DESCRIPTION
AC TIG WELDING METHOD
TECHNICAL FIELD
The present invention relates to a method for AC TIG welding
in which arc welding is performed by alternating a negative-polarity
period and a positive-polarity period.
BACKGROUND ART
In terms of environmental issues, aluminum and magnesium
materials have been extensively used in recent years for buildings,
vehicles, etc. because of their lightweight and highly recyclable
natures. These materials are generally joined using AC TIG welding
apparatus. AC TIG welding apparatus perform arc welding by
alternating a negative-polarity period and a positive-polarity period
(see, for example, Patent Literature 1).
For example, when large-current AC welding is performed using
an output extension cable at large building construction sites, a high
surge voltage is caused by the switching to reverse the polarity. The
high surge voltage may damage a semiconductor device of a secondary
inverter used in AC TIG welding apparatus.
A conventional approach to protecting the semiconductor device
is to decrease the current (to, for example, 100A or less) before the
switching to reverse the polarity, thereby reducing the surge voltage.
However, operator errors or poor precision in working or tools
may bring the TIG electrode and the object to be welded into contact

with each other during welding (generally referred to as "electrode
short circuit" or "short circuit"). As a result, the rate of decrease in
the welding current is lower during a short circuit than during arcing.
This may cause polarity reversal during a short circuit without a
sufficient decrease in the current. Then, the high surge voltage
caused by the switching to reverse the polarity may damage the
semiconductor device of the secondary inverter used in TIG welding
apparatus.
The operation of a conventional AC TIG welding apparatus will
be described as follows with reference to Figs. 8 and 9. Fig. 8 is a
schematic configuration view of the conventional AC TIG welding
apparatus, and Fig. 9 shows the change in a welding current waveform
with time in the conventional AC TIG welding apparatus.
The operation of the AC TIG welding apparatus shown in Fig. 8
will be described with reference to the change in the welding current
waveform with time shown in Fig. 9. The following is a description of
a non-consumable electrode AC TIG welding apparatus in which
welding is performed by alternating a negative-polarity period and a
positive-polarity period.
In Fig. 8, AC TIG welding apparatus 101 includes welding
output unit 102, welding controller 103, current detector 104, setting
unit 107, and first time keeper 108. AC TIG welding apparatus 101 is
electrically connected to welding torch 110 having electrode 109, and to
base material 112, which is the object to be welded. Apparatus 101
supplies electric power to electrode 109 and base material 112 so as to
create arc 111 between them.
Fig. 9, which shows the change in the welding current waveform

with time, also shows the following periods: a positive-polarity period
TEN, a negative-polarity period TEP, a positive-polarity peak period
TP1, a positive-polarity base period TBI, a negative-polarity peak
period TP2, and a negative-polarity base period TB2. The waveform
shows the following currents: a positive-polarity peak current IENP, a
negative-polarity peak current IEPP, a positive-polarity base current
IENB, a negative-polarity base current IEPB, a welding current
before-polarityreversal (the current during a short circuit in the
positive-polarity period) IEN1, and a welding current
before-polarityreversal (the current during a short circuit in the
negative-polarity period) IEP1. Fig. 9 also shows a time point E1
when a short circuit occurs, and a time point E2 when an arc is
recreated.
In Fig. 8, welding output unit 102 includes unillustrated
primary and secondary inverters for alternating a positive-polarity
period and a negative-polarity period based on a control signal from
welding controller 103. Welding output unit 102 receives commercial
power (e.g., three-phase 200V) from outside of AC TIG welding
apparatus 101, and outputs a welding voltage and a welding current
suitable for welding.
The primary inverter is generally composed of unillustrated
metal-oxide semiconductor field effect transistors (MOSFETs) and
unillustrated insulated gate bipolar transistors (IGBTs), both of which
are driven by a pulse width modulation (PWM) operation or a phase
shift operation, an unillustrated primary rectifier diode, a smoothing
electrolytic capacitor, a transformer for power conversion, and other
components.

The secondary inverter, which is generally a half- or full-bridge
inverter including unillustrated insulated gate bipolar transistors
(IGBTs), switches the output polarity.
A positive polarity means that arc plasma electrons move in the
direction from electrode 109 to base material 112, and that electrode
109 is negative, and base material 112 is positive. A negative polarity,
on the other hand, means that arc plasma electrons move in the
direction from base material 112 to electrode 109, and that electrode
109 is positive, and base material 112 is negative.
Setting unit 107, which can be composed of a CPU, sets the
following values: the positive-polarity peak period TP1 (e.g., 9.5 msec),
the positive-polarity base period TBI (e.g., 0.5 msec), the
negative-polarity peak period TP2 (e.g., 3.81 msec), the
negative-polarity base period TB2 (e.g., 0.47 msec), the
positive-polarity peak current IENP (e.g., 400A), the negative-polarity
peak current IEPP (e.g., -400A), the positive-polarity base current
IENB (e.g., 100A), and the negative-polarity base current IEPB (e.g.,
-100A). Setting unit 107 then outputs these values to welding
controller 103. These values can be either set by the operator
entering each parameter, or set automatically by a table or a
mathematical formula based on a current (the RMS or mean value) or
frequency that is set separately from the above-mentioned values.
First time keeper 108, which can be composed of a CPU, counts
the time from the start of the positive-polarity period or the time from
the start of the negative-polarity period. Current detector 104, which
can be composed of a CT, detects a welding current.
Welding controller 103 outputs an output command signal to

