a r t i c l e i n f o a b s t r a c t
Article history: A welding arc is considered a composite of electron flow and electrically neutral arc plasma. Their interac-
Received 12 April 2017 tion significantly affects the transmission and conversion process of the heat and momentum in welding
Accepted 16 May 2017 process. Studying the electrical property of arc plasma is significant for understanding this interaction,
Available online 24 May 2017
but directly testing such property in arc column is still a challenge. In this paper, a novel method based
on active probes was proposed/designed and used to test such a property in a welding arc plasma. The
Keywords:
probes were partially coated to electrically insulate except for their tips. During test, the probes moved
Arc
into the arc zone to detect the resistance between the tips. To minimize the damage to the probes, the
Arc plasma
Probe
movement of the probes was relatively quick such that the stay of the tips in the arc zone is minimal. In the
Detection meantime, electrical signals of the detection circuit and arc images were collected by a data acquisition
system and a high-speed camera system. Images suggested that the movement of the probes imposed
no noticeable effect on the behavior of the arc. On the other hand, the voltage to current ratio (V–I ratio)
between the probe tips reduces as the moving speed of the tips reduces. Analysis suggests that the detec-
tion circuit is not purely resistive. As such, when the moving speed was reduced sufficiently, the V–I ratio
approaches to a constant without changing with the moving speed. The effect from the dynamic property
of the circuit on the measurement of the V–I ratio was thus eliminated. As such, the moving speed can be
selected such that the dynamic behavior of the detection circuit becomes insignificant while the speed is
still sufficient to minimize possible damage to the tips due to high temperature of the arc. Furthermore,
the parameters of detection circuit were also discussed. As such, a novel method becomes available to
study the electrical behavior of the arc plasma in the future study.
© 2017 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jmapro.2017.05.013
1526-6125/© 2017 The Society of Manufacturing Engineers. Published by Elsevier Ltd. All rights reserved.
S. Chen et al. / Journal of Manufacturing Processes 27 (2017) 276–283 277
2. Experimental procedure probes. Since the distance of the two points being tested changes
as the probes rotate, the current in the detection circuit and voltage
2.1. Experimental principle between the two probes also change. Once the tips of the probes
rotate out of the arc column, the detection circuit is turned off. The
The major challenge for the probe method in measuring a weld- V–I (R0 ) is thus used here to represent the electrical conductivity
ing arc column is the ultra-high temperature from the arc that of the selected arc zone such that the Ohm’s law is applicable to
damages the probes quickly because the arc temperature is gen- the arc. Hence, the V–I (R0 ) of the arc between the two tips of the
erally considered to be in the range from 5000 K to 30000 K, or probes can be derived below:
even higher. To resolve this issue, the authors propose to rotate two
probes as shown in Fig. 1 such that they both pass through the arc U2 = U − U 1 (1)
zone rapidly to avoid damages. Both probes are coated to insure
U2 U − U1
U
such that only their tips will determine the two points between
R0 = = R= −1 R (2)
which the responded voltage is measured. By controlling/adjusting I U1 U1
the trajectories of the scans of the tips and their synchronization,
the measurements may be made for different pairs of points. where U is the constant voltage of the power source which drives
The schematic diagram of the detection circuit in the turn-on the detection circuit, U1 is the voltage across the sampling resis-
state is shown in Fig. 2. When the tips of two probes both enter into tor R, and I is the current flowing through the detection circuit. A
the arc column, the detection circuit is closed and a small current smallest V–I (R0 ) will be obtained during the rotation when the dis-
flows through the arc zone. The corresponding voltage to the small tance of the two probes is the smallest and can thus be used as a
current in the detection circuit is measured to analyze the electrical measurement of the electrical property between these two points
characteristics of the concerned arc zone between the tips of two in the arc zone.
