. No. PCT/GB92/02203 and

GB Patent Appl. No. 9125978.8 (1991) U.S. Patent No. 5,460,317.

(a) 500min-1

RS AS

Crack

(b) 750min-1

RS AS

(c) 1000min-1

RS AS

(d) 1250min-1

RS AS

(e) 1500min-1

RS AS

(f) 1750min-1

RS AS

Fig. 16 Schematic illustration of the material flow in the stir zone at

various rotation speeds (14.2 kN, 500mm/min).

1.0~1.2

1.2~1.4

1.4~1.6

1.6~1.8

1.8~2.0

2.0~2.2

3.0~3.2

2.2~2.4

2.4~2.6

2.6~2.8

2.8~3.0

3.2~3.4

3.4~3.6

3.8~4.0

4.0~3.6~4.8


0

50

100

150

200

250

300

350

400

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber

0

50

100

150

200

250

300

350

400

0.5~0.7

0.7~0.9

0.9~1.1

1.1~1.3

1.3~1.5

1.5~1.7

2.5~2.7

1.7~1.9

1.9~2.1

2.1~2.3

2.3~2.5

2.7~2.9

2.9~3.1

3.3~3.5

3.9~3.1~3.3

3.5~3.7

3.7~3.9

Top

Retreating side

Middle

Advancing side

Bottom

Num

ber


Fig. 17 Size and aspect ratio distribution of Si particles in the stir zone for

abnormal stirring (14.2 kN, 1500min�1 and 750mm/min).

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

Pro

babi

lity

(%)

Top

Retreating side

Middle

Advancing side (2.4mm)

Bottom

0~10

10~20

20~30

30~40

40~50

50~60

100~110

60~70

70~80

80~90

90~100

110~120

120~130

140~150

170~180

130~140

150~160

160~170


Advancing side (1.8mm)

Cavity

RS AS

Fig. 18 Direction angle probability and schematic illustration of the

material flow in the stir zone for abnormal stirring (14.2 kN, 1500min�1

and 750mm/min).


2) Y. Li, L. E. Murr and J. C. McClure: Mater. Sci. Eng. A 271 (1999)

213–223.

3) Y. Li, L. E. Murr and J. C. McClure: Scr. Mater. 40 (1999) 1041–1046.

4) K. Colligan: Welding J. 78 (1999) 229S–237S.

5) A. P. Reynolds: Sci. Technol. Welding J. 5 (2000) 120–124.

6) Y. S. Sato, H. Kokawa, K. Ikeda, M. Enomoto, S. Jogan and

T. Hashimoto: Metall. Mater. Trans. A 32A (2001) 941–948.

7) S. Xu, X. Deng, A. P. Reynolds and T. U. Seidel: Sci. Technol.

Welding J. 6 (2001) 191–193.

8) D. P. Field, T. W. Nelson, Y. Hovanski and K. V. Jata: Metall. Mater.

Trans. A 32A (2001) 2869–2876.

9) T. U. Seidel and A. P. Reynolds: Metall. Mater. Trans. A 32A (2001)

2879–2884.

10) J. H. Ouyang and R. Kovacevic: J. Mater. Eng. Perform. 11 (2002) 51–

63.

11) M. Guerra, C. Schmidt, J. C. McClure, L. E. Murr and A. C. Nunes:

Mater. Character. 49 (2003) 95–101.

12) L. KE, L. Xing and J. E. Indacochea: Metall. Mater. Trans. B 35B

(2004) 153–160.

13) J. A. Schneider and A. C. Nunes, Jr.: Metall. Mater. Trans. B 35B

(2004) 777–783.

14) Y. G. Kim, K. Nakata, T. Murakami, M. Maeda, M. Ushio and

T. Komazaki: Preprints of the national meeting of J.W.S., No. 73

(2003) pp. 44–45.

15) H. Fujii, Y. Aoki and K. Nogi: Trans. JWRI 30 (2001) 105–109.

16) K. Nogi, Y. Aoki, H. Fujii and K. Nakata: Acta Mater. 46 (1998) 4405–

4413.

17) R. M. German: Powder Metall. Sci., second edition, pp. 51–54.

18) Y. G. Kim, H. Fujii, T. Tsumura, T. Komazaki and K. Nakata: Mater.

Sci. Eng. A 415 (2006) 250–254.


Documents

Estimation of Material Flow in Stir Zone during Friction ... of Material Flow in Stir Zone during Friction Stir Welding by Distribution Measurement of Si Particles Hidetoshi Fujii