welding output unit 102 based on the output of setting unit 107, the
elapsed time counted by first time keeper 108, and the welding current
value detected by current detector 104. The output command signal is
one of the following currents: a positive-polarity peak current during
the positive-polarity peak period, a positive-polarity base current
lower than the positive-polarity peak current during the
positive-polarity base period, a negative-polarity peak current during
the negative-polarity peak period, and a negative-polarity base current
lower (smaller in the absolute value) than the negative-polarity peak
current during the negative-polarity base period.
Welding output unit 102 receives the output command signal
from welding controller 103. In welding output unit 102, during the
positive-polarity period, the secondary inverter switches the output
polarity such that electrons move in the direction from electrode 109 to
base material 112. During the negative-polarity period, on the other
hand, the secondary inverter switches the output polarity such that
electrons move from base material 112 to electrode 109.
In welding output unit 102, the primary inverter outputs a
positive-polarity peak current (e.g., 400A) during the positive-polarity
peak period, and outputs a positive-polarity base current (e.g., 100A)
during the positive-polarity base period. The primary inverter also
outputs a negative-polarity peak current (e.g., -400A) during the
negative-polarity peak period, and outputs a negative-polarity base
current (e.g., -100A) during the negative-polarity base period.
The welding current and the welding voltage outputted by
welding output unit 102 are supplied to welding torch 110 connected to
AC TIG welding apparatus 101. As a result, arc 111 is created

between the tip of electrode 109 and base material 112, thereby
performing AC TIG welding. Electrode 109 is a TIG electrode made,
e.g., of tungsten, and base material 112 is the object to be welded made,
e.g., of aluminum.
The following is a description, with reference to Fig. 9, of the
conventional AC TIG welding, and the welding current waveform.
As shown in Fig. 9, before the time point E1 when a short circuit
occurs, when a transition is made from the positive-polarity peak
period TP1 to the positive-polarity base period TB1 in the
positive-polarity period TEN, the welding current decreases from the
positive-polarity peak current IENP (e.g., 400A) to the
positive-polarity base current IENB (e.g., 100A). When the
positive-polarity base period TB1 is ended, a transition is made to the
negative-polarity period TEP.
When a transition is made from the negative-polarity peak
period TP2 to the negative-polarity base period TB2 in the
negative-polarity period TEP, the welding current decreases from the
negative-polarity peak current IEPP (e.g., -400A) to the
negative-polarity base current IEPB (e.g., -100A). When the
negative-polarity base period TB2 is ended, a transition is made to the
positive-polarity period TEN.
Thus, the welding current decreases to the target command
value, namely the positive-polarity base current IENB or the
negative-polarity base current IEPB before polarity reversal. As a
result, the surge voltage caused by the switching of the secondary
inverter to reverse the polarity can be decreased (to, for example,
about 300V). This prevents damage of the semiconductor device of

the secondary inverter used in welding output unit 102.
The following is a description, with reference to Fig. 9, of the
welding current waveform when a short circuit occurs between
electrode 109 and base material 112 during the conventional AC TIG
welding.
In Fig. 9, assume that the welding current is commanded to
decrease from the positive-polarity peak current IENP (e.g., 400A) to
the positive-polarity base current IENB (e.g., 100A) when a transition
is made from the positive-polarity peak period to the positive-polarity
base period during the short circuit in the positive-polarity period TEN
after the time point E1 when a short circuit occurs. In this case,
however, the welding current decreases only to the welding current
before-polarity.reversal IENl (e.g., 300A), which is larger than the
positive-polarity base current IENB. As a result, a transition is made
to the negative-polarity period when the welding current is at the
welding current before-polarityreversal IENl (e.g., 300A). The
reason for this will be described later.
Assume that the welding current is commanded to decrease
from the negative-polarity peak current IEPP (e.g., -400A) to the
negative-polarity base current IEPB (e.g., -100A) when a transition is
made from the negative-polarity peak period to the negative-polarity
base period while electrode 109 and base material 112 are still
short-circuited in the negative-polarity period TEP. In this case,
however, the welding current decreases only to the welding current
before-polarity-reversal IEP1 (e.g., -300A), which is larger than the
negative-polarity base current IEPB. As a result, a transition is made
to the positive-polarity period when the welding current is at the

welding current before-polarity-reversal IEP1 (e.g., -300A). The
reason for this will be described as follows.
Assume that it is necessary to significantly decrease the
welding current from the negative-polarity peak current IEPP (e.g.,
-400A) to the negative-polarity base current IEPB (e.g., -100A). In
this case, the primary inverter of welding output unit 102 is stopped,
and welding output unit 102 outputs a voltage output of 0V. During
arcing, the welding current is significantly decreased due to an arc
resistance. During a short circuit between electrode 109 and base
material 112, on the other hand, the welding current is not
significantly decreased because the arc resistance becomes zero. As a
result, the welding current before polarity reversal cannot be
decreased to the target command value, namely, the positive-polarity
base current IENB or the negative-polarity base current IEPB.
As a result, while electrode 109 and base material 112 are
short-circuited, the polarity is reversed when the welding current is at
the welding current before-polarityreversal IENl or the welding
current before-polarityreversal IEP1, which is higher than the target
command value, namely, the positive-polarity base current IENB or
the negative-polarity base current IEPB. As a result, the switching of
the secondary inverter causes a high surge voltage (e.g., about 600V),
possibly damaging the semiconductor device of the secondary inverter.
If an extension cable (its length is, e.g., 40 m) is connected to a
cable used on the output side of AC TIG welding apparatus 1, the cable
has a large inductance. This further increases the surge voltage, and
the semiconductor device is more likely to be damaged.
According to the above-described conventional technique, when

