WELDING RESEARCH
SUPPLEMENT TO THE WELDING JOURNAL, AUGUST 1989
Sponsored by the American Welding Society and the Welding Research Council
All papers published in the Welding Journal'sVJeM'mg Research Supplement undergo Peer Review before
publication for: 1) originality of the contribution: 2) technical value to the welding community; 3) prior
publication of the material being reviewed; 4) proper credit to others working in the same area; and 5)
justification of the conclusions, based on the work performed.
The names of the more than 170 individuals serving on the AWS Peer Review Panel are published
periodically. All are experts in specific technical areas, and all are volunteers in the program.
Measurement of Transient Temperature
Response during Resistance Spot Welding
High-speed cinematography and infrared emission monitoring
are used to evaluate heat transfer with various electrodes
BY E. W. KIM AND T. W. EAGAR
ABSTRACT. The transient temperature
response during resistance spot welding
was measured as a function of various
process
parameters.
The
variables
include the electrode face thickness,
coolant flow rate, welding schedule, electrode force and type of zinc coating.
Two different experimental methods
were used. The first one used an infrared
emission monitoring method which measures the surface temperature using a
Thermovision system. The other method
used high-speed cinematography of a
cross-sectioned welding setup. The
cross-sectioned surface was painted with
thermosensitive paint in order to measure the movement of the isotherm during the welding process.
It was found that the slowest step of
heat transfer during the axial direction in
resistance spot welding is convective
heat transfer to the coolant at the electrode/coolant interface. There exists a
E 14/. KIM and T. W. EAGAR are with the
Department of Materials Science and Engineering, Massachusetts Institute of Technology,
Cambridge, Mass.
Based on a paper presented at the 69th Annual
AWS Meeting, held April 17-22, 1988, in New
Orleans, La.
critical electrode face thickness for which
the minimum electrode temperature can
be obtained. The critical electrode thickness changes with weld parameters such
as welding current, welding time and the
heat transfer coefficient at the water
cooling surface. A discontinuity in temperature was proven to exist across the
electrode/workpiece interface. The heat
transfer characteristics across this interface significantly affect the nugget development mechanism in zinc-coated, lowcarbon steel. It was also found that galvannealed steel and bare steel induce
lower currents in comparison to galvanized steels. Nonetheless, the overall
temperature profiles are higher for gal-
KEY W O R D S
Temperature Response
Transient Temperature
Temperature Measure
Resistance Welding
Spot Welding
Nugget Initiation
Interfaces
Electrode Thickness
Heat Transfer
Calvannealed Steel
vannealed steel and bare steel due to a
lower heat transfer rate at the electrode/
workpiece interface. These t w o materials
are less sensitive to the electrode force
than is the galvanized steel. It was also
found that nugget formation initiated at
the faying interface with truncated-cone
electrodes and at the electrode interface
with dome-type electrodes.
Introduction
Since the development of resistance
spot welding, a large amount of work has
been devoted to understanding the
mechanism of nugget formation and
electrode wear. As a part of these
efforts, various thermal models have
been produced (Refs. 1-8). Nevertheless,
such modeling work alone is not useful
and must be supported by experiments
to verify the suitability of the model as a
prediction tool. It is felt that the measurement of temperature profiles developed
during the welding process is very important in this respect.
In this work, the transient temperature
change on the electrode surface was
measured using various electrode face
thicknesses and coolant flow rates. To
see the effect of the electrode force and
zinc coating composition on galvanized
steel, the temperature distribution during
WELDING RESEARCH SUPPLEMENT I 303-s
IR Camera
Image Rotator
Video Editor
Computer Analysis
Fig. 1 — Infrared monitoring system
welding was monitored in a one-dimensional experimental simulation of the process. The heat propagation into the electrode was also observed on the surface
of the cross-sectioned electrode and the
workpiece. These experimental observations were then related to the observed
welding behavior.
Experimental Procedures
Two different experimental methods
were used. The first one used an infrared
monitoring method that measures the
surface temperature using a Thermovision system (Inframetrics Model 600). The
other used high-speed cinematography
of the cross-sectioned welding system
(Refs. 9-12). The propagation of an isotherm can be traced using cinematography coupled with thermosensitive paints
on the flat surface of the sectioned electrodes. Further explanation of these procedures is given below.
Infrared Emission Monitoring
The Thermovision system monitors the
emission intensity from the surface of the
electrode and the workpiece in the light
wavelength range of 8 to 14 micrometers. The experimental apparatus is shown
in Fig. 1. The scanning speed of this
system is 4 kHz in the horizontal direction
and 60 Hz in the vertical direction. The
data are displayed in NTSC video format.
Data are recorded on videotape and can
be analyzed later using a computer inter-
face. The measurement resolution is 8
bits, which is equivalent to 256 steps for a
given temperature measuring range.
The most crucial factor in accurate
measurement of temperature in this
experiment is the emissivity calibration of
the emitting surface. As the temperature
increases, the surface condition of the
electrode and of the workpiece will
change; therefore, it is very important to
have a known emissivity throughout the
experiment. To achieve this, the surface
was painted with temperature-sensitive
lacquer, which remains solid to 1371 °C
(2500°F). The emissivity of this lacquer
was calibrated by comparing the Thermovision temperature measurement with
thermocouple readings on a statically
heated sheet held at various temperatures.