278 S. Chen et al. / Journal of Manufacturing Processes 27 (2017) 276–283
2.2. Experimental setup and purpose sidered ideally resistive. The change in the distance between two
probes changes the resistance obviously. We expect that the resis-
The schematic of the experimental setup is shown in Fig. 3. tance changes relatively slowly in comparison with the effect of
It consists of a welding power source, welding torch, high-speed the dynamic property due to the inductance. To this end, the mov-
camera, detection circuit (probe A, probe B, sampling resistor R, ing speed of the probes must also be relatively slow. It is thus also
constant voltage power source U, voltage sensor B), voltage sen- interesting to see how slowly the probe must rotate in order to
sor A, current sensor, etc. The detection circuit is in turn-on state omit the effect from the inductance in the detection circuit. To this
(Fig. 3(a)) when the probes are in the arc zone. Otherwise, it is in end, experiments are needed. Furthermore, as a current carrying
turn-off state (Fig. 3(b)). plasma, the welding plasma arc has its own electrical field whose
The sweeping motion system is an important part of the exper- direction, from the tungsten electrode to the workpiece, is differ-
imental setup. Its front view is shown in Fig. 4. A stepping motor ent from that of the current from one probe tip to another. It thus
drives the main shaft to rotate the gear set. Probe A and probe B are may have special characteristics different from purely resistive as
rotated clockwise and anti-clockwise respectively. They simultane- the output of the detection circuit changes. For the detection cir-
ously pass through the arc zone and the variations of the sampling cuit, an effective way to change the output from the probe tips is to
voltage will be collected and recorded as they rotate. The detection change the sampling resistance or the constant voltage. The major
circuit and welding circuit were insulated from each other via poly- parameters for the designed experiments are given in Table 1.
tetrafluoroethylene probe supports. A water-cooled copper block In all experiments, the sampling voltage, the voltage and the
is used as workpiece to avoid the impact of metal vapor on the current of the welding circuit are synchronously collected and
measurements. recorded by an acquisition system based on NI data acquisition card
Though the movement of two probes significantly decreases the and LabVIEW software. The arc images are synchronously collected
heat damage, it may bring new issues. Moving probes may interfer- and recorded by means of a high-speed camera system. A plasma
ence with the welding arc being tested. Such possible interference welding torch with 3 mm orifice diameter, 4.8 mm tungsten diam-
tends to increase with the movement speed. It will be interest- eter and 4 mm tungsten setback aims to the workpiece vertically.
ing to see how fast we can move the probes without producing Pure argon (99.99%) is used for the shielding gas and plasma gas. The
noticeable interference with the arc being tested. Further, in the flow rate of shielding gas is 18.0 L/min, and the flow rate of plasma
previous analysis, the tested arc zone and detection circuit are con- gas is 3.0 L/min. The arc length remains at 10 mm and the welding
S. Chen et al. / Journal of Manufacturing Processes 27 (2017) 276–283 279
Table 1
Parameters in the designed experiments.
1# 47.6, 27.5, 22.4, 12.9, 10.5, 10.0, 6.0, 4.0, 3.8, 1.4 200 24
2# 10.0 100, 200, 300, 400, 500, 1000, 2000 24
3# 10.0 200 5, 15, 24, 29, 39
current is 100 A. The rotate radius of the probe is 40 mm. The two a conductor. The resistance of this conductor depends on the dis-
end-faces (tips) of probes are closest, their horizontal and vertical tance between two probes. When the probes enter the arc zone,
distance is 4 mm. The vertical distance between the workpiece and this conductor is connected into the detection circuit but its resis-
the probes is 6 mm. The tungsten bars with diameter 2 mm are used tance, thus the current of the detection circuit as well as the voltage
as the probes. across the sampling resistor whose resistance is a fixed constant,
changes as the probes rotate. In particular, as the distance between
the two probes decreases during the rotation, the sampling voltage
3. Results and discussion
measured increases. The maximum sampling voltage, thus the min-
imum V–I ratio, occurs when the distance between the two probes
3.1. The effects of the probe rotation on the arc stability
is the closest. It also can be seen from Figs. 6(a) and 7(a) that there is
no obvious fluctuation appeared in both arc voltage and arc current
The influence exerted on the arc being studied by the probes is
during the movement of probes in arc column. To clarify further the
studied in this subsection. The sampling resistance was 200 and
stability of the arc as the probes pass through the arc zone, the areas
the constant voltage was 24 V. As the probes continually passed
in the rectangular of Figs. 6(a) and 7(a) are zoomed in Figs. 6(b) and
through the arc zone, the arc images were recoded and typical arc
7(b) respectively. It can be seen that there are also no obviously
images are shown in Fig. 5. It can be seen that there is no significant
abnormal fluctuations found from the waveforms of the arc volt-
change in arc morphology as the probes pass through the arc zone.