a short circuit occurs between electrode 109 and base material 112
during a large-current AC TIG welding process, the welding current
cannot be significantly decreased. As a result, the welding current
before polarity reversal cannot be decreased to the target command
value, namely, the positive-polarity base current IENB or the
negative-polarity base current IEPB. As a result, the polarity is
reversed when the welding current is at the welding current
before-polarityreversal IEN1 or the welding current
before-polarityreversal IEP1, which is higher than the target
command value, namely, the positive-polarity base current IENB or
the negative-polarity base current IEPB.
As a result, the switching of the secondary inverter causes a
high surge voltage (e.g., about 600V), possibly damaging the
semiconductor device of the secondary inverter.
As described above, in the conventional AC TIG welding
apparatus, if a short circuit occurs between the electrode and the base
material while welding is performed with a large current (in the range
of 300A to 500A), the polarity is reversed without a sufficient decrease
in the welding current. As a result, the switching of the secondary
inverter causes a high surge voltage, possibly damaging the
semiconductor device of the secondary inverter.

SUMMARY OF THE INVENTION
The present invention provides a high-quality method for AC
TIG welding in which even if a short circuit occurs between the
electrode and the base material while welding is performed with a
large current, the semiconductor device of the inverter is prevented
from being damaged.
To solve the above-described problem, in the method of the
present invention for AC TIG welding, welding is performed by
alternating a positive-polarity period and a negative-polarity period.
This method includes: detecting the contact between a TIG electrode
and the object to be welded during welding,' and disabling a transition
from one polarity period to the other when the contact is detected.
According to the present invention, polarity reversal is disabled
during a short circuit. As a result, even if the TIG electrode and the
object to be welded come into contact with each other during welding,
no surge voltage is caused by the switching to reverse the polarity.
This achieves high-quality AC TIG welding which prevents the
semiconductor device from being damaged.
BRIEF DESCRIPTION OF DRAWINGS
Fig. 1 is a schematic configuration view of an AC TIG welding
apparatus used in a method for AC TIG welding according to a first
exemplary embodiment of the present invention.
Fig. 2 shows the change in a welding current waveform with
time in the method for AC TIG welding according to the first
exemplary embodiment.
Fig. 3 is a schematic configuration view of an AC TIG welding

apparatus used in a method for AC TIG welding according to a second
exemplary embodiment of the present invention.
Fig. 4 shows the change in a welding current waveform with
time in the method for AC TIG welding according to the second
exemplary embodiment.
Fig. 5 shows the change in another welding current waveform
with time in the method for AC TIG welding according to the second
exemplary embodiment.
Fig. 6 is a schematic configuration view of an AC TIG welding
apparatus used in a method for AC TIG welding according to a third
exemplary embodiment of the present invention.
Fig. 7 shows the change in a welding current waveform with
time in the method for AC TIG welding according to the third
exemplary embodiment.
Fig. 8 is a schematic configuration view of a conventional AC
TIG welding apparatus.
Fig. 9 shows the change in a welding current waveform with
time in the conventional AC TIG welding apparatus.
DESCRIPTION OF EMBODIMENTS
Embodiments of the present invention will be described as
follows with reference to the accompanied drawings. In the second
and third embodiments, like components are labeled with like
reference numerals with respect to the preceding embodiments, and
hence the description thereof may not be repeated.
FIRST EXEMPLARY EMBODIMENT

Fig. 1 is a schematic configuration view of an AC TIG welding
apparatus used in a method for AC TIG welding according to the
present first exemplary embodiment. Fig. 2 shows the change in a
welding current waveform with time in the method for AC TIG welding.
The operation of the AC TIG welding apparatus shown in Fig. 1 will be
described with reference to the change in the welding current
waveform with time shown in Fig. 2.
The following is a description of a non-consumable electrode AC
TIG welding apparatus in which welding is performed by alternating a
negative-polarity period and a positive-polarity period.
As shown in Fig. 1, AC TIG welding apparatus 1 includes
welding output unit 2, welding controller 3, current detector 4, voltage
detector 5, AS determination unit 6, setting unit 7, and first time
keeper 8. Welding output unit 2 outputs a welding output. Welding
controller 3 controls welding output unit 2. Current detector 4
detects a welding current. Voltage detector 5 detects a welding
voltage. AS determination unit 6 determines whether electrode 9 and
base material 12 are short-circuited or an arc is created therebetween,
based on the detection result of voltage detector 5. Setting unit 7 sets
welding conditions. First time keeper 8 counts time.
AC TIG welding apparatus 1 is electrically connected to welding
torch 10 having electrode 9 and base material 12, which is the object to
be welded. Apparatus 1 supplies electric power to electrode 9 and
base material 12 so as to create arc 11 between them.
Fig. 2, which shows the change in the current waveform with
time, also shows the following periods: a positive-polarity period TEN,
a negative-polarity period TEP, a positive-polarity peak period TP1, a