Two different experiments were performed using this technique. The first was
a measurement of the electrode surface
temperature. A series of welds was produced that simulated a robotic welder in
an automotive assembly line. Twenty
welds were made with 1-in. nugget spacing at a repetition rate of 45 welds per
minute followed by a period of 23 s with
no weids. This weld-no weld cycle was
repeated three times, making a total of
60 welds. The coupon consisted of t w o
strips, each 1 in. (25 mm) wide and 22 in.
(56 cm) long. The variables that were
evaluated included the electrode face
thickness and the coolant flow rate.
The second experiment
fig. 3 Cinema tography
setup on an edge
weld
involved a
Upper Electrode
/ "
-_
f
Workpiece
'""-^^SCcNii^tfHSSiP = . W
9& ad^' ^6 ^ M l^ ^
rA^L^^-^^
Nugget -*
X
- Lower Electrode
304-s | AUGUST 1989
Fig. 2 — One-dimensional simulation of spot
welding
one-dimensional simulation of the welding process. The setup is shown in Fig. 2.
The length of the slender solid cylindrical
electrode was 19 mm (0.75 in.) and the
diameter was 4.8 mm (0.18 in.). The
coupons were made by punching out
disks from sheet stock which were then
statically pressed at 500 Ib (227 kg) to
eliminate the shear lips. To keep the
temperature low enough during the weld
simulation so that the collapse of the disk
coupon could be avoided, the welding
current was reduced by inserting an electrically resistive material, such as Inconel®
or stainless foil, between the electrode
shank and the welding machine. The
increase of the electrical resistance of the
secondary loop was about tenfold, which
reduced the secondary current to
acceptable levels. This test was used to
determine development of the temperature field in the workpiece and in the
electrode. The variables studied in this
experiment included changes in the electrode force as well as the zinc coating of
the steel.
High-Speed Filming
The experimental setup for high-speed
cinematography is shown in Fig. 3. The
cross-sectioned electrode surface and
the workpiece edges were painted with
thermosensitive paint, which melted at
the relatively low temperature of 371 °C
(700°F). The propagation of the 371 °C
isotherm into the electrode can be measured by tracing the melt line of the paint.
The pictures were taken at the rate of
1200 pps (pictures per second). The
response time of the lacquer is claimed to
be a few milliseconds and to have an
accuracy of ± 1 % by the manufacturer
(Ref. 13).
Thermal RfiRlafemcfi at:
j RC : convective heat transfer
I
j
Results and Discussion
j RE : electrode
In general, conductive heat flow across
different materials is analogous to flow
through a series of thermal resistances. A
simple one-dimensional heat flow model
of resistance spot welding is shown in Fig.
4. The heat flow in the radial direction for
the case of spot welding of steel sheets is
considered to be very small and hence is
neglected, especially when the thickness
of the sheet is small (Ref. 14). If one of the
thermal resistances is very much greater
than the others, the greatest thermal
resistance will control the flow of heat in
the system.
Heat flow within the molten nugget is
influenced by magnetohydrodynamic
convection induced by the welding current. This convection is very strong,
which will increase the effective thermal
conductivity of the liquid. For simplicity, it
is assumed that the convection-enhanced
thermal conductivity of the liquid metal is
much larger than the thermal conductivity of the solid. The thermal resistance of
the workpiece, RW/ and of the electrode
material, RE, are the reciprocals of the
thermal conductivity of each material.
The thermal resistance at the electrode
interface, RE|, acts as a thermal barrier to
heat flow from the workpiece to the
electrode. Another thermal resistance is
included at the boundary layer of the
coolant, which is symbolized as Rci- The
convective heat transfer of the coolant
itself is assumed to be very large so that
the effective thermal resistance of the
coolant, Rc, can be neglected. In the
following sections, experimental results
will be discussed with regard to each of
these thermal resistances.
/
\
I
Fig. 4 — Thermal
resistances in
one-dimensional
heat flow model
443,5
421,0,
398.9
377,8
354,0
331,9
310,7
289,7
268,2
245.8
223.3200,5
,175,7
,148,5
103,1
WORK
w
Fig. 5—Two-dimensional temperature profile on the electrode surface
111 lllll Mill
iPSiwir
Electrode Temperature
One example of the two-dimensional
electrode surface temperature field as
measured with the Thermovision system
is shown in Fig. 5. The relatively even
temperature field in the horizontal direction is clearly seen, and is strong justification for the simplicity of the one-dimensional thermal model. A typical cascade
display of a high-speed line scan along
the centerline of a similar weld is shown
in Fig. 6. The temperature drop at the
center is due to the low temperature of
the workpiece edge, which is far from
the nugget-formation zone. In this experiment, the temperature of the workpiece
does not have any significance as the
purpose is to measure heat flow in the
electrodes. The vertical direction in this
figure represents the time axis, and the
PIECE
418,0
389,7
364,1
387,2 liiWfi';!-''
!78,5
•--''i,":J:!
SiBii'
5jr;i;;i!ii!!iiii|!!liii:li;':;:ifei^i!l!»ii!
