A Study on Arc Characteristics and Its

Application to Height Control

in Plasma Arc Cutting
The arc voltage signal is shown to be an effective characteristic for
automatically controlling electrode tip-to-workpiece height

BY T. -J. L H O , D. - G . U M A N D S. -J. N A

ABSTRACT. When cutting curved or thin the workpiece will be changed. In order been widely used.
plates, w h i c h can be easily deformed to avoid a collision between the torch In this study, therefore, a new mea-
during plasma arc cutting, it is indis- and the workpiece and to maintain good suring and processing method using the
pensable to control the motion of the cut quality, the torch must be kept at a cutting arc voltage signal is proposed for
torch to keep a constant electrode tip- constant distance from the surface of the sensing and controlling torch height. By
to-workpiece distance (arc length). In workpiece. applying this cutting arc sensor to a
this study, the characteristics of the arc For the plasma arc cutting process, plasma arc cutting system with a recti-
were analyzed and the results were used the plasma arc itself can play an impor- fied three-phase constant current power
for detecting the electrode tip-to-work- tant role in detecting the electrode tip- source, a height control system has been
piece distance during cutting. to-workpiece distance (arc length) dur- developed to keep a constant electrode
Analyzing the frequency power spec- ing cutting. It has the advantages of re- tip-to-workpiece distance.
trum of the arc voltage signal as mea- quiring no extra detectors and being un-
sured with a FFT analyzer, the first dom- affected by the arc heat, spatter and arc Analysis of Raw Arc Voltage Signal
inant frequency component was found light. From the viewpoint of reliability
to be 360 Hz, due to the current ripples and production cost, using the arc as a As shown in Fig. 1 , the cutting arc
of the rectified three-phase power sensor is suitable, especially for a high- sensor proposed in this study detects the
source, and the second dominant fre- current plasma power source. distance ( L \ between the electrode tip
quency was 19 kHz, due to the anode H. Fujimura, etal. (Ref. 2), has devel- and top surface of the workpiece by mea-
spot motion. The arc voltage signal, de- oped a joint tracking system using the suring the lower peak value of the cut-
tected from the lower peak value and fil- welding arc sensor in G M A W . H. No- ting arc voltage (Va) corresponding to
tered by the moving averaging method, mura, etal. (Refs. 3, 4), has developed the arc length (La) between the electrode
increased almost linearly with the in- an automatic real-time torch height con- tip and anode spot. Generally, the cut-
crease in electrode tip-to-workpiece dis- trol system with the welding arc sensor ting arc voltage measured can be re-
tance. This arc characteristic could be in GTAW and in multipass G M A W . The garded as a linear function of the arc
effectively used to realize a reliable automatic welding systems mentioned length for plasma arc cutting with a con-
height control system for plasma arc cut- above are now applied to many com- stant current power source (Ref. 7).
ting processes. mercial arc welding robots and auto- The power source used here has a
matic welding setups. The cutting arc rectified direct current (DC) output and
Introduction sensor, however, has been rarely used a constant current characteristic with
for detecting the electrode tip-to-work- thyristor control, for w h i c h the main
Plasma arc cutting (PAC) is a process piece distance in the plasma arc cutting specification is shown in Table 1. The
that severs a metal by melting a local- process (Refs. 5, 6), while various joint mixture of 5 0 % argon and 50% nitrogen
ized area with the heat of a constricted clearance sensors such as eddy current, is undoubtedly widely used as a cutting
arc and removing the molten material capacitance, pressure and tactile have gas for cutting low-carbon steel plates.
with a high velocity jet of the hot, ion- However, since a pure argon gas was
ized plasma gas issuing from the orifice easily available, it was employed in this
(Ref. 1). The automatic plasma arc cut- study for cutting experiments, because
ting system has been widely adopted in the main objective of this study was to
many manufacturing fields to improve KEY W O R D S investigate the feasibility of using the arc
cutting productivity and to release op- as a sensor based on the arc character-
erators from a harsh working environ- Plasma Arc Cutting istics. The shielding gases that are used
ment. Arc Characteristics for cutting low-carbon steels are oxygen,
When cutting curved or thin plates, Torch Control carbon dioxide, air and nitrogen in de-
w h i c h can be easily deformed during Tip-to-Workpiece scending order of the performance (Ref.
thermal cutting, the distance between Arc Voltage Signal 8). Compressed air at a rate of 5 kg/cm 2
the electrode tip and the top surface of Dominant Frequency was supplied for shielding and cooling
Thin Plate in this study because of its lower cost
Curved Plate and easy availability. The conditions of
T. -j. LHO is a Ph.D. Candidate, D. -C. UM Cutting Sensor the experiments that may have some ef-
is a Researcher and S. -J. NA is an Associate
Mathematical Analysis fects on the relationship between the arc
Professor, Department of Production Engi-
neering, Korea Advanced Institute of Science characteristics and electrode tip-to-
and Technology, Seoul, Korea. workpiece distance are shown in Table

