IIW Commission III Doc. III-1290-04
Properties of Low Carbon Steel Joint by Low Heat Input Friction Welding Method
by Masaaki KIMURA, Masahiro KUSAKA, Kenji SEO and Akiyoshi FUJI
Properties of Low Carbon Steel Joint by Low Heat Input Friction Welding Method by Masaaki KIMURA*, Masahiro KUSAKA*, Kenji SEO* and Akiyoshi FUJI** *Department of Mechanical and System Engineering, Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji, Hyogo 671-2201, Japan **Department of Mechanical Engineering, Faculty of Engineering, National University Corporation-Kitami Institute of Technology 165 Koen-cho, Kitami, Hokkaido 090-8507, Japan. Abstract This report describes the properties of friction welded joint of low carbon steel by using Low Heat Input Friction Welding Method (LHI method) that was developed by authors. A direct drive friction welding machine was used for joining. In case of LHI method, the relative speed between both specimens (weld faying surfaces) simultaneously and suddenly decreased to zero at the close of friction time. While the initial seizure and the joining began at the periphery portion by high friction pressure regardless of friction speed, they began at center portion by low friction pressure. The initial seizure was not coincided with the joining under various friction pressures as concerning to above results. However, the friction torque reached to initial peak torque when the welded interface was joined completely and upsetting of both base metals started. The friction welded joints made by LHI method had more than 100% joint efficiency in tensile strength and similar fatigue strength to that of the base metal. That is, there is no necessity for using forging (upsetting) stage. In this case, same results could be obtained in spite of friction pressure. As a conclusion, it was clarified that the welded joints made by LHI method, i.e., without the forging stage, have the same mechanical properties as those welded by the conventional friction welding process including forging stage, regardless of applied friction pressure. Furthermore, LHI method has a lot of advantages for industrial usage, e.g., the machining process of post-weld can be simplified, and so on.
1. Introduction Friction welding method is well known as one of the solid state joining methods. In particular, this method is very useful for the joining of dissimilar materials. Friction welding process, one of them, is widely used in the automobile industry, and it is applied to fabricate important parts such as drive shafts, engine valves, and so on. In the previous reports[1-3], the authors had clarified the joining mechanism during the friction stage of the friction welding process of low carbon steel joints. In addition, we had also suggested that the welded joints obtained 100% joint efficiency by using only the first stage (up
Kimura - 1 -
to the initial peak torque) of the friction stage[4,5]. That is, upsetting stage was not applied during friction welding. We named this method as Low Heat Input Friction Welding Method (henceforth it is referred to as LHI method). LHI method provides more advantages than the conventional method, i.e., less axial shortening (burn-off), less flash (burr or collar) and so on. However, those results were obtained under the following friction welding condition: the friction -1
speed of 27.5s and the friction pressure of 30MPa. Therefore, further investigation is necessary to clarify the joint property under another friction welding conditions such as high friction pressure for adaptability instance of LHI method. In addition, the evaluations of welded joints were carried out in static strength tests such as joint tensile test and bend test. In particular, it is important to evaluate the welded joints under the dynamic strength tests such as joint fatigue test and impact test. The fatigue strength of the friction welded joints that had 100% joint efficiency in joint tensile test, was less than that of base metal in some reports[6,7]. This study presents the characteristics of welded joints made by LHI method. In this report, the authors describe the relationship between the mechanical properties and friction time of low carbon steel joints. We also present the test results of joint tensile and fatigue properties of welded joints that were joined under various friction pressures.