— — thermal scan line
/
N
work piece
Fig. 6 - Cascade display of a high-speed thermal line scan
y
WELDING RESEARCH SUPPLEMENT I 305-s
;a.e
ELECTRODE : CLASS 2, A CAP
MATERIAL : EG 70/70
ELECTRODE FORCE : 720 LBS
WELD TIME : 12 CYCLE
unmachined
r— 15.8 mm
-my
-j, 12.8 mm
N>
o
o
3
3
3
3
1
2
3
4
588.0
_
'05
"
OD
u
cn
3
3
3
]
i~r
\
6.2 mm
\
\
CJ
458.8
—
:
:
:
:
2.8
4.7
6.6
8.5
mm
mm
mm
mm
(0.11")
(0.19")
(0.26 - )
(0.34 - )
UJ
/
/
/
CD
ry
3
3
~>
<
QL
488.8
t-
6.35 mm
LU
C_
358.8
UJ
machined
308.8
15.8 mm
—
12.8 mm
CO
CO
IS
^3
3
=1
3
f.
OD.
3
_*
258.8
OJ
3
3
cr,
\Vn
\
•
i
i ' 6.35 mm
/
I
I
8.8
3
'
I
I
|
i—i—r
10.8
NUMBER
CD
CD
t r o d e surface as a f u n c t i o n of the n u m b e r
of welds. The welds were made at
10.6 ± 0 . 1 kA with coolant flow rate
fixed at 0.7 gpm (2.6 L/min). These
experiments were conducted to investigate the effect of electrode face thickness and the evolution of electrode temperature in successive welding. It is seen
that the maximum temperature increases
during the first three to five welds and
then stabilizes. The temperature rise
toward the end of the twentieth weld is
due to the heat built up in the workpiece
as the welding progresses towards the
end of the coupon.
1 : 8.5 mm
2 : 6.6 mm
3 : 4.7 mm
4 : 2.8 mm
Fig. 7 —Electrode geometry
horizontal direction is the geometric position along the electrode axis. The internal
geometry of the electrodes was modified
by machining conventional RWMA Class
2, A cap electrode shapes to thinner
electrode faces. Four different face thicknesses were tested: 2.8, 4.7, 6.6 and 8.5
mm (0.11, 0.18, 0.26 and 0.33 in.). The
overall geometry of the electrode that
was used is shown in Fig. 7. Welding was
performed on 0.8 mm (0.03 in.) electrogalvanized sheet steel, which had 70
g / m 2 of zinc on both sides. For all cases,
the electrode force was 720 Ib (327 kg)
and the welding time was 12 cycles with
t w o different current settings at the same
tap.
As the face thickness is reduced, the
maximum temperature decreases but
then increases when the face thickness
becomes too thin. Thus, there exists a
critical face thickness which minimizes the
electrode surface temperature. In this
experiment for a weld of 12 cycles, the
minimum temperature rise occurs at a
face thickness of around 4.7 mm, while
the maximum temperature was decreased by about 60 °C (108°F).
Table 1 summarizes another set of
experiments wherein the maximum electrode face temperature was monitored
as a function of various coolant flow rates
and electrode face thicknesses. The
welding current ranged from 12.5 kA to
Effect of Electrode Face Thickness. Figure 8 shows the maximum temperature
observed experimentally on the elec-
Table 1 — Effect of Coolant Flow Rates and Electrode Face Thickness on the Maximum
3
Electrode Surface Temperature* '
Thickness
8.5
6.6
4.7
2.8
0.9 gpm
mm
mm
mm
mm
460
(421)
388
(487)
(a) Temperatures in Centigrade.
306-s | AUGUST 1989
OF
Fig. 8 — Change of maximum electrode surface temperature
3
3
Face Thickness
Number
Number
Number
Number
"i—r
5.0
Flow Rate
0.7 gpm
472
392
413
466
0.5 gpm
0.2 gpm
485
400
429
491
481
420
450
505
i—i—r
15.0
28.8
WELD
as a function of number of welds
12.7 kA. The plots of these data are
shown in Fig. 9, which depict the maximum electrode temperature change during the welding cycle. Twenty welds
were made in a series. The experimental
data were taken from the four welds of
these 20 welds for every other cycle. The
curves were generated by curve fitting
the discrete experimental data with a
correlation factor exceeding 0.98 for all
cases. The symbols in the graphs were
merely inserted to identify each line.
Compared with the data in Fig. 8, the
temperatures are generally higher due to
the increased welding current. It is clear
from Fig. 9 and Table 1 that the electrode
face thickness has a much greater effect
on the maximum electrode temperature
than the coolant flow rate. In Table 1, the
temperatures for the 6.6- and 2.8-mm
(0.26- and 0.11-in.) electrodes at the flow
rate of 0.9 gpm seem to be abnormally
high. Two possible reasons can be contemplated. The first is the high heat
generation rate caused by the contaminated electrode contact surface. The other reason may be the change in the
emissivity due to the uneven thickness of
the high-temperature lacquer or the smut
produced by the contaminated contact
surface during welding. Taking the abnormalities into account, it may be said that
the lowest temperatures occur in the
6.6-mm electrode for this high-current
experiment. The lowest maximum electrode temperature for the 0.7 gpm flow
rate occurred in the 6.6-mm-thick electrode. This can be compared with the
previous experiment where the lowest
temperature occurred in the 4.7-mmthick electrode with the same coolant
flow rate. By increasing the welding current from 10.6 to 12.6 kA, the optimum
electrode face thickness had changed
-
ELECTRODE FACE THICKNESS
8.5
500.0
453.0
FLOW RATE i
a 0.9 GPM
^ 0 . 7 GPM
O 0 5 GPM
©0.2 GPM
—
420. e
LU
OL
4S8
350.0
300.0
a
250.0
200.0
150.0
OC
t±J
—
-
a.