W E L D I N G RESEARCH SUPPLEMENT I 277-s

 1. In order to establish the fundamental
Table 1 Test Parameters
relationship between the cutting arc
voltage and electrode tip-to-workpiece
Output current range 40-100 A
distance, the behavior of the anode spot
Used output current 70 A for 1.6 mm thickness
100 A for 6.4 mm thickness and the arc characteristics of the power
Power source
Rated output voltage 150 V source must be first investigated for the
Open circuit voltage 205 V plasma arc cutting process. The anode
Electrode material Thoriated tungsten spot is generally defined as the portion
Electrode diameter 2.4 mm of the positively charged workpiece
Orifice diameter 1.5 mm within which the electrons are absorbed.
Cutting torch Cooling system Air cooling type The motion of anode spots during
Flow delivery type Tangential plasma arc cutting brings about the vari-
Arc type Transferred ation of the arc length, and subsequently
Torch travel Lead-screw carriage affects the arc voltage signal measured.
Cutting gas Ar gas 7.5 L/min. For plasma arc cutting, the anode spot
Cases Shielding gas Compressed air 5 kg/cm2 is carried away from the upper to lower
(above 125 L/min.) side and vice versa on the front cut sur-
Workpiece low-carbon steel plates with 1.6-mm and 6.4-mm thickness face due to the momentum of the plasma
and cold surrounding gas stream, and
also due to the magnetic force acting on
the anode spot and plasma arc column
(Ref. 9). Consequently, the anode spot
periodically repeats such a motion, so
that the arc voltage fluctuates in saw-
tooth waveforms at a frequency corre-
-e electrode sponding to the anode spot motion.
In the first part of the investigation, a
typical arc voltage signal of 500 sam-
ples was measured for a sampling time
N arc \ of 35 u.s using a data acquisition system
$ ^/voltage \ i/ interfaced to a microcomputer. The cut-
ting was performed on 6.4-mm (0.25-
in.) thick plate with a current of 1 00 A
plasma
arc \ ^
A


-
La
Ls

1
and a speed of 500 mm/min (20 in./min).
The result is plotted in Fig. 2, w h i c h
shows a periodic waveform of sawtooth
type oscillating with a very large
ffl peak-to-peak value.
< work piece

dross
\1 ^^anode sp ot

ejected
f In order to analyze the frequency
components of the arc voltage signal
measured above, the discrete Fourier
transformation of the sampled arc volt-
plasma age signals, X[(f m ), was defined at the i-
th averaging segment as follows (Ref.
10).
Fig. 1 Schematic diagram showing electrode tip-to-workpiece dis-
tance measured by cutting arc voltage signal.
X,(/,) = Ar ]>>,(A-A0 zxpi-j 2 xfj At)

where, f m ( = m / NAt for m = 0, 1, ....,

N - 1) is the frequency, t the time, N the
number of samples, At the sampling time
and Xj(kAt) the arc voltage sampled at t
= kAt for the i-th averaging segment. The
one-sided power density function (PDF)
G x x (f m ) of the signal can be then intro-
duced and described as follows (Ref.
10).

GMJ- naNAt **(/) W J

200 300 Samples 500
N u m b e r of Data N_
m = 0.
"2
Fig. 2 Typical waveform of arc voltage signal measured during
cutting. where Xj (fn is the conjugate complex
of Xj(f m ) and n a the number of averag-

278-s IAUGUST1992
 Figure 3 shows the frequency power POWER 8PEC1 gOAva oxovip Hfnn
spectrum of the measured arc voltage eoo
signal after n a = 30 averages with the
FFT (Fast Fourier Transformation) ana- 7S.0
lyzer. The cutting conditions were the m
same as above, and At-35 ps, N = 500, /Div
f max = 50 kHz and G m a x x x = 600 mV 2 .
The signal shown in Fig. 3 was averaged
30 times to eliminate the noise compo-
nent from the measured raw arc voltage
signal. The frequency power spectrum Maa
of the arc signal is shown again in Fig.
4 after zooming that of Fig. 3 up to the a
frequency of 2 kHz. The dominant fre- V2
quency component is 360 Hz and its
multiples. This is due mainly to the 360
Hz ripples of the direct current (DC) out-
put of the power source used, because
the alternating current (AC) output from
a transformer is fed to a full-wave recti-
fier, which converts it to DC. Since the
0.0
1 Hz SOk
60 Hz three-phase AC was used for the
input of the plasma arc power source,
the DC output should have 360 Hz rip- Frequency, /
ples. As indicated in Fig. 5, a small fluc- m
Fig. 3 - Frequency power spectrum of cutting arc voltage signal (ntl = 30, At 35 \is, N
tuation of the arc current {i.e., ripples of
500, fmax= 50 kHz and G* = 600 mW).
DC output) gives rise to a large variation
of the arc voltage in the constant cur-
rent power source of this plasma arc cut-
ting machine.
The variation of the cutting arc volt-
age filtered by a low-pass filter with a
cut-off frequency of 1 kHz is shown in
Fig. 6. Compared to the waveform of Fig.
2, the signal component of a small os-
cillation and high frequency was almost C3
completely excluded. Moreover, it can
be also seen that the values of the lower
peak are more stable than those of the
upper peak. This is probably due to the
fact that the dross adherence on the
lower part of the cutting surface is not
uniform during cutting.
By increasing the frequency power
spectrum of Fig. 3 up to the maximum
power spectrum of G m a x x x = 6.1 6 mV 2 , Frequency, fm
it can be seen that the power also has a
Fig. 4 The frequency power spectrum of the arc voltage signal of Fig. 3 (fmax = 2 kHz).