age while the sampling voltage changes. These indicate that the
The typical curves of the current and voltage of the welding circuit
electrical property of the arc column is not obviously affected by
and the sampling voltage under the 47.6 rad s−1 and 10.0 rad s−1
the moving probes when they pass through the arc zone. This effect
angular velocity are shown in Figs. 6 and 7 respectively. Figs. 6(b)
thus could be ignored. Each peak value of the sampling voltage is
and 7(b) show the partial enlarged view of the rectangular area in
considered as the measurement result from the selected positions
Figs. 6(a) and 7(a) respectively.
in arc column, and in the following analysis the sampling voltage
As can be seen from Figs. 6(a) and 7(a) that the sampling voltage
will be transformed to the V–I ratio for easily understanding and
demonstrates sharp peaks periodically each time when the probes
evaluating the response characteristic of the detection circuit.
pass through the arc zone. A welding arc is commonly considered
280 S. Chen et al. / Journal of Manufacturing Processes 27 (2017) 276–283
Fig. 6. The voltage and current waveform in a sampling period with 47.6 rad s−1 angular velocity.
Fig. 7. The voltage and current waveform in a sampling period with 10.0 rad s−1 angular velocity.
S. Chen et al. / Journal of Manufacturing Processes 27 (2017) 276–283 281
200 V/A) and no longer decrease as the probes angular velocity con-
tinues to decrease. As such, 10.5 rad s−1 appears to be the optimal
speed.
The change of the minimal V–I ratio with the angular speed
reflects an inductive property of the arc zone that opposes any
changes in current through it. When the probes enter the arc col-
umn, a voltage signal could be measured but it needs enough time
to eliminate inductive effect. Once the inductive effect still works
when the probes reach the shortest distance/selected position, the
V–I ratio we got is higher than it should be. A decrease in the probes
angular velocity can give the detection circuit more time to elim-
inate the inductive effect. As such, as the probes angular velocity
decreases, the minimal V–I ratio (during a rotation of the probes) is
decreased. When the probes angular velocity reaches 10.5 rad s−1
or lower, the inductive effect was fully eliminated before probes
reached the selected position, so the V–I ratio does not further
decrease as the probes angular velocity decreases. Hence the mea-
surement result of 10.5 rad s−1 probes angular velocity or lower
could be considered valid. In this study, 10.0 rad s−1 is selected as
the rotation speed.
Fig. 8. Minimal V–I ratio versus probes angular velocity. A relatively fast speed is needed to avoid damages to the insu-
lation coating. Fig. 9 shows the probes whose coatings have been
damaged by the high temperature in the arc zone. After the coat-
3.2. The effects of the probes angular velocity in the ing is damaged, the melt coating material pollute the probe tip and
measurements the area of the contact of the probes with the arc zone decreases.