positive-polarity base period TB1, a negative-polarity peak period TP2,
a negative-polarity base period TB2, and a period TEN 1 indicating a
period before a short circuit occurs in the positive-polarity period.
The waveform shows the following currents: a positive-polarity
peak current IENP, a negative-polarity peak current IEPP, a
positive-polarity base current IENB, a negative-polarity base current
IEPB, a welding current (the current when the negative-polarity
period is ended during a short circuit) IEP2, an absolute value
IEPBABS of the negative-polarity base current, and an absolute value
IEPPABS of the negative-polarity peak current.
Fig. 2 also shows a time point E1 when a short circuit occurs, a
time point E2 when an arc is recreated, and a time point E6 when the
negative-polarity period is ended.
In Fig. 1, welding output unit 2 includes unillustrated primary
and secondary inverters for alternating a positive-polarity period and a
negative-polarity period based on the output of welding controller 3.
Welding output unit 2 receives commercial power (e.g., three-phase
200V) from outside of AC TIG welding apparatus 1, and outputs a
welding voltage and a welding current suitable for welding.
The primary inverter is generally composed of unillustrated
metal-oxide semiconductor field effect transistors (MOSFETs) and
unillustrated insulated gate bipolar transistors (IGBTs), both of which
are driven by a pulse width modulation (PWM) operation or a phase
shift operation, an unillustrated primary rectifier diode, a smoothing
electrolytic capacitor, a transformer for power conversion, and other
components.
The secondary inverter, which is generally a half- or full-bridge

inverter including unillustrated insulated gate bipolar transistors
(IGBTs), switches the output polarity.
A positive polarity means that arc plasma electrons move in the
direction from electrode 9 to base material 12, and that electrode 9 is
negative, and base material 12 is positive. A negative polarity, on the
other hand, means that arc plasma electrons move in the direction
from base material 12 to electrode 9, and that electrode 9 is positive,
and base material 12 is negative.
Setting unit 7, which can be composed of a CPU, sets the
following values: the positive-polarity peak period TP1 (e.g., 9.5 msec),
the positive-polarity base period TB1 (e.g., 0.5 msec), the
negative-polarity peak period TP2 (e.g., 3.81 msec), the
negative-polarity base period TB2 (e.g., 0.47 msec), the
positive-polarity peak current IENP (e.g., 400A), the negative-polarity
peak current IEPP (e.g., -400A), the positive-polarity base current
IENB (e.g., 100A), and the negative-polarity base current IEPB (e.g.,
-100A). Setting unit 7 then outputs these values to welding
controller 3. These values can be either set by the operator entering
each parameter, or set automatically by a table or a mathematical
formula based on a current (the RMS or mean value) or frequency that
is set separately from the above-mentioned values.
First time keeper 8, which can be composed of a CPU, counts the
time from the start of the positive-polarity period or the time from the
start of the negative-polarity period. Current detector 4, which can be
composed of a CT, detects a welding current.
Welding controller 3 outputs an output command signal to
welding output unit 2 based on the output of setting unit 7, the elapsed

time counted by first time keeper 8, and the welding current value
detected by current detector 4. The output command signal is one of
the following currents: a positive-polarity peak current during the
positive-polarity peak period, a positive-polarity base current lower
than the positive-polarity peak current during the positive-polarity
base period, a negative-polarity peak current during the
negative-polarity peak period, and a negative-polarity base current
lower (smaller in the absolute value) than the negative-polarity peak
current during the negative-polarity base period.
Welding output unit 2 receives the output command signal from
welding controller 3. In welding output unit 2, during the
positive-polarity period, the secondary inverter switches the output
polarity such that electrons move in the direction from electrode 9 to
base material 12. During the negative-polarity period, on the other
hand, the secondary inverter switches the output polarity such that
electrons move from base material 12 to electrode 9.
In welding output unit 2, the primary inverter outputs a
positive-polarity peak current (e.g., 400A) during the positive-polarity
peak period, and outputs a positive-polarity base current (e.g., 100A)
during the positive-polarity base period. The primary inverter also
outputs a negative-polarity peak current (e.g., -400A) during the
negative-polarity peak period, and outputs a negative-polarity base
current (e.g., -100A) during the negative-polarity base period.
The welding current and the welding voltage outputted by
welding output unit 2 are supplied to welding torch 10 connected to AC
TIG welding apparatus 1. As a result, arc 11 is created between the
tip of electrode 9 and base material 12, thereby performing AC TIG

welding. Electrode 9 is a TIG electrode made, e.g., of tungsten, and
base material 12 is the object to be welded made, e.g., of aluminum.
Voltage detector 5, which can be composed of a CT, detects a
welding voltage by measuring the voltage across the output terminal of
AC TIG welding apparatus 1.
AS determination unit 6, which can be composed of a CPU,
receives a voltage detection signal from voltage detector 5, and
calculates the absolute value of the voltage detection signal. Assume
that the absolute value of the voltage detection signal has reached (has
decreased to) a predetermined detection level (e.g., 10V) during arcing.
In this case, it is determined that electrode 9 and base material 12,
which is the object to be welded are in contact with each other
(generally referred to as "electrode short circuit" or "short circuit").
The AS signal indicates that a short circuit is present (low level).
Assume, on the other hand, that the absolute value of the voltage
detection signal has reached (has increased to) a predetermined
detection level (e.g., 15V) during a short circuit. In this case, it is
determined that the contact (the short circuit) between electrode 9 and
base material 12, which is the object to be welded, has opened and an
arc has occurred (generally referred to as arc recreation or during
arcing). The AS signal indicates that an arc is present (high level).
Welding controller 3, which controls welding output unit 2,
receives the AS signal from AS determination unit 6. The AS signal
indicates whether an arc is present or a short circuit is present.
When the AS signal indicates that a short-circuit is present (low level),
welding controller 3 controls welding output unit 2 to disable a
transition from one polarity period to the other even when the timing