3S0.
a
300
a
2S0
a
150. 0
1
3
1
1 '1
12.0
18.0
FLOU RATE
a 0.9 GPM
A 0.7 GPM
© 0 5 GPM
© 0.2 GPM
a
200. 0
1
~
1
1 ' 1 '1
24.0
30.0
36.0
1
1
1
6.0
TIME CCYCLES OF 60 Hz5)
12.3
18.0
ELECTRODE FACE THICKNESS i 6.6 mm
ELECTRODE FACE THICKNESS
OL
QC
UJ
0-
500.0
FLOU RATE
a 0.9 GPM
A 0.7 GPM
© 0 5 GPM
0 0 2 GPM
400.0
1
|
1
1
30.
(c)
530 .0
UJ
|
24.0
TIME CCYCLES OF 60 Hz)
(a)
4S0.0
4 7 mm
409. e
UJ
OL
ZD
ZD
cn.
1x1
ELECTRODE FACE THICKNESS
see
450.0
2.8
—
-
a
A
©
O
400.0
350.0
UJ
OL
Z>
300.0
<
OL
UJ
CL
T.
UJ
2S0.0
200.0
IS0.0 -
mm
FLOU RATE
0.9 GPM
0.7 GPM
3.5 GPM
0.2 GPM
350.0
300.0
250.0
200.0
150.3
~i—'—i—i—i
6.0
12.0
'
18.0
r^
24.0
r
1
30.
1
6.0
TIME CCYCLES OF 60 Hz)
1
1 '1
12.0
18.0
1 ' 1 '1
24.0
33.
TIME CCYCLES OF 60 Hz)
(b)
(d)
Fig. 9-Change of maximum electrode surface temperature during welding. (The symbols are inserted to identify each line)
from 4.7 to 6.6 mm. The maximum
decrease in the electrode temperature is
about 80°C (144°F) for this electrode. For
a given flow rate, the effects of the
electrode face thickness can be explained
below.
The heat diffusion length in a solid
body can be estimated by calculating the
characteristic heat diffusion length. If the
temperature at the electrode interface is
assumed constant during each half cycle
of weld, this diffusion length, I, is equal to
2\/crt for a temperature rise of 16% of
the electrode interface temperature.
Here ot is the thermal diffusivity of the
electrode material, which is roughly 0.9
cm 2 /s (5.8 in. 2 /s), and t is the welding
time, which in this case was 0.2 s (12
cycles). Thus, the estimated characteristic
heat diffusion length during the weld
cycle is 8.5 mm. This means that the heat
generated at the first cycle of welding
time will diffuse a distance of about 8.5
mm from the electrode interface by the
end of the weld cycle. On the other
hand, the heat generated at the 12th
cycle will propagate only about 2.5 mm
before the end of the weld cycle. By
decreasing the electrode face thickness,
the water-cooled surface area can be
increased. If the heat flow at the coolant
interface is the rate-determining step, the
electrode temperature at the watercooled surface will increase, thus producing a greater temperature drop across
the coolant
boundary
layer. The
increased water-cooled surface area and
the larger temperature drop with thinner
electrode faces will help reduce the electrode temperature. However, if the electrode face thickness is too thin, heat will
build up near the electrode face due to
insufficient water cooling. This is pictorially explained in Fig. 10A. The maximum
electrode temperature will be determined by the competition of these t w o
factors, i.e., the heat diffusion length and
the water cooling at the electrode/coolant interface.
As can be seen in Fig. 8 and Table 1,
the 8.5- and 2.8-mm-thick electrodes
exhibit the highest temperatures. For the
8.5-mm-thick electrode, water cooling
has a very small effect on the electrode/
sheet interface temperature during the
time of the weld cycle, since most of the
heat cannot diffuse as far as the water in
the time allowed. This is illustrated in Fig.
9A. In this case, the water cooling merely
cools the electrode after the weld is
completed. There is a very small effect of
the water flow rate on the electrode face
temperature during welding, per se. On
the contrary, the 2.8-mm-thick electrode
experiences cooling by the water during
the welding cycle. However, since heat
transfer through the water, even in the
presence of strong convection, is less
than heat diffusion in the copper, heat
will build up near the electrode face. For
electrodes with thicknesses between
these two cases, maximum temperature
is lower due to optimization of heat
diffusion in the copper and of heat
extraction by the water cooling.
For the electrodes with these thicknesses, the effects of the intensity of heat
input, the electrode thickness, the heat
transfer coefficient at the cooling interface and the thermal conductivity can be
seen as follows. For a simple one-dimensional heat flux equilibrium through the
electrode and the cooling water,
k.(Tf-Tc)/L = h.(Tc-Tw) = Q
(1)
where, k = electrode thermal conductivity, L = electrode face thickness, h = heat
WELDING RESEARCH SUPPLEMENT | 307-s
(a)
FLOU RATE = B.7 GPM
ELECTRODE FACE THICKNESS
r\
U
450 0
w
400 0
LU
ry
3
h-
UJ
n
UJ
i—
350
a
300
a
250
a
230
a
-i
.0
F/g. 10 —Increased
cooling of a thinner
electrode
[
1
12.3
1
i
18.3
|
24.8
i
|
33.0
r
36.3
TIME CCYCLES OF 60 Hz)
(b)
transfer coefficient at the cooling interface, Q = heat flux, T, = temperature at
electrode/workpiece
interface, Tc =
temperature at the water cooling surface
and T w = coolant temperature. Rearranging these equations,
T f = (A + 1) • T c - A • Tw
T, = Tc + B
where,
A =
L •h
Equations 2 and 3 are plotted in Fig. 11
with axes Tf and Tc. The position of the
curves represented by these equations in
the T,-T c plane is determined by the
parameters A and B. Parameter A is the
slope of Equation 2 and is basically a Biot
number. For the hypothetical value of A2
and B2, the crossing point, a, of these t w o
curves determines the electrode face
temperature Tf. If the heat input Q is
increased, the parameter B increases
from B2 to Bi, and hence the electrode
(2)
(3)
LQ
Fig. 11Determination of
electrode
temperature from
electrode thickness,
heat input and heat
transfer coefficient
at the cooling
interface
A!