The decreased contact area will decrease the conductivity and give
The influence exerted on the measurement result, i.e., the V–I the two probes a more share of constant voltage. As such, the volt-
ratio, by the movement of probes is studied in this subsection. The age measured from the sampling resistance will reduce. As shown
sampling resistance was 200 , the constant voltage was 24 V, and in Fig. 10, before and after the coating is damaged, the sampling
ten levels of angular velocity were experimented for the probes, voltage is decreased from 12 V to 4.7 V.
from 47.6 rad s−1 to 1.4 rad s−1 . Different from the fixed probes, the To avoid the burning loss of the coating, the maximum time that
life of the probes is extended by reducing the exposure time of the the probes could exist in the arc zone (that is, the smallest probes
probes in the arc zone as the probes pass through the arc zone in a angular velocity) is discussed. The absorbed heat of the insulating
rotating manner. However, the exposure time of the probes in the coating is mainly through convective heat transfer. Combining with
arc zone has an important effect in the measurements. An optimal the convective heat transfer formula and the circular motion for-
angular speed should be a value below which the measured mini- mula, the minimum angular velocity to prevent the burning loss of
mal V–I ratio no longer changes significantly as the speed reduces the coating can be obtained by Formula (3),
further. If the damage to the probes at this speed is acceptable, then
this speed can be used as the speed to measure the arc property. LhT1 S
= (3)
The change of the measured minimal V–I ratio (during a rota- RcmT2
tion) with the probes angular velocity is shown in Fig. 8. It can be
seen that the V–I ratio gradually decreases as the probes angular where is the minimum angular velocity of the probes (rad·s−1 ), L
velocity decreases. As the probes angular velocity reaches below is the distance of the probe swept in the arc zone (mm), h is the
10.5 rad s−1 , the V–I ratio remains a relatively stable value (about convective heat-transfer coefficient (W m−2 K−1 ) and the forced
Fig. 11. The V–I ratio under different sampling resistance value.
Fig. 10. The sampling voltage before and after the coating damage.
Based on the analysis above, with a fixed probes angular velocity, [3] Nunes AC, Bayless EO, Jones III CS, et al. Variable polarity plasma arc welding
the constant voltage from 15 V to 29 V and the sampling resistance on the area shuttle external tank. Weld J 1984;63(9):27–35.
[4] Chen SJ, Jiang F, Lu YS, Zhang YM. Separation of arc plasma and current in
from 100 to 500 should be used to investigate the electrical electrical arc. Weld J 2013;93(7):253–61.
conductivity distribution in arc column. [5] Chen SJ, Jiang F, Zhang JL, et al. Measurement and analysis of plasma arc com-
The results from this study also show a relatively complex ponents. J Manuf Sci Eng 2015;137(1), 011006-1-9.
[6] Zhang RY, Jiang F, Chen SJ. Comparison of energy acted on workpiece among
dynamic electrical characteristics of welding arc column during twin-body plasma arc welding, non-transferred plasma arc welding and plasma
moving probe experiments, different with the previously under- arc welding. J Manuf Process 2016;24:152–60.
standing which usually considered the arc zone purely resistive. [7] Chabrerie JP, Haidar J, Manenc L, et al. Experimental study of forces acting on arc
electrodes. In: Proceedings of the 14th Interna- tional Conference on Electric
While the details of the complex characteristics are beyond the
Contacts. 1988. p. 20–4. AQ6.
scope of this paper, the authors intend to study further in this [8] Burleigh TD, Eagar TW. Measurement of the force exerted by a welding arc.
direction. Metall Trans A 1983;14(6):1223–4.
[9] Jiang Y, Xu BS, Lu YH, et al. Experimental analysis on the variable polarity plasma
arc pressure. Chin J Mech Eng 2011;24(4):1–6.
4. Conclusions [10] Tao DB, Chen SJ, Bai SJ, et al. Combined measurement of arc current and arc force
based on the split anode method. Electric Weld Machine 2011;41(5):10–3, 25.