comes to do so. When the short circuit continues to be present (low
level) as the AS signal, welding controller 3 continues to disable the
transition.
In the case where the AS signal indicates that an arc is present
(high level), and when the timing comes to make a transition from one
polarity period to the other, welding controller 3 controls welding
output unit 2 to allow the transition. In the case where an arc is
determined to be present while the transition is being disabled due to
the presence of a short circuit, welding controller 3 also controls
welding output unit 2 to allow a transition from one polarity period to
the other. This case will be described in detail later.
The following is a description, with reference to Fig. 2, of the
operation of AC TIG welding apparatus 1 to disable a transition from
one polarity period to the other when a short circuit occurs during
welding, and the welding current waveform.
In Fig. 2, before the time point E1 when a short circuit occurs,
when a transition is made from the positive-polarity peak period TP1
to the positive-polarity base period TB1 in the positive-polarity period
TEN, the welding current decreases from the positive-polarity peak
current IENP (e.g., 400A) to the positive-polarity base current IENB
(e.g., 100A). When the positive-polarity base period TB1 is ended, a
transition is made to the negative-polarity period TEP.
When a transition is made from the negative-polarity peak
period TP2 to the negative-polarity base period TB2 in the
negative-polarity period TEP, the absolute value of the welding current
decreases from the negative-polarity peak current IEPP (e.g., -400A)
to the negative-polarity base current IEPB (e.g., -100A). When the

negative-polarity base period TB2 is ended, a transition is made to the
positive-polarity period TEN.
In Fig. 2, assume that the welding current is commanded to
decrease from the positive-polarity peak current IENP (e.g., 400A) to
the positive-polarity base current IENB (e.g., 100A) when a transition
is made from the positive-polarity peak period TP1 to the
positive-polarity base period TB1 during the short circuit in the
positive-polarity period TEN after the time point E1 when a short
circuit occurs. In this case, however, the welding current decreases
only to a welding current before-polarity-reversal IEN2 (e.g., 300A),
which is larger than the positive-polarity base current IENB. As a
result, the positive-polarity period TEN is ended while the welding
current is at the welding current before-polarity-reversal IEN2. The
welding current cannot be decreased significantly when a short circuit
occurs between electrode 9 and base material 12 during a large-current
AC TIG welding process. The reason for this is the same as described
in the BACKGROUND ART with reference to Fig. 9.
When the positive-polarity period TEN is ended, electrode 9 and
base material 12 have been short-circuited since the time point El.
Therefore, AS determination unit 6 determines that a short circuit is
present, and welding controller 3 disables a transition to the other
polarity period. In the case of Fig. 2, the transition from the positive
to the negative polarity period is disabled.
Thus, polarity reversal is disabled. This prevents polarity
reversal with a large current, and hence, prevents a surge voltage from
being caused by the switching of the secondary inverter to reverse the
polarity. This prevents damage of the semiconductor device of the

secondary inverter.
As shown in Fig. 2, a transition is made by a welding command
to the negative-polarity period without polarity reversal (no polarity
reversal). At this moment, the negative-polarity peak current (no
polarity reversal) becomes the absolute value IEPPABS (e.g., 400A) of
the negative-polarity peak current. The negative-polarity base
current (no polarity reversal) becomes the absolute value IEPBABS
(e.g., 100A) of the negative-polarity base current.
Assume that the welding current is commanded to decrease
from the absolute value IEPPABS (e.g., 400A) of the negative-polarity
peak current to the absolute value IEPBABS (e.g., 100A) of the
negative-polarity base current when a transition is made from the
negative-polarity peak period (no polarity reversal) to the
negative-polarity base period (no polarity reversal). In this case,
however, the welding current decreases only to the welding current
before-polarity.reversal IEP2 (e.g., 300A), which is larger than the
absolute value IEPBABS of the negative-polarity base current. As a
result, a transition is made to the positive-polarity period TEN while
the welding current is at the welding current before-polarityreversal
IEP2.
At this moment, electrode 9 and base material 12 have been
short-circuited since the time point E1. Therefore, AS determination
unit 6 determines that a short circuit is present, and welding
controller 3 disables a transition to the other polarity period.
Thus, polarity reversal is disabled. This prevents polarity
reversal with a large current, and hence, prevents a surge voltage from
being caused by the switching of the secondary inverter to reverse the

polarity. This prevents damage of the semiconductor device of the
secondary inverter.
The following is a description, with reference to Fig. 2, of the
operation of AC TIG welding apparatus 1 and the welding current
waveform to allow a transition from one polarity period to the other in
the case where the transition is disabled due to a short circuit occurred
during welding, and then the short circuit opens and an arc occurs (arc
recreation).
In Fig. 2, at the time point E2 when an arc is recreated, the
short circuit between electrode 9 and base material 12 opens and an
arc occurs. AS determination unit 6 determines that an arc is present,
and allows a transition to the other polarity period.
Welding controller 3 controls the current to have the current
waveform of a reference timing in which a positive-polarity period and
a negative-polarity period are alternated at a predetermined timing.
At this moment, welding controller 3 waits for the reference
timing to make a first transition from the positive to negative polarity
period after the time point E2 when an arc is recreated. At a time
point E5 when the positive-polarity period TEN is ended, welding
controller 3 controls welding output unit 2 to perform a transition from
the positive to negative polarity period.
At the time point E5, the arc is present, and the current is
significantly decreased before a transition is made to the negative
polarity period as shown in Fig. 2. Therefore, no surge voltage is
caused by the switching of the secondary inverter.
If an arc is recreated at a time point E4 in Fig. 2, welding
controller 3 controls not to make a transition at a time point E6 when