2. Low Heat Input Friction Welding Method (LHI method) Figure 1 shows the schematic diagram of relationship between friction torque, relative speed and friction time by the conventional friction welding and LHI methods. When the joints were made by a direct (continuous) drive friction welding machine, i.e., the conventional method, the rotation of base metal are not instantly stopping. It is necessary of the braking time for rotation stop, because the rotation of base metal is not stopped completely at setting friction time. Therefore, the relative speed between both base metals was not instantly decreased to zero as
Rotating side
Fixing side
speed
Friction pressure
Rotating side
Fixing side
Friction pressure
torque
speed
torque Friction time
(a) Conventional method
Friction time
(b) LHI method
Fig.1 Schematic diagram of conventional method and LHI method. Kimura - 2 -
Table 1 Comparison of conventional method and LHI method. Welding type
Conventional method
LHI method
Axial shortening, mm
6.0
1.5
Joint tensile strength, MPa
450
444
Friction time, s
2.5
1.5
Heat input, kJ
8.9
3.6
shown in Fig.1(a), and the welded joints included the deformation during braking times. In contrast, LHI method was performed for the joining as follows. The base metals were joined by using an electromagnetic clutch in order to prevent braking deformation during rotation stop. When the clutch was released, the relative speed between both base metals simultaneously decreased to zero as shown in Fig.1(b). In this case, the friction pressure could be maintained (loaded), so that the effect of braking time on deformation for joints could be negligible. The main characteristics of welded joints made by using LHI method were as follows. Table 1 summarizes the axial shortening, joint tensile strength and heat input of the welded joints through the conventional and LHI methods. These joints were welded by the following friction -1
welding conditions: the friction speed of 27.5s and the friction pressure of 30MPa. Each heat input is calculated from the friction torque curve. The joint tensile strength obtained with LHI method was close to those obtained with the conventional method. That is, the welded joints by LHI method had similar joint strength to the tensile strength of the base metal. However, the welded joints by LHI method had less axial shortening and less flash comparing to those of the conventional method. In particular, it was clarified that the heat input of LHI method was much more less than that of the conventional method. The detail joint characteristics for LHI method had mentioned in the previous reports[4,5].
3. Experimental procedure The material used in this experiment was low carbon steel. The chemical compositions and Table 2 Chemical compositions and tensile properties of material used. (a) Chemical compositions (mass%) C
Si
0.15
0.29
Mn
P
S
Fe
0.41 0.012 0.019
420
315
75 Weld faying surface (a) Type A
120
Bal. 15
(b) Tensile properties T.S., MPa Y.S., MPa
20
El., %
R.A., %
37
62
75 Weld faying surface (b) Type B
120 (unit:mm)
Fig.2 Shapes and dimensions of the friction welding specimens. Kimura - 3 -
Welded interface
tensile properties of the material used are given in Table 2. The dimensions and configurations of the test specimens to be joined are shown in Fig.2. Type A specimens as shown in Fig.2(a) were
Removal of burr
used for the observation of joining phenomena and the joint tensile test, and Type B as shown in Fig.2(b) was used for the joint fatigue test. All of the weld faying (contacting) surfaces of the base
R3
R3
(unit:mm)
(a) Joint tensile test specimen Welded interface
metal were polished by surface grinding machine before joining. The surface roughness of the weld faying surface was ranged from 0.05 to 0.14µm as the
Removal of burr
center line average asperities. A direct drive friction welding machine was employed for the joining. R36
Both specimens were welded by using
30
(unit:mm)
LHI method. During friction welding
(b) Joint fatigue test specimen
operations, the friction speed and the
Fig.3
friction
pressure
following
were
set
combinations:
to
(a)
the
R36
Preparation process of joint tensile test specimen and joint fatigue test specimen.
27.5 -1
revolutions per second (rps) and 30MPa, and (b) 27.5s and 90MPa. No forge pressure was applied. The friction torque during the friction stage was measured with a load-cell, and recorded with a personal computer through an A/D converter by sampling time of 0.05s. The axial shortening of all friction welded joints were measured. Figure 3 shows the preparation process of joint tensile test and joint fatigue test specimens. The joint tensile test was carried out for the welded joints that were removed flash, as shown in Fig.3(a). The joint fatigue specimens were machined to 12mm in diameter and 30mm in length, as shown in Fig.3(b). The fatigue test was carried out using a rotating bending fatigue testing machine.
4. Results and discussion 4.1 Joining phenomena Figure 4 shows the relationship between the friction time and the cross-sectional appearance of the welded interface region by friction pressure of 30MPa. In order to clarify the joining process, each cross-sectional appearance is shown corresponding to friction torque curve. At the friction time of 0.7s, the boundary interfaces of both base metals at central region (center axis)
Kimura - 4 -
were not clear, i.e., they were joined.
