Materials Transactions, Vol. 47, No. 2 (2006) pp. 446 to 449
#2006 The Japan Institute of Metals
RAPID PUBLICATION
Effect of Cadmium Chloride Flux in Active Flux TIG Welding
of Magnesium Alloys
Liming Liu, Zhaodong Zhang, Gang Song and Yong Shen
State Key Laboratory of Material Surface Modification by Laser, Ion, and Beams, Dalian University of Technology,
Dalian 116024, P.R. China
The Cadmium Chloride flux increases the weld penetration evidently in the Alternating Current Tungsten inert gas (AC TIG) welding of
magnesium alloy. In the present study, in order to investigate the effect of the CdCl2 active flux on the weld shape and arc voltage, bead-on-plate
specimens are made on AZ31B magnesium alloy pre-placed with CdCl2 active flux by the AC TIG process. Weld pool cross-sections and the arc
voltage are analyzed under different welding parameters, welding speed, weld current and arc length. The results showed that compared to the
conventional AC-TIG, welding penetration and the weld depth/width with CdCl2 flux are both two times greater than that of without flux under
optimal parameters. The voltage decreases with decreasing of travel speeds and arc length decreasing. Besides, the phenomenon of arc trailing in
the EN period and arc contraction in the EP period were observed in AC TIG welding of magnesium alloy with CdCl2 flux. It found that the arc
voltage increases with the increases of welding current, more energies are supplied for welding, resulting in the increases of arc voltage and weld
penetration.
(Received August 31, 2005; Accepted December 14, 2005; Published February 15, 2006)
Keywords: magnesium alloy, flux-assisted gas tungsten arc welding, chloride flux, alternate current
1.
Introduction
A number of studies regarding active flux gas tungsten arc
welding (A-TIG) have been published. In the mid 1960s,
research1,2) showed weld penetration could be augmented as
much as three times when the base materials were coated
with fluxes. Such fluxes remain interesting today because
they allow full-penetration welding at greater rates while still
employing the inexpensive and clean gas tungsten arc as the
heat source. To date, however, fluxes have been developed
only for joining titanium alloys2–6) and steels,7,8) and their
compositions are not published.
Several mechanisms for the augmented penetration observed in A-TIG welding have been given.5–10) The relative
importance of each mechanism is a function of the chemical
composition of the flux and bade metal, as well as the process
parameters. Although not entirely resolved, the mechanism
of arc constriction that raises the current density and creates
welds with greater depth-to-width ratios has often been
invoked.5,8,9) Because the flux also chemically interacts with
the molten material, a surface-energy contribution that
originates a Marangoni flow has also been proposed.8,10)
In TIG welding of light metals, the alternate current (AC)
was used to cathodically disrupt the refractory surface oxide.
For Aluminum alloys, Sire11,12) developed a flux-bounded
coating (FBTIG technique), which gave very promising
results on weld penetrations. For magnesium alloys,
Marya13,14) investigated the A-TIG of magnesium alloys
with a direct current (DC) and a chloride flux. In this study,
the effects of welding speed, weld current and arc length, on
the weld shape and the weld arc voltage in AC current TIG
welding with cadmium chloride flux were investigated on an
AZ31B magnesium alloy substrate.
Table 1
Parameters
Electrode type
Value
AC, W–2%ThO2
Diameter of electrode
2.4 mm
Vertex angle of electrode
60
Shield Gas and flow rate
Arc length
Ar, 10 Lmin
1 mm
Welding current
80 A
Welding speed
300 mm/min
1
<0:03%Fe, <0:10%Cu, <0:005%Ni, <0:04%Ca and the
rest of Mg, were selected for the welding experiments and
machined into 100 50 5 mm rectangular plates as test
piece. AC-TIG bead-on-plate welds were made with an
automatic control system. During welding the arc voltage
was continuously monitored. The standard welding conditions are listed in Table 1. The images of the electric arc for
TIG welding both with and without activating were obtained
with a High Speed Camera and stored in a computer with a
frame grabber.