the negative-polarity period TEP is ended. Instead, welding
controller 3 waits for the next reference timing and makes a transition
from the positive to negative polarity period at the time point E5 when
the positive-polarity period TEN is ended.
As described hereinbefore, when it has been detected that
electrode 9, which is the TIG electrode and base material 12, which is
the object to be welded are short-circuited during welding, transition
from one polarity period to the other is disabled. This prevents
polarity reversal with a large current, and hence, prevents a high
surge voltage from being caused by the switching of the secondary
inverter to reverse the polarity. This prevents damage of the
semiconductor device of the secondary inverter.
Assume that conventional AC TIG welding apparatus 101
includes an overvoltage protection circuit to protect the secondary
inverter from damage. In this case, the surge voltage tends to be high
because the polarity is reversed even while electrode 109 and base
material 112 are short-circuited. As a result, the overvoltage
protection circuit stops AC TIG welding apparatus 101, thereby
reducing welding process efficiency. In AC TIG welding apparatus 1
of the present first exemplary embodiment, on the other hand, polarity
reversal is disabled during a short circuit between electrode 9 and base
material 12, thereby reducing the surge voltage. This prevents AC
TIG welding apparatus 1 from being stopped by the overvoltage
protection circuit, and hence, prevents a reduction in welding process
efficiency.
Thus, in the method of the present invention for AC TIG
welding, welding is performed by alternating a positive-polarity period

and a negative-polarity period. This method includes: detecting the
contact between a TIG electrode and the object to be welded during
welding,' and disabling a transition from one polarity period to the
other when the contact (short circuit) is detected.
With this method, polarity reversal is disabled during a short
circuit, preventing a surge voltage caused by the switching to reverse
the polarity. This achieves a high-quality AC TIG welding which
prevents the semiconductor device from being damaged.
The method may alternatively include: maintaining the current
at the time of the contact between the TIG electrode and the object to
be welded after the contact is detected during welding until the contact
opens and an arc occurs.
This method prevents polarity reversal with a large current,
and hence, prevents a surge voltage from being caused by the
switching of the secondary inverter to reverse the polarity. This
prevents damage of the semiconductor device of the secondary inverter.
The method may alternatively include: decreasing the current
value of the current when the contact is detected during welding.
This method prevents polarity reversal with a large current,
and hence, prevents a surge voltage from being caused by the
switching of the secondary inverter to reverse the polarity. This
prevents damage of the semiconductor device of the secondary inverter.
The method may alternatively include: allowing the transition
from one polarity period to the other when the contact between the TIG
electrode and the object to be welded during welding is detected and
then the opening of the contact is detected.
This method prevents polarity reversal with a large current,

and hence, prevents a high surge voltage from being caused by the
switching of the secondary inverter to reverse the polarity. This
prevents damage of the semiconductor device of the secondary inverter.
The method may alternatively include: controlling, when the
contact between the TIG electrode and the object to be welded during
welding is detected and then the opening of the contact is detected, the
current to have the current waveform of a predetermined reference
timing in which a positive-polarity period and a negative-polarity
period are alternated.
This method prevents polarity reversal with a large current,
and hence, prevents a high surge voltage from being caused by the
switching of the secondary inverter to reverse the polarity. This
prevents damage of the semiconductor device of the secondary inverter.
In the present first exemplary embodiment, a short circuit
occurs in the positive-polarity period. When a short circuit occurs in
the negative-polarity period, the same effect can be achieved by the
same control.
SECOND EXEMPLARY EMBODIMENT
Fig. 3 is a schematic configuration view of an AC TIG welding
apparatus according to a second exemplary embodiment of the present
invention, and Figs. 4 and 5 show the changes in welding current
waveforms with time in the AC TIG welding apparatus.
The present second exemplary embodiment will describe a
non-consumable electrode AC TIG welding apparatus in which welding
is performed by alternating a negative-polarity period and a
positive-polarity period. The operation of the AC TIG welding

apparatus shown in Fig. 3 will be described with reference to the
welding current waveforms shown in Figs. 4 and 5.
As shown in Fig. 3, AC TIG welding apparatus 21 includes
current-lower-than-during-a-short-circuit setting unit 13, which sets a
current obtained when electrode 9 and base material 12 are
short-circuited. Apparatus 21 differs from device 1 of the first
exemplary embodiment shown in Fig. 1 in having unit 13.
In Figs. 4 and 5, a current- IS lower than during a short circuit
indicates a current obtained when electrode 9 and base material 12 are
short-circuited.
In Fig. 3, current-lower-than-during-a-short-circuit setting unit
13, which can be composed of a CPU, receives the AS signal from AS
determination unit 6, and a current signal from current detector 4.
Unit 13 then sets the current IS, which has a current value lower than
the current value obtained upon detection of a short circuit between
electrode 9 and base material 12.
Welding controller 3 receives the AS signal from AS
determination unit 6 and the output of
current-lower-than-during-a-short-circuit setting unit
current-lower-than-during-a-short-circuit setting unit. During a
short circuit between electrode 9 and base material 12, welding
controller 3 controls the welding current to be the current IS.
In AC TIG welding apparatus 21 of the present second
exemplary embodiment, the current IS is calculated and set based on
the output current of current detector 4 obtained upon detection of the
short circuit. The current IS may alternatively be calculated and set
based on the welding current command held by welding controller 3.