80
The heat affected zone and the joined
70
region
to
60
periphery (outer surface) at 1.1s. In
50
this time, the flush was generated on
40
the outer surface. The friction torque
30
rapidly increased from 0.7s, at which
20
time joining started at center portion
10
extended
from
center
of the welded interface. The friction torque
reached
to
approximately
initial peak torque when the welded
FT : 1.1s
R
Fig.4
80
between the friction time, friction
70
torque,
cross-sectional
60
appearance of the welded interface
50
region by 90MPa. At 0.04s, the
40
boundary interfaces of both base
30
metals at the peripheral region were
20
not clear, so that they were joined.
10
The friction torque rapidly increased
0
region of the welded interface. The welded interface was joined and upsetting of both base metals started
FT : 1.3s
F
R
WI
F
R : Rotating side F : Fixing side WI : Welded interface FT : Friction time
5mm
0
Figure 5 shows the relationship
when joining started at the peripheral
F
0
1.3s.
the
WI
FT : 0.7s
interface was joined completely at
and
WI
R
0.5
1.0 1.5 2.0 Friction time, s
2.5
3.0
Relationship between friction time, friction torque, and cross-sectional appearance of welded interface region: friction pressure of 30MPa.
FT : 0.11s
R
WI
F
R
WI
F
FT : 0.3s R
WI
FT : 0.04s
0
Fig.5
0.5
F R : Rotating side F : Fixing side WI : Welded interface 5mm FT : Friction time
1.0 1.5 2.0 Friction time, s
2.5
3.0
Relationship between friction time, friction torque, and cross-sectional appearance of welded interface region: friction pressure of 90MPa.
at 0.11s. When the friction torque reached its approximately initial peak torque at 0.3s, the welded interface was joined completely. It is clarified from these results that the initial joining portion was different under both friction pressures. However, the friction torque reached to initial peak torque when the welded interface was joined completely and upsetting of both base metals started regardless of friction pressure.
4.2 Joint tensile strength Figure 6 shows the relationship between the friction time and the joint tensile strength plotted on the friction torque curve by the friction pressure of 30MPa. Figure 7 shows the appearances of the joint tensile test specimens after testing. At the friction time of 0.7s, the joint tensile strengths scattered (Fig.6), and fractures occurred at the welded interface (Fig.7). When friction
Kimura - 5 -
Fractured portion : Welded interface : Mixed mode : S15CK base metal 500
80
Friction time
400
70 S15CK base metal 60
0.7s
300
50
0.9s
Welded interface Rotating side Fixing side
40 200
30 20
100 0
1.1s
10
friction torque 0
0.5
1.0 1.5 2.0 Friction time, s
2.5
1.3s
0 1.5s
3.0
Fig.6 Relationship between friction time and joint tensile strength corresponding to friction torque curve: friction pressure of 30MPa.
2.0s 3.0s
80
500
70
5mm
Fig.7 Appearances of joint tensile test specimens after testing: friction pressure of 30MPa.
400
S15CK base metal 60
300
50
200
friction torque 40 30
times were 0.9s and 1.1s, the joint tensile
20
strength increased, and scatter decreased.
10
The joint tensile strengths were close to
0
that of the base metal, and fractures
Fractured portion : Welded interface : Mixed mode : S15CK base metal
100 0
0
0.5
1.0 1.5 2.0 Friction time, s
2.5
3.0
Fig.8 Relationship between friction time and joint tensile strength corresponding to friction torque curve: friction pressure of 90MPa.
propagated from welded interface to the base metal (mixed mode fracture). The welded joints had similar strength of base metals when friction time was 1.3s or longer. The fracture occurred in the base
metals (not at the welded interface) in all welded joints at 1.3 and 1.5s of which friction torque was close to initial peak torque. However, the several joints fractured at the welded interface including the base metal when friction time was 2.0s or longer (Fig.7). Figure 8 shows the relationship between the friction time and the joint tensile strength plotted on friction torque curve by the friction pressure of 90MPa. At 0.04s, the joint tensile strengths scattered, and fractures occurred at the welded interface. Then, the joint tensile strength increased, and the scatter decreased as increasing friction time. The welded joints had similar strength of base metals when friction time was 0.3s (close to initial peak torque) or longer. Almost welded joints fractured at the base metals. Kimura - 6 -
4.3 Joint fatigue strength
: Fractured at welded interface : Not fractured
10
8
relationship between the friction time
10
7
70
and the number of cycles to failure for
10
6
60
the
10
5
50
pressures of 30 and 90MPa, respectively.