A2
T,
A3
/P
/
/tJ
b
A
jT
//
/
/
c
0
/
\ vr
¥a
/
a
/
//
/
B/
Bf
Tw
B<
Tw
308-s I AUGUST 1989
Tc
temperature will increase to the point b in
Fig. 11. However, if the electrode face
thickness is increased, the temperature
drops to the point c by the increase of A
from A2 to A-i, and then moves in the
direction d with additional increases in
the parameter B, i.e., the final position of
d may be either lower or higher than that
of the point b depending on the relative
positions of the t w o curves. This suggests
the possibility that it may be advantageous to use a thicker electrode for a
higher welding current. By the same argument, if the electrode face thickness
decreases, the electrode face temperature wili follow the path a-b-e-f. The final
position of f is also dependent on the
relative position of the t w o curves. The
thinner electrode may be beneficial or
not, depending on the relative values of
A and B. In this experiment, the lowest
temperature was observed in the 4.7-mm
electrode for low-current welding, while
lower temperatures were found in the
6.6-mm electrode for high-current welding.
The effect of changes in the electrode
face thickness, L, for a given welding
current is the same. The temperature will
follow the path a-g-d or a-h-f, depending
on whether the electrode thickness is
increased or decreased. An increase in
the heat transfer coefficient, h, always
decreases the electrode temperature
from a to g.
Effect of Coolant Flow Rate. As was
discussed in the previous section, the
increase in flow rate does not show any
significant reduction in the maximum
electrode temperature for the 8.5-mm
electrode. However, the time held above
the threshold is responsive to the flow
rate as can be seen in Fig. 9A. By increasing the flow rate, the time above the
threshold decreases due to the faster
subsequent water cooling. This subsequent cooling is more effective when the
electrode face thickness is thinner. This is
shown in Fig. 10B. The cooling temperature gradient is steeper for a thinner
electrode and the final temperature is
generally lower for such an electrode.
It is believed that the electrode wear is
related to both the maximum time duration at temperature and the magnitude of
the maximum temperature. From this
point of view, it is also important to
optimize the flow rate and the internal
geometry in terms of the coolant flow. It
is seen in Fig. 11 that the effect of
increased heat transfer coefficient is very
important in lowering the electrode temperature. Therefore, thermal optimization of the electrode design should
include both the electrode face thickness
and also the characteristics of convective
cooling behavior of the water. When the
electrode is thick enough, the factor controlling the maximum temperature of the
electrode depends on the heat diffusion
characteristics of the electrode body.
When the face thickness is less than the
longest diffusion distance, convective
heat transfer at the coolant interface will
control the maximum face temperature
during welding. In general, the ratecontrolling step for the electrode thermal
behavior or the slowest process of heat
transfer in resistance spot welding is convective heat transfer of the coolant at the
cooling interface.
Effect of the Coating and the Electrode
Force
Figure 12 shows a typical temperature
profile developed in the one-dimensional
simulation. The t w o vertical lines marked
A near the center show the location of
the electrode interfaces. Another set of
vertical lines marked B is 3 mm (0.12 in.)
from the interface where the electrode
temperature was monitored. The temperature was also measured at the workpiece center and the electrode interface.
The measurement was performed when
the highest temperature was reached at
the faying interface. As would be expected, the temperature always reached its
maximum value at the end of the weld
cycle. Specimens of EG (electrogalvanized steel, 70 g / m 2 , both sides), G60 (hot
dip galvanized steel), A40 (galvannealed
steel) and bare steel (zinc removed by
HCI etching from EG 70/70) were used.
All were 0.8-mm-thick, low-carbon steel.
Welding was done at 450, 550, 650 and
750 Ib (205, 250, 295 and 341 kg) with
the same current setting in the welding
machine. Therefore, differences in electrical properties of the workpieces and of
the contact interfaces will induce different welding currents resulting in different
temperature profiles. The results are
shown in Figs. 13 and 14.
For a given welding machine setting,
A40 and bare steel lead to lower currents
in comparison to G60 and EG. Nonetheless, the overall temperature profiles are
higher for A40 and bare steel. A material
with a high electrical contact resistance
will produce more heat and will produce
higher temperatures provided that the
welding current is fixed. But the situation
of this experiment is different. The
materials that are believed to have higher
electrical contact resistance induce lower
welding currents and exhibit higher temperatures in the workpieces. The higher
temperature may be due either to a
higher heat generation rate or to a lower
heat loss rate or to both. The heatgeneration rate is linearly dependent on
the electrical resistance and is quadratically dependent on the current. The ratio of
the squared current of the low-current
materials, A40 and bare steel, to the
high-current materials, G60 and EG, is
approximately 0.8. This means that the
overall electrical resistance of the A40
cylindrical electrode
disk coupon
Fig. 12-Temperature profile of a high-speed line scan during one-dimensional simulation of the
spot welding process
and bare steel must be about 25% higher
than that of the G60 and the EG in order
to produce the same temperature profile.