It is alternatively possible to set the lower of the two values: a
predetermined current value (e.g., 100A) and a welding current value
obtained when a short circuit is detected. It is also possible to set a
value obtained by multiplying a predetermined coefficient less than 1
(e.g., 0.5) and the welding current value obtained when a short circuit
is detected.
The following is a description, with reference to Fig. 4, of the
operation to disable a transition from one polarity period to the other
when a short circuit occurs between electrode 9 and base material 12 at
the time point E1 during AC TIG welding, and the welding current
waveform.
In Fig. 4, at the time point E1 when a short circuit occurs, AS
determination unit 6 determines that electrode 9 and base material 12
are short-circuited. Welding controller 3 disables the transition from
one polarity period in which a short circuit is detected, to the other
polarity period, based on the output of AS determination unit 6. In
the case of Fig. 4, the transition from the positive to negative polarity
period is disabled.
While electrode 9 and base material 12 are short-circuited,
welding controller 3 controls the welding current to be the current IS
(e.g., 100A) set by current-lower-than-during-a-short-circuit setting
unit 13. The current IS has a current value lower than the current
value (a positive-polarity peak current IENP of 400A in Fig. 4)
obtained when the short circuit is detected.
The following is a description, with reference of Fig. 4, of the
operation to allow a transition from one polarity period to the other
when an arc is recreated while the transition is disabled due to the

occurrence of a short circuit, and the welding current waveform.
In Fig. 4, at the time point E2 when an arc is recreated, AS
determination unit 6 determines that an arc is present, so that the
transition from one polarity period to the other is allowed. Welding
controller 3 controls the welding current such that current control is
started from a polarity period (the negative.polarity period TEP)
opposite to the polarity period in which a short circuit has been
detected (in this case, the positive-polarity period TEN) as shown by
the current waveform.
Alternatively, as shown in Fig. 5, at the time point E2 when an
arc is recreated, welding controller 3 can alternatively control the
welding current such that current control is started from a polarity
period (in this case, the positive-polarity period TEN) in which a short
circuit has been detected as shown by the current waveform.
As described hereinbefore, according to the present exemplary
embodiment, when an arc is determined to be present during welding,
a transition to the other polarity period is disabled. This prevents a
high surge voltage from being caused by the switching of the secondary
inverter to reverse the polarity as in the first exemplary embodiment.
This prevents damage of the semiconductor device of the secondary
inverter.
When a short circuit is present, the current is reduced to the
current IS, which is lower than the current obtained when the short
circuit has been detected. This prevents unnecessary consumption or
damage of the electrode.
A normal welding condition is quickly resumed and welding
defect is unlikely to occur because normal AC welding is achieved

immediately after an arc is recreated.
In the method of the present invention for AC TIG welding,
welding is performed by alternating a positive-polarity period and a
negative-polarity period. This method includes: detecting the contact
between a TIG electrode and the object to be welded during welding,'
and disabling a transition from one polarity period to the other when
the contact (short circuit) is detected. The method may alternatively
include: allowing the transition from one polarity period to the other
when the contact between the TIG electrode and the object to be
welded during welding is detected and then the opening of the contact
is detected. The method may alternatively include: controlling, when
the contact between the TIG electrode and the object to be welded
during welding is detected and then the opening of the contact is
detected, the current to have the same current waveform as the
current waveform appearing at the start of a polarity period in which
the contact has been detected.
This method prevents a high surge voltage from being caused by
the switching of the secondary inverter to reverse the polarity. This
prevents damage of the semiconductor device of the secondary inverter.
The method may alternatively include: controlling, when the
contact between the TIG electrode and the object to be welded during
welding is detected and then the opening of the contact is detected, the
current to have the same current waveform as the current waveform
appearing at the start of the polarity period opposite to the polarity
period in which the contact has been detected.
This method prevents a high surge voltage from being caused by
the switching of the secondary inverter to reverse the polarity. This

prevents damage of the semiconductor device of the secondary inverter.
In the present second exemplary embodiment, a short circuit
occurs in the positive-polarity period. When a short circuit occurs in
the negative-polarity period, the same effect can be achieved by the
same control.
THIRD EXEMPLARY EMBODIMENT
Fig. 6 is a schematic configuration view of an AC TIG welding
apparatus according to a third exemplary embodiment of the present
invention, and Fig. 7 shows the change in a welding current waveform
with time in the AC TIG welding apparatus.
The present third exemplary embodiment will describe a
non-consumable electrode AC TIG welding apparatus in which welding
is performed by alternating a negative-polarity period and a
positive-polarity period are alternated. The operation of AC TIG
welding apparatus 31 shown in Fig. 6 will be described with reference
to the welding current waveform shown in Fig. 7.
In Fig. 6, AC TIG welding apparatus 31 includes
short-circuit-welding-output-stop-time setting unit 14 and second time
keeper 15, which counts the time since the occurrence of a short circuit.
Apparatus 31 differs from apparatus 1 of the first exemplary
embodiment shown in Fig. 1 in having unit 14 and second time keeper
15.
In Fig. 7, a period TRES is started when a short circuit occurs.
In the period TRES, the welding output is stopped. A time point E3 is
when the period TRES has elapsed since the occurrence of the short
circuit.