10
4
40
In this case, the applied stress amplitude
10
3
was
stress
10
2
amplitude value is fatigue strength of
10
1
base metal. The number of cycles to
10
0
Figures
9
welded
and
joints
206MPa.
10
show
by
That
the
is,
the
friction
this
curve. We determined the safety fatigue
Stress amplitude 30 :206MPa 20 10
friction torque 0
failure was plotted on the friction torque Fig.9
80
0.5
1.0 1.5 2.0 Friction time, s
2.5
0
3.0
Relationship between friction time and number of cycles to failure by stress amplitude of 206MPa corresponding to friction torque curve: friction pressure of 30MPa.
life for welded joints as approximately 7
10 cycles to failure. When the friction pressure was 30MPa (Fig.9), the number 7
of cycles was less than 10 cycles at 0.7 and 0.9s, and those joints fractured from welded
interface.
However,
it was
7
10 cycles or over when the friction time was 1.1 through 2.0s. It decreased and the
joints
fractured
from
welded
interface
when
the
friction
exceeded
2.5s.
When
the
pressure
was
90MPa
time
friction
(Fig.10),
the
number of cycles to failure was not 7
10 cycles
at
0.04s.
It
reached
to
: Fractured at welded interface : Not fractured
10
8
10
7
70
10
6
60
10
5
50
10
4
10
3
friction torque 40 30
10
2
20
10
1
10
0
Stress amplitude 10 :206MPa 0 1.0 1.5 2.0 2.5 3.0 Friction time, s
0
0.5
7
10 cycles or over when the friction time was 0.08s or longer. Figures 11
and
12 show the
relationship between the friction time and the number of cycles to failure for
Fig.10
80
Relationship between friction time and number of cycles to failure by stress amplitude of 206MPa corresponding to friction torque curve: friction pressure of 90MPa.
welded joints with friction pressures of 30 and 90MPa, respectively. In this case, the applied stress amplitude was 230MPa. This stress amplitude value is higher than fatigue strength of base metal. When the friction pressure was 7
30MPa (Fig.11), the number of cycles to failure were more than 10 cycles at 1.3s, of which 7
torque was close to initial peak torque. It was less than 10 cycles and the joints fractured from the welded interface for other friction times. When the friction pressure was 90MPa (Fig.12), the
Kimura - 7 -
: Fractured at welded interface : Not fractured
10
8
10
7
70
10
6
60
10
5
50
10
4
40
10
3
10
2
10
1
10
0
(a) 30MPa-3.0s
80
Stress amplitude 30 :230MPa 20 10
friction torque 0
0.5
1.0 1.5 2.0 Friction time, s
2.5
0
3.0
(b) 90MPa-1.5s
Fig.11 Relationship between friction time and number of cycles to failure by stress amplitude of 230MPa corresponding to friction torque curve: friction pressure of 30MPa. : Fractured at welded interface : Not fractured
10
8
10
7
70
10
6
60
10
5
50
10
4
10
3
friction torque 40 30
10
2
20
10
1
10
0
Stress amplitude 10 :230MPa 0 1.0 1.5 2.0 2.5 3.0 Friction time, s
0
0.5
80
Fig.12 Relationship between friction time and number of cycles to failure by stress amplitude of 230MPa corresponding to friction torque curve: friction pressure of 90MPa.