The published literature shows the ratio
of dynamic resistance of a bare steel or a
galvannealed steel to a galvanized steel is
roughly 1.3 to 1.5 (Refs. 15, 16). Thus, the
heat generation in bare steel or galvannealed steel is higher than that in galvanized steel by approximately 0 to 20%.
Considering the large temperature differences in the maximum coupon temperatures compared to the small differences
in the electrode temperatures as seen in
Fig. 13, it may be concluded that the
reason for the higher temperature of A40
and bare steel is more likely due to a
higher thermal contact resistance than to
a higher rate of heat generation. The
higher interface temperatures in these
materials are also believed to be due to
slower heat dissipation rather than to
greater heat generation.
Another observation made from these
tests is that the temperature profiles for
A40 and bare steel are less sensitive to
electrode force than G60 and EG. G60
and EG show sharp overall temperature
drops as the electrode force increases.
Generally speaking, the temperature
decreases as the electrode force
increases. However, the induced welding
current increases with electrode force as
shown in Fig. 14. This may be explained
by the decreasing electrical and thermal
contact resistances produced with the
increasing electrode force. The final temperature field will depend on the com-
bined effect of these t w o contact resistances. These are believed to be related
to the morphology of the contact surface
and the deformation characteristics of
the sheets. The free zinc on the surface
of G60 and EG softens and melts very
quickly creating a low electrical and thermal resistance, while the surface of bare
steel and the Fe-Zn compound of the
A40 resist severe deformation and maintain high electrical and thermal resistance
even at elevated temperatures.
As stated previously, the thermal contact resistance will affect the heat distribution in spot welding. The temperature
difference between the faying interface
and the electrode interface is much larger
for A40 and bare steel. The overall temperature in the sheet is also higher for
these materials. The low interface heat
transfer coefficient may have increased
the temperature rise within the coupon
even though the induced current level is
much lower than that of EG and G60. This
illustrates the importance of the thermal
contact resistance at the electrode interface in the nugget growth mechanism.
This observation may explain the reason
why spot welding of galvanized sheets
requires a higher current level compared
to the bare materials. Previously, the
formation of a zinc halo surrounding the
weld nugget has been an explanation for
the effectively larger nugget size and
consequently the higher current requirement for welding of galvanized materials
(Refs. 15, 17, 18). In addition to this halo
effect, the enhanced heat transfer char-
WELDING RESEARCH SUPPLEMENT 1309-s
• EG 7 0 / 7 0
G60
G80, 8.8 mm
E0 7 8 / 7 8 ,
Uj
OL
408.8
8.8
-
ZD
H
^
388.8
EG 70/70
G60
A40
BARE STEEL CZn
UJ
Q.
z:
_
B i MAX. TEMP. I N COUPON
X i INTERFACE TEMP.
i i ELECTRODE TEHP. C3 »m
-p - 1 —i
488.8
588.8
ero
i r - 1 r— r 608.8
700.8
T
830.0
400.8
~i—'—i—'—i
500,0
630.0
B
'
700.0
530,0
ETCHING
_
n
O
LU
OL
Z>
H
<
OL
LU
0,
X.
LU
tMAX. TEMP. IN COUPON
INTERFACE TEHP.
ELECTRODE TEMP. C3 mrm off?
~i—'—I—•—i—•
488.8
588.8
683 0
I
708.8
MAX. TEMP. IN COUPON
INTERFACE TEMP.
ELECTRODE TEHP. C3mm
•—r~
888.8
ELECTRODE FORCE CLBS)
~l—'—i—'
488.8
533.8
off)
D
I ' r-1 r
600.0
780 8
888.8
ELECTRODE FORCE CLBS)
Fig- 13-Effects of zinc coating and electrode force on the temperature development in
one •dimensional simulation welding
acteristics at the electrode interface of
the zinc coated steel is also seen to be
important. As the nugget size increases,
the heat loss to the electrode becomes
greater and will demand higher heat
input. Figure 15 shows the lobe curve for
these coated materials. The relative positions of the lobe curve qualitatively
matches the thermal behavior observed
in this experiment (Ref. 19).
Heat Generation and Propagation
Figure 16 is an example of the heat
propagation pattern observed during the
high-speed cinematography experiment.
As described in the experimental procedure section, the melting propagation
front of the thermosensitive paint
310-s | AUGUST 1989
' 1
600.0
' I
700 0
^~
60
Fig. 14 — Induced welding current
888.8
EO 78/70, 8.8 mm
Zn COATING REMOVED BY
S00.0
r~
• Bare steel
A40
' I
ELECTRODE FORCE CLBS)
ELECTRODE FORCE CLBS)
ELECTRODE FORCE CLBS)
I
400.0
MAX. TEMP. IN COUPON
INTERFACE TEHP.
