You are on page 1of 8

Effect of friction welding parameters on

mechanical and metallurgical properties of


aluminium alloy 5052 A36 steel joint
W. B. Lee, Y. M. Yeon, D. U. Kim and S. B. Jung
The mechanical and metallurgical properties of friction welded joints between type 5052 aluminium alloy and type
A36 steel have been studied in the present work. Joint strength increased with increasing upset pressure and friction
time until it reached a crictical value. The strength of the joint settled at a lower value, compared with that of the
base metal, in the case of increasing friction time, caused by the formation of an intermediate phase (intermetallic
compound, oxides). The microstructure of 5052 alloy was greatly deformed near the weld interface, and underwent
dynamic recrystallisation owing to frictional heat and deformation resulting from the friction welding process.
Therefore, a very ne and equiaxed grain structure was observed near the interface. Elongated grains were observed
outside the dynamic recrystallisation region at the peripheral part, while the A36 steel side was not deformed. The
hardness of the near interface was slightly softer than that of the 5052 alloy base metal, and maximun softened width
was ~8 mm from the interface. In the present work, the conditions of friction time t1~0.5 s and upset pressure
P2~137.5 MPa gave maximum joint strength of 202 MPa when the friction pressure, upset time and rotation speed
were xed at 70 MPa, 5 s and 2000 rev min2 1, respectively, and these were the optimum friction welding conditions
for the aluminium alloy 5052 A36 steel joint.
MST/5494
Dr Lee and Professor Jung (sbjung@skku.ac.kr) are at the Advanced Materials and Process Research Centre for IT,
Sungkyunkwan University, 300 Cheoncheon-dong, Jangan-gu, Suwon, Gyounggi-do, 440 746, Korea, Dr Yeon is in the
Department of Automatic-Welding Engineering, Suwon Science College, Whasung, Gyounggi-do, 445 742, Korea and Dr Kim
is at the Research Institute, Hyundai Mobis Co., Youngin, Gyounggi-do, 449 910, Korea. Manuscript received 22 April 2002;
accepted 27 August 2002.
# 2003 IoM Communications Ltd. Published by Maney for the Institute of Materials, Minerals and Mining.

Introduction
Friction welding is a solid state welding process, which
means that the joining is carried out at a temperature
below the melting points of the metals to be joined.1 ,2 The
basis of the friction welding process is that the relative
motion of two faying surfaces is used to generate heat,
and this is accompanied by a compressive force which
plastically deforms the surfaces, resulting in the joint.
Practical structures are becoming increasingly complicated, so it is dif cult to manufacture them using only
one material to satisfy environmental and service requirements. Ideally, such structures should be manufactured
from a combination of materials. It is necessary, therefore,
to establish a bonding technique that can readily join dissimilar metals.3 ,4
There are numerous commercial applications for aluminium alloy steel joints, ranging from cryogenic pressure
vessels and anode assemblies to the manufacture of plate
iron and cooking utensils. Friction welding is an obvious
contender in the quest for acceptable properties, compared
with other joining methods for dissimilar metal combinations.5 In the case of fusion welding in the Fe Al system,
excess formation of brittle intermetallic compounds degrades
the joint strength. Nevertheless, since friction welding is a
solid state bonding processes, low thickness intermetallic
layers are formed at the weld interface.6 However, in the
Fe Al system, solid solubility is almost zero, so some intermetallic compounds will be formed despite the application
of friction welding. The major problem in friction welding
aluminium steel joints is the formation of intermetallics
and oxides at the interface. The thickness of such layers
must be controlled if sound welds are to be achieved. In
some studies, Fe2 Al5 , FeAl3 , FeAl and Fe 3 Al intermetallics
were observed at the weld interface in pure aluminium
austenitic stainless steel welds.7 ,8 In other studies, Fe2 Al 5
DOI 10.1179/026708303225001876

was formed at the weld interface in 6061 aluminium alloy


304 stainless steel joints9 and Fe 4 Al 1 3 and Fe2 Al 5 were
observed in pure aluminium carbon steel joints.1 0
It is still unknown how the welding parameters affect
the formation of intermetallic layers in and the mechanical
properties of aluminium alloy steel joints. In the present
study, commercial 5052 aluminum alloy was bonded to
grade A36 steel using a brake type friction welding machine.
To elucidate the relationship between microstructure and
mechanical properties, the microstructural change from
the interface to the unaffected part was observed by optical
microscopy, and the variation of hardness was measured.
The intermetallic layer formed at the interface was studied
by scanning electron microscopy (SEM) and energy dispersive spectrometry (EDS) analysis. The mechanical properties were determined using tensile and hardness testing. The
aim was to establish optimum conditions of friction welding
5052 A36 joints.