[11] Huang Y, Wang XX, Qu HY, et al. Effects of arc parameters on arc anode current
In this paper, a novel method is designed and built based on
density of coupling AA-TIG arc. Trans China Weld Inst 2014;35(2):5–9.
active probes, to measure the electrical conductivity of arc plasma [12] Drouet MG. Measurement of the Current Distribution at the Anode of an Arc
directly. The V–I ratio was obtained as the testing result and assess- Burning in a Gas and in a Vacuum. J Phys E: Sci Instrum 1987;20(6):494625–6.
ment index. The effect of probes and its movement on the welding [13] Liang GL, Yuan SQ. Study on the temperature measurement of AZ31B magne-
sium alloy in gas tungsten arc welding. Mater Lett 2008;62(15):2282–4.
arc, the effect of probes angular velocity and characteristics of [14] Alsabti T, Alshawaf A, Liu S. Flux-assisted gas tungsten arc and laser welding
detection circuit on testing result are investigated by carefully of titanium with cryolite-containing fluxes: arc spectroscopy and corrosion
designed experiments. resistance studies. Weld J 2014;93(10):379S–87S.
[15] Farmer AJD, Haddad GN. Local thermodynamic equilibrium in free-burning arcs
It is found that the effect of probes and its movement on arc in argon. Appl Phys Lett 1984;45(1):24–5.
plasma shape and the electrical property could be ignored. The bal- [16] Haddad GN, Farmer AJD. Temperature measurements in gas tungsten arcs.
ance between the testing time needed and the damage avoidance Weld J 1985;64(12):339–42.
[17] Zhang YM, Zhang SB, Liu YC. A plasma cloud charge sensor for pulse keyhole
time of probes in arc zone is the key factor during the experiment. process control. Meas Sci Technol 2001;12(8):1365–70.
The movement of probes makes the arc column to illustrate induc- [18] Shan P, Yi XL, Hu SS, et al. Detection of plasma sheath voltage in plasma arc
tive characteristic, which could be eliminated as the probe angular welding. Trans China Weld Inst 2003;24(2):19–21, 26.
[19] Zhang YM, Jiang M, Lu W. Double electrodes improve GMAW heat input control.
velocity decreases to a certain value.
Weld J 2004;83(11):39–41.
Using a probe method to test welding arc column will bring a [20] Li KH, Zhang YM. Metal transfer in double-electrode gas metal arc welding. J
boundary resistance to experiment result, and the boundary will Manuf Sci Eng Trans ASME 2007;129:991–9.
[21] Li KH, Chen SJ, Zhang YM. Double-electrode GMAW process and control. Weld
tend to be a constant as the output of detection circuit is large
J 2007;86(8):231–7.
enough. However, once the output of detection circuit is too large, [22] Chen SJ, Lu Y, Li XR, et al. Gas tungsten arc welding using an arcing wire. Weld
the property of testee arc zone will be changed. J 2012;91(10):261–7.
[23] Shao Y, Zhang YM. Pulsed laser enhanced GMAW. Weld J 2014;93(6):205-s–14-
s.
Acknowledgements [24] Shao Y, Wang ZZ, Zhang YM. Monitoring of liquid droplets in laser enhanced
GMAW. Int J Adv Manuf Technol 2011;57:203–14.
This work is supported financially by the National Natural Sci- [25] Chen SJ, Zhang L, Wang XP, et al. Stability of cross arc process–a preliminary
study. Weld J 2015;94(5):158–68.
ence Foundation of China (Grant No. 51375021), the Beijing Natural [26] Chen SJ, Zhang L, Wang XP, et al. Feasibility study of cross arc welding process.
Science Foundation (Grant No. 3172004), the National Science and AWS Professional Program and Poster Session, Chicago 2013.
Technology Major Project of China (Grant No. 2014ZX04001171), [27] S J Chen, L Zhang, N Huang, et al. Gas Tungsten Arc Welding with Cross AC
Arcing Twin Wires. US patent application no. US 9457420 B2, 2016.
and the China Scholarship Council (CSC).
References
[1] Kammerhuber Ch, Sommerfeld R. High deposition MAG welding: used for
welding bridges and structures. Weld Word 1996;38:227–343.
[2] Irving B. Why aren’t airplanes welded? Weld J 1997;76(1):31–41.