2mm Fig.13
Fractured surfaces after rotating bending fatigue testing by stress amplitude of 230MPa.
number of cycles to failure were also 7
10 cycles at 0.3s, of which torque was close to initial peak torque. Figure 13 shows the examples of the fractured surfaces of the fatigue specimens welded with 30MPa-3.0s and 90MPa-1.5s, respectively. The welded joints had axial shortening
of
approximately
5mm
for
30MPa-3.0s joint, and approximately 7mm for 90MPa-1.5s one. That is, these joints had sufficient quantity of flash. However, the fracture started from the peripheral region of welded interface indicated by the arrows as shown in Fig.13. Since the welded joints had not-joining region at the peripheral region at welded interface of which color is pale black. The not-joining region at the peripheral region was produced during friction process after initial peak torque (steady state). The occurrence of not-joining region was due to repeated the cycle of joining and Kimura - 8 -
separating at the welded interface during steady state in friction process. The joining and the separating at the welded interface were clarified by Hasui et al.[8], and we also observed not-joining region at the cross-sectional appearances[5]. The not-joining region could be equated with crack, so that the crack caused failure of the welded joints during fatigue test. That is, the forge pressure is practically applied to joints during forging (upsetting) stage in order to progress joining at the peripheral region of the welded interface. However, the welded interface could be joined completely when the friction torque reached to initial peak torque by LHI method. Therefore, it was clarified that the forge pressure is not necessary for friction welding by LHI method theoretically. According to these results, it was clarified that the welded joints with LHI method had similar tensile strength to that of base metal, regardless of friction pressure. These joints also had similar fatigue life to that of base metal. Furthermore, the post-weld processing, e.g., machining, can be simplified because of less flash of the welded joints.
5. Conclusions This report described the properties of friction welded joint of low carbon steels by using Low Heat Input Friction Welding Method (LHI method) that was developed by authors. The properties included the joint tensile strength and the joint fatigue strength of the welded joints. The following conclusions are provided. (1) While initial seizure and joining began at periphery portion by high friction pressure regardless of friction speed, they began at center portion by low friction pressure. The initial seizure was not coincided with the joining under various friction pressures. However, the friction torque reached to initial peak torque when welded interface was joined completely and upsetting of both base metals started. (2) The friction welded joints that had more than 100% joint efficiency in tensile strength could be performed by using only the friction stage up to initial peak torque regardless of friction pressure. (3) The friction welded joints that was performed by using only the friction stage up to initial peak torque, regardless of the friction pressure, also had similar fatigue strength to that of base metal. In particular, the joints fractured from the welded interface when the friction time was set such as the time of the initial peak torque or longer. The number of cycles to failure 7
of those welded joints was less than 10 cycles. (4) It was clarified that the welded joints made by LHI method, i.e., without forging stage, had the same mechanical properties as those welded by the conventional friction welding process including forging stage. Furthermore, LHI method has a lot of advantages for industrial usage, e.g., machining process of post-weld can be simplified, and so on.
Kimura - 9 -
Acknowledgements The authors acknowledge financial support from the Hyogo Science and Technology Association, Japan. The authors also wish to thank the staff of the Machine and Workshop Engineering at Graduate School of Engineering of University of Hyogo (old, Himeji Institute of Technology). References [1] KIMURA M., MIOH H., KUSAKA M., SEO K. and FUJI A.: Quarterly Journal of the Japan Welding Society, 2002, 20, (3), 425-431 (in Japanese). [2] KIMURA M., KUSAKA M., SEO K. and FUJI A.: Quarterly Journal of the Japan Welding Society, 2002, 20, (3), 432-438 (in Japanese). [3] KIMURA M., OHTSUKA Y., AN G.B., KUSAKA M., SEO K. and FUJI A.: Quarterly Journal of the Japan Welding Society, 2003, 21, (4), 615-622 (in Japanese). [4] KIMURA M., KUSAKA M., SEO K. and FUJI A.: JSME International Journal(Series A), 2003, 46, (3), 384-390. [5] KIMURA M., KUSAKA M., SEO K. and FUJI A.: Quarterly Journal of the Japan Welding Society, 2002, 20, (4), 559-565 (in Japanese). [6] NAKAYAMA H., OHHIRA K., OKITA K., ARITOSHI M. and HASUI A.: Journal of the Society of Materials Science, Japan, 1984, 33, (367), 447-453(in Japanese). [7] HASEGAWA M.: Quarterly Journal of the Japan Welding Society, 1995, 13, (3), 463-469 (in Japanese). [8] HASUI A. and FUKUSHIMA S.: Journal of the Japan Welding Society, 1975, 44, (12), 1005-1010 (in Japanese).
Kimura - 10 -