Experimental procedure
The materials used in the present work were commercially
available 5052 aluminium alloy and type A36 steel (0.2%
carbon), which were machined to a rod shape 20 mm in
diameter and 120 mm in length. The chemical compositions
are given in Table 1. The 5052 alloy was supplied in the
form of cold drawn bar.
The surfaces of the parts were ground with SiC paper
(grade 100) and cleaned with acetone before welding. Friction welding was carried out using a brake type friction
welding machine (Nitto Seike Co. Ltd). Friction welding
parameters are rotation speed N, friction time t1 , upset
time t2 , friction pressure P1 and upset pressure P2 when
welding with this machine. In this present work, t2 , P1
and N were xed at 5 s, 70 MPa, and 2000 rev min2 1 ,
Materials Science and Technology

June 2003

Vol. 19

773

774 Lee et al. Effect of friction welding on properties of Al alloy 5052 A36 steel joint

1 Schematic illustration of friction welding process and shape of welded specimens

2 Cross-sectional macrostructures of friction welded joints for given welding conditions

respectively. The friction time was varied from 0.1 to 3.0 s


and the upset pressure from 70 to 150 MPa. A schematic
illustration of the friction welding process and geometry
of the welded specimens is shown in Fig. 1. The 5052
alloy was rotated, while the A36 steel was held and exerted
the force.
The resultant welds were sliced using a diamond cutting
wheel to provide cross-sections, avoiding thermal degradation. They were then ground with SiC paper (grade 100
2000), and nally micropolishedusing 0.05 mm Al2 O3 powder.
The 5052 alloy specimen faces were etched with Kellers
reagent1 1 (150 mL water, 3 mL nitric acid, 6 mL hydro uoric acid, 6 mL hydrochloric acid) for 0.18 ks and A36
steel specimen faces were etched with nital (3% nitric acid
in methyl alcohol) for 15 s, before metallurgical examination. Microstructures of the friction welded interfaces were
observed by optical microscopy and SEM.
The width of the dynamic recrystallisation (DRX)
region at the weld interface was determined using an
image analyser.
Tensile tests were carried out to evaluate the mechanical
properties of the joints. Testing was conducted at room
temperature using an Instron type machine with 1.676
102 2 mm s2 1 crosshead speed. The Vickers hardness distribution of each material in the vicinity of the weld interface was
measured with a load of 0.98 N, for 10 s.

Results
MICROSTRUCTURE EVOLUTION
Macrographs of welded joint cross-sections are shown in
Fig. 2. In the case of dissimilar material joints, formation of
the ash depends on the mechanical properties of the two
Table 1 Chemical compositions of materials used in present study, wt-%
C

Si

Mn

Cr

Mg

Al

Fe

5052 alloy
0.094

0.113 0.006 2. 51 Bal.


A36 steel 0. 18 0.18 0.58 0.25 0.1

Bal.