In the previous series of experiments,15) the cadmium
chloride (CdCl2 ) flux has the most pronounced effect on weld
penetration increase of magnesium alloy welding. In this
paper, the cadmium chloride flux was selected as flux
ingredient to investigate the influences on weld penetration
and arc voltage under different welding parameters. After
welding, specimens for the weld shape observation were
prepared and etched to reveal the bead shape and size. The
cross-sections of the weld bead were photographed using an
optical microscope.
3.
2.
Standard welding conditions.
Results and Discussion
Experimental
AZ31B magnesium alloy plates with the average composition of 3.10%Al, 0.65%Zn, >0:20%Mn, <0:10%Si,
3.1 Arc voltage and heat input
Generally, the welding arc can be separated into three
regions: cathode, anode and arc column. The arc voltage
Effect of Cadmium Chloride Flux in Active Flux TIG Welding of Magnesium Alloys
3.2 Effect of arc length on weld shape and arc voltage
The weld penetration (D), depth/width (D=W) ratio and
arc voltage are shown in Figs. 1 and 2, respectively, under
different arc length from 0.5 to 0.75 mm. The weld D and the
weld D=W decrease with the increasing arc length. However,
the arc voltage increases with the arc length. The results
Weld, D /mm
2.5
0.5
0.3
1.5
3.2
D (without flux)
D (with CdCl2 flux)
D/W (without flux)
D/W (with CdCl2 flux)
2.8
0.6
0.4
2.0
Effect of welding speed on weld shape and arc
voltage
The role of the welding speed in determining the weld
shape and arc voltage was studied by varying the welding
speed from 120 to 570 mm/min, with an interval of 60 mm/
min, as shown in Figs. 3 and 4. The D and the weld D=W
ratio decrease with the increasing welding speed. In the
Weld, D /mm
D (without flux)
D (with CdCl2 flux)
D/W (without flux)
D/W (with CdCl2 flux)
3.0
3.3
0.7
Weld D/W ratio
3.5
showed that the measured voltage values are plotted as a
function of the arc length, yielding a straight line. At a
constant welding current, the large arc length will directly
increase the arc voltage. The slope of this line is a measure of
the electric field strength in the column of the arc, whereas
extrapolation to zero arc length yields the sum of the anodic
fall voltage and the cathodic fall voltage. It appears that
CdCl2 addition gives rise to an increase of the field strength
in the arc column and also to an increase of the sum of the
voltages. Therefore, the overall heat supply from the welding
power system will increase when the arc length becomes
large. However, the arc efficiency will reduce when the arc
length increases.21) Tsai reported that a large arc length will
broaden the heat distribution of the arc on the weld pool
surface significantly,22) which will enlarge the anode size and
lower the heat density on the pool. Therefore, the weld D and
the weld D=W ratio should decrease with an increase in the
arc length.
2.4
0.50 0.75 1.00 1.25 1.50 1.75
0.3
1.2
0.1
0.2
0.1
100
200 300 400 500 600
Welding speed, V /mm min-1
Arc length, l /mm
Fig. 1 Effect of arc length on weld D, weld D=W ratio with and without
CdCl2 flux.
0.6
0.4
1.6
0.8
0.5
0.7
0.5
2.0
0.2
1.0
0.8
Weld D/W ratio
includes the three sections: cathodic fall voltage, arc column
fall voltage and anodic fall voltage.16–18) A conclusion that
the arc voltage is higher with flux than without flux has been
given up by many researchers.19,20) Increases of voltage first
suggest weld pool formation only occurred after the flux layer
had been broken up and possibly removed from the arc
column. This process of displacing the flux, either by
vaporization or fluid flow, is energy consuming. The
increases of voltage, which can be interpreted as a strengthening of the electrical fields, indicate greater energies.