ELECTRODE TEMP. C3 mm <>rr>
matches the 371 °C isothermal line. This
figure shows t w o different combinations
of weld time and weld current. The heat
generation and propagation pattern is
quite different for these t w o cases. The
heat propagation front in the electrode is
more convex when long time and low
current are used. In this case, the heat
propagation pattern is symmetric. Weld B
in Fig. 16, which used high currents and
short weld times, shows localized heating
in the early stages of the process. Due to
a larger change in the current amplitude
for high-current welding, the fluctuation
in the temperature is also greater and is
more localized. This is the usual case for
higher current welding, i.e., there is less
symmetry and much more localization of
the heating pattern. In both cases, the
temperature builds up in the workpiece
first and then propagates into the electrode. This also confirms that there is a
significant thermal discontinuity at the
electrode interface.
Table 2 shows the effect of the different zinc coatings on the heat propagation
and generation pattern. The times shown
in the first t w o columns indicate the
starting frame of the high-speed movie at
which a phase change was observed in
the thermosensitive paint. The third column indicates the time at which a visible
red glow in the workpiece started. The
time was measured from the onset of the
welding current. The distance between
the isotherms indicates the distance
between the 371 °C isotherms in the
upper electrode and the lower electrode
after the completion of the weld cycle.
These were measured at the center of
the cross-sectioned electrode on the projected image. Thus, the units in this column are only relative. The incipient phase
change of the thermosensitive paint
occurs in the workpiece. A40 shows an
early buildup of temperature and glow in
the workpiece as compared to G60 and
EG, but the temperature rise in the electrode is slower. This is manifested by the
shorter distance of heat diffusion into the
electrode in the last column of Table 2.
This phenomenon is thought to be related to the difference in the contact heat
transfer characteristics of these materials
as discussed in the previous section. This
result is another example of the importance of the thermal characteristics of the
electrode interface as a heat transfer
barrier in developing the weld nugget.
Another observation from the highspeed cinematography experiment is the
effect of the electrode outer geometry.
The nugget starts to melt at the faying
interface for truncated-cone electrodes,
but with the dome-type electrode, melting begins at the electrode interface. The
smaller contact area for heat and current
transfer with the dome-type electrodes
results in a concentrated heat generation
pattern at the electrode interface. This
may explain the poor wear behavior of
domed electrodes (Ref. 15).
Table 2—Nugget Melting in Relation to the Breakdown of Thermal Discontinuity
Initiation of
Paint Melting
on Steel
(ms)
Material
Conclusions
Based on these observations and discussion, one can conclude the following:
1) There exists a critical electrode face
thickness above which heat conduction
in the electrodes controls the maximum
electrode surface temperature, and
below which convective heat transfer at
the water coolant interface is rate determining.
2) For a given welding machine setting, A40 and bare steel induce lower
currents in comparison to G40 and EG
steel; nonetheless, the overall temperatures are higher for A40 and bare steel
and are less sensitive to the electrode
force.
3) A discontinuity in temperature
exists across the electrode interface. The
heat transfer characteristics across this
interface significantly affect the nugget
development mechanism in zinc-coated,
low-carbon steel.
4) Nugget formation initiates at the
faying interface with truncated-cone
electrodes and at the electrode interface
with dome-type electrodes.
5) The slowest step in heat transfer in
the axial direction in RSW is convective
transfer to the water at the electrode/
coolant interface.
A40
G60
EG 70/70
Initiation of
Paint Melting
on Electrode
(ms)
Nugget glow
(ms)
Current
(kA)
Maximum Distance
between
Isothermal
Lines
(Relative Units)
25.8
19.2
17.5
32.5
44.2
38.3
11.0
10.9
11.7
8
K)
9
- • :
14.2
13.3
z
LU
s
a.
o
o
<r
<
H
</>
ui
Q.
B A48 GALVANNEALED
^ G60XS HOT DIP
0 70/78
ELECTROGALVANIZED
Ul
S
oa.
>
UJ
a
x
o
a.
<
LJ
3
4
ui
co
—
"1
8000
1
18000
I—
12000
14000
Fig. 15 — Lobe curves of zinc-coated
2
oa.
-I
CURRENT, amps
materials
ui
>
Ul
(11 0.5 cycle
a
o
a.
<
~—
Ul
to
Ul
Acknowledgments
W e would like to express appreciation
to our sponsors for this project, namely:
Ford Motor Company, General Motors
Company and International Lead Zinc
Organization, Inc. Special thanks are
extended to M. H. Eagar for help in
preparing graphs.
ce
z
<2> 1.5 cycle
UJ
2
o.
O
_i
ui
>
X
otr
References
1. Greenwood, J. A. 1961. Temperature in
spot welding. British Welding Journal (6):316322.
2. Nied, H. A. 1984. The finite element
modeling of the resistance spot welding process. Welding Journal b3[4)A23-s to 132-s.
3. Houchens, A. F., Page, R. E., and Yang,
W . H. 1977. Numerical modeling of resistance
spot welding. Numerical Modeling of Manufacturing Process, ed. R. F. Jones, Jr., D. W .
Taylor, H. Armen, |. T. Hong, ASME, pp.
117-129.
4. Rice, W „ and Funk, E. |. 1967. Analytical
investigation of the temperature distribution
during resistance welding. Welding Journal
46(4):175-s to 186-s.
5. Gould, ). E. 1987. An experimentation of
nugget development during spot welding,
using both experimental and analytical technique. Welding Journal 66(1):1-s to 10-s.