Materials Science and Technology

June 2003 Vol. 19

parent materials. In the present case, ashes from the


interface were formed symmetrically around the weld circumference on the 5052 alloy side. The amount of ash increased
with increasing friction time and upset pressure. The A36
steel side was not deformed because this material has
higher strength than the aluminium alloy, and is thus more
resistant to deformation. Hence, the formation of ashes
was restricted to 5052 alloy only.
Optical microstructures of weld interfaces and base metals
are shown in Fig. 3. The 5052 alloy microstructure at the
weld interface was very different from that of the base
metal. However, the A36 steel microstructuredid not change,
relative to the base metal. Microstructural features of the
5052 alloy could be divided into four regions similar to
those observed in friction stir welding:1 2 ,1 3 (a) the DRX
region,1 4 ,1 5 showing re ned and equiaxed grains caused by
strong plastic deformation and heat generation during
the friction welding process, with grain size 10 times smaller
than that of the base metal; (b) the heat and deformation
affected zone (HDZ) showing a owing grain structure
from the central part to the peripheral part, so elongated
grains were formed; (c) the heat affected zone (HAZ)
located outside the HDZ, showing no difference in grain
size with respect to the optical appearance of the base metal,
and only distinguished by hardness tests; and (d) the base
metal. The deformed region increased with increasing friction time and decreased with increasing upset pressure.
The peripheral part showed a wider deformed region
than the central part owing to the difference in rotation
speed (illustrated in Figs. 4 and 5).
It is known that plastic deformation and heat generation occurring simultaneously during the friction welding
process lead to DRX. Figure 6 shows the width of the
DRX region in 5052 alloy from the weld interface to the
base metal for various welding conditions. The width of
the DRX region was narrow at the centre and broader
towards the periphery. The reason for this is that the peripheral part underwent a stronger plastic ow and experienced more heat than the central part. At the edge of
the peripheral part, the DRX region was narrow again
because of the formation of the ash outside the interface.
In the case of increasing the upset pressure from 70 to
92.5 MPa, the width of the DRX region was decreased

Lee et al.

Effect of friction welding on properties of Al alloy 5052 A36 steel joint

775

3 Microstructures of weld interfaces at central and peripheral regions and base metals

at the centre but the maximum width was greater. Increasing the friction time diminished the difference in DRX
width between the centre and the periphery so that the
width of the DRX region was similar regardless of location. Aluminium alloys are high thermal conductivity
materials, and a longer friction time caused the present
5052 alloy to attain a homogeneous temperature near the
interface.
Fukumoto et al. reported that the hardness pro le of the
cross-section of a weld interface could be lower or higher
than the hardnesses of the base materials, depending on the
metals welded and the welding parameters.1 6 1 8 Generally,
aluminium alloys are divided into two groups: precipitation hardening alloys, after a speci c heat treatment (aging),
including the 2000, 6000 and 7000 alloy systems, and solid
solution strengthening alloys, including the 3000, 4000 and

5000 alloy systems.1 9 The former is known to undergo


thermal softening near the weld interface owing to the
loss of precipitates as a result of the welding heat, but the
latter does not exhibit this effect. Figure 7 shows hardness
pro les of the weld interface on the 5052 alloy side. A
softened area was observed near the interface, different from
the above assumption. The present 5052 alloy, recovered
and recrystallised as a result of friction heat and deformation, was slightly softened2 0 compared with the original,
cold drawn 5052 alloy, which was already work hardened
before the friction welding procedure. There was no difference in hardness values among the measured locations
(centre, periphery, half radius R/2). The softened area
showed lower hardness at about 60 70 HV(0.98 N) than
that of the base metal, which showed a scattered hardness range from 85 to 90 HV(0.98 N). The softened area
Materials Science and Technology

June 2003

Vol. 19

776 Lee et al. Effect of friction welding on properties of Al alloy 5052 A36 steel joint

a P2~70 MPa, central region; b P2~70 MPa, peripheral region;


c P2~115 MPa, central region; d P2~115 MPa, peripheral
region

a t1~0.5 s, central region; b t1~0.5 s, peripheral region;


c t1~2.0 s, central region; d t1~2.0 s, peripheral region

Optical microstructure of 5052 alloy near interface: variation with increasing friction time

Distributions of width of dynamic recrystallisation


(DRX) region at weld interface for given welding
conditions

4 Optical microstructure of 5052 aluminium alloy near


interface: variation with increasing upset pressure

could be divided into regions A, B and C (Fig. 7a)