In Fig. 6, short-circuit-welding-output-stop-time setting unit 14,
which can be composed of a CPU, sets the period TRES indicating a
short-circuit welding-output stop period. The period TRES is a
threshold of the period elapsed since a short circuit occurs. The
period TRES also is a predetermined period in which the welding
output is stopped when electrode 9 and base material 12 have been in
contact with each other for a predetermined period during welding.
Second time keeper 15, which can be composed of a CPU, counts the
time since a short circuit occurs.
Welding controller 3 receives the output of AS determination
unit 6, the output of short-circuit-welding-output-stop-time setting
unit 14, and the output of second time keeper 15. If electrode 9 and
base material 12 are in contact with each other during welding for the
period TRES (e.g., 1 sec), which is the predetermined period, welding
controller 3 controls welding output unit 2 to stop the welding output.
The following is a description, with reference to Fig. 7, of the
operation to disable a transition from one polarity period to the other
when a short circuit occurs between electrode 9 and base material 12 at
the time point E1 during AC TIG welding, and the welding current
waveform.
In Fig. 7, at the time point E1 when a short circuit occurs, AS
determination unit 6 determines that electrode 9 and base material 12
are short-circuited based on the output of voltage detector 5. Welding
controller 3 disables the transition to the other polarity period based
on the output of AS determination unit 6. In the case of Fig. 7, the
transition from the positive to the negative polarity period is disabled.
After the detection of the short circuit between electrode 9 and

base material 12, welding controller 3 controls the welding current to
have the current value (in this case, the value of the positive-polarity
peak current IENP) at the time point E1 when a short circuit occurs
until the short circuit opens and an arc occurs.
Welding controller 3 receives the output of AS determination
unit 6, the output of short-circuit-welding-output-stop-time setting
unit 14, and the output of second time keeper 15. If the short circuit
continues for the period TRES (e.g., 1 sec) since the time point E1 when
a short circuit occurs, welding controller 3 stops the welding output.
Thus, if a short circuit continues for a long period, welding
controller 3 considers that this is an abnormal condition, and stops the
welding output. This prevents the continuous flow of an abnormal
short-circuit current, thus providing a safe welding process.
As described hereinbefore, according to the present third
exemplary embodiment, a transition to the other polarity period is
disabled if a short circuit occurs during welding. This prevents a high
surge voltage from being caused by the switching of the secondary
inverter to reverse the polarity. This prevents damage of the
semiconductor device of the secondary inverter.
In the method of the present invention for AC TIG welding,
welding is performed by alternating a positive-polarity period and a
negative-polarity period. This method includes: detecting the contact
between a TIG electrode and the object to be welded during welding;
and disabling a transition from one polarity period to the other when
the contact (short circuit) is detected. The method may alternatively
include: stopping the welding output when the contact between the
TIG electrode and the object to be welded continues for a

predetermined period during welding.
This method prevents the continuous flow of an abnormal
short-circuit current, thus providing a safe welding process.
The short-circuit welding-output stop period TRES may
alternatively be a predetermined fixed value, or a value determined
based on the output welding current. In the present third exemplary
embodiment, a short circuit occurs in the positive-polarity period.
When a short circuit occurs in the negative-polarity period, the same
effect can be achieved by the same control.
INDUSTRIAL APPLICABILITY
As described hereinbefore, according to the present invention, if
the electrode and the object to be welded come into contact with each
other during a large-current welding, an output polarity reversal is
disabled during a short circuit. This prevents a surge voltage from
being caused by the switching to reverse the polarity, and also prevents
the semiconductor device from being damaged. This method for AC
TIG welding is particularly applicable to industries using aluminum
and magnesium materials, such as automotive and construction
industries.
REFERENCE MARKS IN THE DRAWINGS
1, 21, 31 AC TIG welding apparatus
2 welding output unit
3 welding controller
4 current detector
5 voltage detector

AS determination unit
setting unit
first time keeper
electrode
welding torch
arc
base material
current-lower-than-during-a-short-circuit setting unit
short-circuit-welding-output-stop-time setting unit
second time keeper

1. A method for AC TIG welding in which welding is performed by
alternating a positive-polarity period and a negative-polarity period,
the method comprising:
detecting a contact between a TIG electrode and an object to be
welded during welding: and
disabling a transition from one polarity period to an other
polarity period when the contact is detected.

2. The method of claim 1, comprising:
maintaining a current at a time of the contact between the TIG
electrode and the object to be welded after the contact is detected
during welding until the contact opens and an arc occurs.

3. The method of claim 1, comprising:
decreasing a current value of a current when the contact is
detected during welding.
4. The method of any one of claims 1 to 3, comprising:
allowing the transition from one polarity period to an other
polarity period when the contact between the TIG electrode and the
object to be welded during welding is detected and then the opening of
the contact is detected,

5. The method of claim4, comprising:
controlling, when the contact between the TIG electrode and the

object to be welded during welding is detected and then the opening of
the contact is detected, the current to have a same current waveform
as a current waveform appearing at a start of a polarity period in
which the contact has been detected.
6. The method of claim4, comprising:
controlling, when the contact between the TIG electrode and the
object to be welded during welding is detected and then the opening of
the contact is detected, the current to have a same current waveform
as a current waveform appearing at a start of a polarity period
opposite to a polarity period in which the contact has been detected.
7. The method of claim4, comprising:
controlling, when the contact between the TIG electrode and the
object to be welded during welding is detected and then the opening of
the contact is detected, the current to have a current waveform of a
predetermined reference timing in which a positive-polarity period and
a negative-polarity period are alternated.
8. The method of any one of claims 1 to 7, comprising:
stopping a welding output when the contact between the TIG
electrode and the object to be welded continues for a predetermined
period during welding.

ABSTRACT

Disclosed is a method for AC TIG welding in which welding is
performed by alternating a positive-polarity period and a
negative-polarity period. It is detected whether the TIG electrode and
the object to be welded comes into contact with each other during
welding. When they are in contact with each other, a transition from
one polarity period to the other is disabled. This prevents a surge
voltage to be caused by the switching to reverse the polarity. This
prevents damage of the semiconductor device.