For the ACTIG welding, it includes EN (negative electrode
period) cycle and EP (positive electrode period) cycle. When
the base plate becomes negative (EP cycle), the refractory
surface oxide is cathodically disrupt, and at the same time,
the arc wanders and becomes long due to the low thermal
emissivity of magnesium. So the arc voltage of EP cycle is
higher than that of the EN cycle. In this paper, the arc voltage
of EN period is named as UEN with a positive number, and
then the arc voltage of EP period is named as UEP with a
negative number due to the contrary polarity.
447
Fig. 3 Effect of welding speed on weld D, weld D=W ratio with and
without CdCl2 flux.
UEN
UEN
40
40
30
D (without flux)
D (with CdCl2 flux)
20
10
0
l /mm
0.50 0.75 1.00 1.25 1.50 1.75
-10
-20
-30
-40
Arc voltage, U /V
Arc voltage, U /V
30
D (without flux)
D (with CdCl2 flux)
20
10
0
100
200
300
V /mm min-1
400 500 600
-10
-20
-30
-40
UEP
UEP
Fig. 2
Effect of arc length on arc voltage with and without CdCl2 flux.
Fig. 4 Effect of welding speed on weld D, weld D=W ratio with and
without CdCl2 flux.
L. Liu, Z. Zhang, G. Song and Y. Shen
Front view
Side view
5
Weld, D /mm
Without flux
4
D (without flux)
D (with CdCl2 flux)
D/W (without flux)
D/W (with CdCl2 flux)
1.2
1.0
0.8
3
0.6
2
0.4
With CdCl2 flux
1
0
10mm
EN
EP
EN
EP
Weld D/W ratio
448
0.2
60
70 80 90 100 110
Welding current, I /A
0.0
Fig. 7 Effect of welding current on weld D, weld D=W ratio with and
without CdCl2 flux.
Fig. 5 Arc image of ACTIG welding of Mg alloy with and without CdCl2
flux.
UEN
Arc voltage, U /V
Tungsten electrode
(a)
Molten pool
40
32
24
16
8
0
-8
-16
-24
-32
-40
D (without flux)
D (with CdCl2 flux)
I /A
60
70
80
90
100
110
UEP
Tungsten electrode
Fig. 8 Effect of welding current on weld D, weld D=W ratio with and
without CdCl2 flux.
(b)
Molten pool
Fig. 6
Effect of CdCl2 on arc shape.
effective arc length and, hence, in a large voltage drop over
the arc column [Fig. 6(b)]. The increasing of arc voltage with
CdCl2 flux is caused by the increase of arc trailing with
increasing welding speed.
3.4
conventional TIG welding, the arc voltage increase a little
with the increasing welding speed. However, the arc voltage
increases much with increasing travel speed when the
cadmium chloride flux is used.
This result can be understood by taking two different
phenomena into account: arc trailing and arc contraction.
Figure 5 shows the front view and side view arc images with
and without CdCl2 flux under the standard welding conditions (Table 1). The front views for EN period suggest that
the arc was dispersed on each side when cadmium chloride
was present. The front and side views for EP period suggest
that the arc was constricted. Arc contraction causes the
welding arc to shift to a new equilibrium, which affects both
the arc column and the anode and cathode fall areas
[Fig. 6(a)]. The side views for EN period with the CdCl2
flux shows that the arc was trailing, which implied that a
fraction of the total current possibly traveled a longer
distance than if it would have traveled straight in the
extension of the tungsten cathode as the welding arc in EN
period without flux. Arc trailing results in an increased
Effect of welding current on weld shape and arc
voltage
A series of experiments were carried out varying welding
currents from 60 to 110 A, while the other parameters were
kept at their standard values (Table 1). The results of these
experiments are shown in Fig. 7 and clearly demonstrate that,
with increasing welding current, the weld D, the weld D=W
and arc voltage all increasing. Figure 8 also shows that
increases in voltage correspond to increases in current as
expected from Ohm’s law. The large welding current will
also increase the electro-magnetic force, which strengthens
the downward body convection in the welding pool and
increase the weld D=W ratio.