with different hardness values. The hardness of region A
was slightly higher than that of region B. Region A
underwent both plastic deformation and heating, so
the microstructure was effectively forged, but region B
received only heat.4 Region C corresponded to the recovered
region. The softened area became wider as friction time
increased, and narrower with increasing upset pressure.
The maximum and minimum width of the softened
area from the weld interface was ~8 mm and 5 mm,
respectively.
Figure 8 shows SEM images of the weld interface.
An intermetallic layer was observed by SEM with t1 ~
0.1 s and t1 ~1.5 s. The thickness of the intermetallic
reaction layer increased slightly with increasing friction
time. Generally, a thicker intermetallic layer was observed
at the peripheral region.2 0 When friction time was 0.1 s,
the layer was almost the same thickness at both central and peripheral regions; an intermetallic layer was
formed only in 5052 alloy. At a friction time of 1.5 s,
the thickness of the reaction layer was increased, and
formed in both 5052 alloy and A36 steel. Many elements
(Fe, Al, Mg, Si, O, C) were detected by EDS analysis

(Fig. 8). Probably, this reaction layer was composed of


mainly FeAl intermetallic and MgO oxide. X-ray diffractometry was used to determine the phases of this
layer, but types of intermetallic compound and oxide
could not be detected because the layer was very thin,
within 1 mm.

a given measured locations; b central region, given welding conditions; c peripheral region, given welding conditions

7 Vickers hardness distributions of 5052 alloy near weld interface

Materials Science and Technology

June 2003 Vol. 19

Lee et al.

Effect of friction welding on properties of Al alloy 5052 A36 steel joint

777

a P2~70 MPa; b P2~92.5 MPa; c P2~137. 5 MPa

10

a t1~0 .1 s; b t1~1.5 s

8 Images (SEM) and EDS analysis of weld interface

JOINT STRENGTH OF FRICTION WELDED


JOINTS
It is known that the parameters t1 and P2 have the greatest effect on joint strength, among the friction welding
parameters (t1 , t2 , P1 , P2 , N). In the present work, t1 and
P2 were varied and the other welding parameters were
xed.
The relationship between upset pressure P2 and joint
strength and amount of ash is shown in Fig. 9a. Tensile
strength increased with increasing upset pressure and nally
showed a settled value, 202 MPa, which was ~98% of the
strength of the 5052 alloy base metal. The joint strength
became constant at an upset pressure of 137.5 MPa.

Tensile fracture surfaces of A36 steel side

Figure 10 shows the appearance of the fracture surface


with increasing upset pressure on the A36 steel side. Fracture occurred exactly at the weld interface when the upset
pressures were 70 and 92.5 MPa. Both peripheral and
central parts showed a few scars on the 5052 alloy side, and
the unbonded or intermetallic layer region was observed
across the whole weld interface, resulting in lower joint
strength than that of the base aluminium alloy. Over an
upset pressure of 137.5 MPa, a fully bonded region had
formed at the peripheral part and a partially unbonded
region existed at the central part, and fracture occurred in
the 5052 alloy or HAZ; in this case, joint strength was
increased, relative to that with a lower upset pressure. The
amount of ash increased proportionally with increasing
upset pressure.
Figure 9b shows the relationship between friction time
t1 and joint strength and amount of ash. Joint strength
was increased slightly with increasing friction time up to
0.5 s. At friction times over 0.5 s, joint strength settled
at a constant value of 180 MPa. The location of fracture
was near the weld interface. The possibility of intermetallic
compound formation increased with increasing friction
time because the temperature of the weld interface was
high. However, the friction welding process included an

joint strength; d amount of ash


a upset pressure; b friction time

9 Relationship between given welding parameters and tensile strength and amount of ash

Materials Science and Technology

June 2003

Vol. 19

778 Lee et al. Effect of friction welding on properties of Al alloy 5052 A36 steel joint

upsetting stage, which produced deformation ash. The


reaction layers and oxides might have been expelled from
the weld interface so that few existed at the weld interface
(thickness less than 1 mm, described above). With increasing
friction time, a larger amount of ash was formed from
the faying surfaces. However, joint strength settled at a
constant value regardless of friction time. A very thin
intermetallic layer was formed at the interface, and therefore this layer had only a slight effect on joint strength
regardless of friction welding conditions.
In the present work, the conditions of t1 ~0.5 s and
P2 ~137.5 MPa resulted in the maximum joint strength
(202 MPa), and these were the optimum friction welding
parameters for the 5052 A36 joint.