6. Archer, G. R. 1960. Calculation for tern-
<
Ul
CO
Ul
ce
Ul
2
a.
O
(5) 2.5 cycle
>
ui
a
o
tr
<
ui
co
Fig. 16— Heat propagation
pattern on an edge weld
WELDING RESEARCH SUPPLEMENT!311-s
perature response in spot welds. Welding
Journal 39(8):327 to 330.
7. Bentry, K. P., Greenwood, J. A.,
McKnowlson, P., and Baker, R. G. 1963. Temperature distribution in spot welds. British
Welding Journal (12)«13-619.
8. Nishiguchi, K„ and Matsuyama, K. 1985.
Influence of current waveform on nugget
formation phenomena in spot welding of thin
sheets. IIW doc. No. 111-805-85.
9. Satoh, T., Katayama, J., and Abe, H.
1970. Temperature distribution and breakd o w n of oxide layer during resistance spot
welding using a two-dimensional model.
Report 1, Journal of Japan Welding Society
39(1):38-48.
10. Satoh, T., Katayama, J., and Abe, H.
1970. Temperature distribution and break-
down of oxide layer during resistance spot
welding using a two-dimensional model.
Report 2, Journal of Japan Welding Society
39(2):124-137.
11. Upthegrove, W . R., and Key, J. F. 1972.
A high-speed photographic analysis of spot
welding. We/ding Journal 51(5):233-s to
244-s.
12. Lane, C. T., Sorensen, C. D., Hunter, G.
B., Gedeon, S. A., and Eagar, T. W. 1987.
Cinematography of resistance spot welding of
galvanized steel sheet. Welding Journal 66(9):
260-s to 265-s.
13. Tempil Div., Big Three Industries, Inc.
14. Kim, E. W., and Eagar. T. W . 1988.
Parametric analysis of resistance spot welding
lobe curve. SAE paper 880287.
15. Gedeon, S. A. 1984. Resistance spot
welding of galvanized steel sheet. MS Thesis,
MIT.
16. Lee, A., and Nagel, G. L. 1988. Basic
phenomena in resistance spot -welding. SAE
880277.
17. Natale, T. V. 1986. A comparison of the
resistance spot weldability of hot-dip and electrogalvanized
sheet
steels. SAE paper
860435.
18. Howe, P., and Kelly, S. C, 1988. A
comparison of the resistance spot weldability
of bare, hot-dipped, galvannealed, and electrogalvanized DQSK sheet steels. SAE paper
880280.
19. Bowers, R. )., and Eagar, T. W. 1987.
Spot welding of galvanized steel. Part 2, An
Evaluation of Galvanized Steels, Progress
Report, Materials Processing Center, MIT.
WRC Bulletin 330
January 1988
This Bulletin contains two reports covering the properties of several constructional-steel weldments
prepared with different welding procedures.
The Fracture Behavior of A588 Grade A and A572 Grade 50 Weldments
By C. V. Robino, R. Varughese, A. W. Pense and R. C. Dias
An experimental study was conducted on ASTM A588 Grade A and ASTM A572 Grade 50 microalloyed
steels submerged arc welded with Linde 40B weld metal to determine the fracture properties of base
plates, weld metal and heat-affected zones. The effects of plate orientation, heat t r e a t m e n t , heat input,
and postweld heat t r e a t m e n t s on heat-affected zone toughness were included in the investigation.
Effects of Long-Time Postweld Heat Treatment on the Properties of Constructional-Steel Weldments
By P. J. Konkol
To aid steel users in the selection of steel grades and fabrication procedures for structures subject to
PWHT, seven representative carbon and high-strength low-alloy plate steels were welded by shielded
metal arc welding and by submerged arc welding. The weldments were PWHT for various times up to 100
h at 1 1 0 0 ° F ( 5 9 3 ° C ) and 1 2 0 0 ° F ( 6 4 9 ° C ) . The mechanical properties of the weldments were
determined by means of base-metal tension tests, transverse-weld tension tests, HAZ hardness tests,
and Charpy V-notch (CVN) impact tests of the base m e t a l , HAZ and weld metal.
Publication of these reports was sponsored by the Subcommittee on Thermal and Mechanical Effects
on Materials of the Welding Research Council. The price of WRC Bulletin 330 is $20.00 per copy, plus
$5.00 for postage and handling. Orders should be sent with payment to the Welding Research Council,
345 E. 47th St., Suite 1 3 0 1 . New York, NY 10017.
WRC Bulletin 333
May 1988
Bibliography on Fatigue of Weldments and Literature Review on Fatigue Crack Initiation from Weld
Discontinuities
By C. D. Lundin
The bibliography together with a review of the present state of assessment of the factors which affect
fatigue crack initiation make up this d o c u m e n t . The bibliography was compiled through the efforts of
many students at The University of Tennessee utilizing the previously available bibliographies and
c o m p u t e r searches.
Publication of this report was sponsored by the S u b c o m m i t t e e on Failure Modes in Pressure Vessel
Components of the Materials and Fabrication Division of the Pressure Vessel Research C o m m i t t e e of the
Welding Research Council. The price of WRC Bulletin 333 is $20.00 per copy, plus $5.00 for postage and
handling. Orders should be sent with payment to the Welding Research Council, Suite 1 3 0 1 , 345 E. 4 7 t h
St., New York, NY 10017.
312-s | AUGUST 1989