Cadmium has a high first ionization potential (8.99 eV)
than magnesium (7.5 eV).23) In the plasma physics, elements
of high first ionization potentials necessitate greater temperatures to release a first electron. Because electrical conductivity depends upon electron density, conductivity is
lower with elements of high first ionization potential. The
ohmic impedance of the arc, defined by the ratio of the
voltage to the current, would be greater with cadmium
Effect of Cadmium Chloride Flux in Active Flux TIG Welding of Magnesium Alloys
chloride flux than without flux, resulting in the increase of arc
voltage.
4.
Conclusions
This paper depicts the influence of welding parameters on
weld D, D=W and arc voltage during AC-TIG welding
magnesium alloy with CdCl2 . Compared to the conventional
AC-TIG, the weld D and weld D=W ratio with CdCl2 flux are
both two times greater than that of without flux under optimal
parameters. The results of experiments also demonstrate that
the voltage decreases with decreasing of travel speeds and arc
length. Besides, the phenomenon of arc trailing in the EN
period and arc contraction in the EP period were observed in
ACTIG welding of magnesium alloy with CdCl2 flux. Arc
contraction increases the weld penetration and arc trailing
increases the arc voltage. It found that the arc voltage
increases with the increases of welding current due to the
ohmic impedance of the arc increasing with the elements of
high first ionization potentials in the arc. So more energies
are supplied for welding, resulting in the increases of arc
voltage and weld penetration.
Acknowledgements
The authors gratefully acknowledge the sponsorship from
Supported by Program for New Century Excellent Talents in
University under project NCET-04-0271 and the Excellent
Young Teachers Program of MOE, P.R.C.
449
REFERENCES
1) S. M. Gurevich and V. N. Zamkov: Avtom. Svarka. 12 (1966) 13–16.
2) S. M. Gurevich, V. N. Zamkov and N. A. Kushmienko: Avtom. Svarka.
9 (1965) 1–4.
3) A. G. Simonik: Svar. Proiz. 3 (1974) 52–53.
4) T. Paskell, C. Lundin and H. Castner: Weld. J. 76 (1997) 57s–62s.
5) B. E. Paton, V. N. Zamkov and V. P. Prilutsky: Paton Weld. J. 1 (2000)
5–11.
6) Z. Sun and D. Pan: Sci. & Tech. of Weld. & Join. 9 (2004) 337–344.
7) W. Lucas and D. Howse: Weld. Met. Fabr. 64 (1996) 11–17.
8) C. R. Heiple and J. R. Roper: Weld. J. 61 (1982) 97s–77s.
9) H. C. Ludwig: Weld. J. 47 (1968) 234s–240s.
10) M. Tanaka, T. Shimizu, H. Terasaki, M. Ushio, F. Koshi-ishi and C. L.
Yang: Sci. & Tech. of Weld. & Join. 5 (2000) 397–402.
11) S. Sire and S. Marya: Proc. of the 7th Int. Symp. JWS (2001) pp. 113–
118.
12) S. Marya and S. Sire: Proc. of the 7th Int. Symp. JWS (2001) pp. 107–
112.
13) M. Marya and G. R. Edwards: Weld. J. 81 (2002) 291s–298s.
14) M. Marya: Welding in the world 46 (2002) 7–21.
15) Z. D. Zhang, L. M. Liu, Y. Shen and L. Wang: Chinese Journal of
Nonferrous Metals 15 (2005) 912–916.
16) C. E. Jackson: Weld. J. 39 (1970) 129s–146s.
17) C. E. Jackson: Weld. J. 39 (1970) 177s–190s.
18) C. E. Jackson: Weld. J. 39 (1970) 225s–230s.
19) Y. V. Kazakov and K. B. Koryagin: Weld. Int. 16 (2002) 886–889.
20) W. Middel and G. D. Ouden: Sci. & Tech. of Weld. & Join. 4 (1999)
335–339.
21) P. Burgardt and R. Heiple: Weld. J. 65 (1986) 150s–155s.
22) N. S. Tsai and T. W. Eagar: Metall Trans B 16 (1985) 841–846.
23) J. F. Lancaster: The Physics of welding UK (Pergamon Press; 1984).