Conclusions
Aluminium alloy 5052 and type A36 steel were joined
using a brake type friction welding machine. The mechanical and metallurgical properties were investigated by
means of tensile testing, microstructural examination and
hardness testing. The major conclusions are summarised
as follows.
1. Aluminium alloy 5052 was greatly deformed, relative
to the base metal, and the features could be divided into
four regions: (a) dynamic recrystallisation region, (b) heat
and deformation affected zone, (c) HAZ and (d) base metal.
The grains were ne and equiaxed near the weld interface
and grains were elongated beside the dynamic recrystallisation region. However, the A36 steel was not deformed.
2. A softened area was formed near the weld interface.
Increasing upset pressure minimised that area and increasing friction time widened the area. The maximum width
of the softened area was ~8 mm from the weld interface.
3. The thickness of the intermetallic reaction layer increased with increasing friction time. It was thicker in the peripheral region than in the central region. This layer was
identi ed as probably FeAl and MgO.
4. Joint strength increased and then settled at a constant
value after reaching a maximum with increasing upset
pressure and friction time. A very thin intermetallic layer
was formed at the weld interface.

Materials Science and Technology

June 2003 Vol. 19

Acknowledgements
This work was supported by the Advanced Materials
and Process Research Centre for IT at Sungkyunkwan
University (Grant no. R12 2002 057 03001 0).

References
1. b. s. yilbas, a. z. sahin, n. kahraman and a. z. al-garni:
J. Mater. Process. Technol., 1995, 49, 431 443.
2. j. ruge, k. thomas, c. eckel and s. sundaresan: Weld. J.,
August 1986, 28 31.
3. s. fukumoto, t. inuki, h. tsubakino, k. okita, m. aritosh and
t. tomita: Mater. Sci. Technol., 1997, 13, 679 686.
4. r. a. bell, j. c. lippod and d. r. anolphson: Weld. Res. Suppl.,
1984, 11, 325 332.
5. g. m. workmanande. d. nicholas: Met. Mater., 1986,2, 138 140.
6. s. fukumoto, h. tsubakino, k. okita, m. aritosh and
t. tomita: Scr. Mater., 2000, 42, 807 812.
7. s. sunderesan and k. g. murte: Mater. Forum, 1993, 17, 301 307.
8. s. fukumoto, h. tsubakino, k. okita, m. aritosh and
t. tomita: Mater. Sci. Technol., 1998, 14, 333 338.
9. s. fukumoto, m. ohashi, h. tsubakino, k. okita, m. aritosh,
t. tomita and k. goto: J. Jpn Inst. Light Met., 1998, 48, 36.
10. m. kikuchi, h. takeda and s. morozumi: J. Jpn Inst. Light
Met., 1984, 34, 165 173.
11. g. liu, l. e. murr, c.-s. niou, j. c. mcclure and f. r. vega:
Scr. Mater., 1997, 37, 365 367.
12. c. j. dawes: Proc. 6th Int. Symp. of JWS, Nagoya, Japan,
November 1996, 711 718.
13. k. v. jata and s. l. semiatin: Scr. Mater., 2000, 43, 743 749.
14. b. ren and j. g. morris: Metall. Mater. Trans. A, 1995, 26A,
31 40.
15. w. blum and h. j. mcqueen: Mater. Sci. Forum, 1996, 217, 31 42.
16. h. ochi, k. ogawa, s. kaga, h. ohke and y. suga: J. Jpn Inst.
Light Met., 1993, 43, 365 371.
17. t. asahima, k. kato and h. tokisue: J. Jpn Inst. Light Met.,
1991, 41, 674 680.
18. s. fukumoto, m. ohashi, h. tsubakino, k. okita, m. aritosh
and t. tomita: Proc. 3rd Paci c Rim Int. Conf. on Advanced
materials and processing, Hawaii, July 1998, TMS, 2247
2252.
19. t. h. courtney: Mechanical behavior of materials, 162 214;
1990, New York, McGraw-Hill.
20. s. fukumoto, h. tsubakino, k. okita, m. aritosh and
t. tomita: Mater. Sci. Technol., 1999, 15, 1080 1086.

Copyright of Materials Science & Technology is the property of Maney Publishing and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.

You might also like