200
Constant Current Volts
Power Source
150

Arc Voltage M 100

Fluctuation x

50

100 200 300 400 500

Number of Data
Fig. 6 Waveform of cutting arc voltage processed by a low-pass
filter with a cut-off frequency of 1 kHz.
Current

>
*K^Current Ripple

Fig. 5 Fluctuation of arc voltage according to cutting current rip-

ples of a power source with constant current characteristics.
W E L D I N G RESEARCH SUPPLEMENT I 279s
 Real-Time Signal Processing
and Height Control
X-19kHz Y - 6 . 15599lfiV2
Ya-5.35691mV2 Figure 8 represents the v a r i a t i o n of
POWER SPEC1 30AVQ
B. 16 t h e r a w arc v o l t a g e signal m e a s u r e d for
H the linearly varying electrode tip-to-
w o r k p i e c e distance under the cutting
769 c u r r e n t of 1 0 0 A a n d c u t t i n g v e l o c i t y of
5 0 m m / m i n (2 i n . / m i n ) . It c a n b e seen
/Div
that the arc voltage signal represents
considerably high peak-to-peak values,
w h i l e its a v e r a g e d v a l u e rises w i t h t h e
8 increasing e l e c t r o d e - t o - w o r k p i e c e dis-
t a n c e of 2 t o 1 0 m m (0.08 to 0.4 in.). As
Mag
i n d i c a t e d i n t h i s f i g u r e , h o w e v e r , it is
Q v e r y d i f f i c u l t t o d i r e c t l y use t h e r a w arc
CL,
rmi v o l t a g e signal for d e t e c t i n g the e l e c t r o d e
V2 tip-to-workpiece distance w i t h o u t any
special processing of the signal because
the f l u c t u a t i o n (i.e., p e a k - t o - p e a k value)
of the signal is t o o large.
Figure 9 s h o w s t h e result of p l o t t i n g
t h e l o w e r peak v a l u e x p (i) of t h e r a w sig-

Fxd
O.O
Y 0
j

Hz
A ... ,
50k
nals Xj(kAt) s a m p l e d w i t h At = 4 0 ps a n d
N = 1 0 0 0 in t h e i-th p e a k d e t e c t i n g p e -
riod f r o m t = (i-1 )T to t = iTp, T p = 4 0
Frequency, /
111 ms, w h e r e t h e c u t t i n g c o n d i t i o n w a s the
s a m e as a b o v e . T h i s o u t p u t s i g n a l w a s
Fig. 7 Frequency power spectrum of arc voltage signal of Fig. 3 for Cmaxxx = 6.16 m V2 (na
processed u s i n g a l o w e r p e a k d e t e c t i n g
= 30, At = 35 us, N = 500 and fnwx = 50 kHz).
m e t h o d , w h i c h detects the l o w e r peak
v a l u e t h r o u g h the use of a software p r o -
g r a m . As s h o w n in Fig. 9 , h o w e v e r , it is
s m a l l f r e q u e n c y i n t e n s i t y p e a k v a l u e at surface of the c u t t i n g z o n e , w h i c h cor- a l s o d i f f i c u l t t o d i r e c t l y use t h e l o w e r
a b o u t 19 k H z , as s h o w n in Fig, 7. T h i s responds t o the c o n d i t i o n o f the b o t t o m p e a k signals for d e t e c t i n g t h e e l e c t r o d e
f r e q u e n c y is c o n s i d e r e d t o be the d o m - of the o u t p u t c u r r e n t , the c u t t i n g arc v o l t - t i p - t o - w o r k p i e c e distance, because the
i n a n t f r e q u e n c y o f the a n o d e s p o t m o - age s h o u l d b e c o m e a l o w e r peak v a l u e s i g n a l is s o m e w h a t u n s t a b l e a n d its
t i o n , w h i c h is nearly consistent w i t h t h e d u e t o t h e shortest arc l e n g t h . In o r d e r p e a k - t o - p e a k v a l u e is still t o o large.
p r e v i o u s o b s e r v a t i o n t h a t t h e arc v o l t - to identify the d i s t a n c e b e t w e e n the elec- The o u t p u t signal V m ( i ) of a m o v i n g
age signal f l u c t u a t e s in s a w t o o t h w a v e s trode tip and t o p surface of the w o r k - a v e r a g i n g f i l t e r (Ref. 11) at t i m e t = i T p
at a b o u t 2 0 k H z in p l a s m a arc c u t t i n g p i e c e d u r i n g c u t t i n g , it is r e a s o n a b l e t o c a n be d e f i n e d as f o l l o w s :
(Ref. 9). T h i s k i n d of a n o d e spot m o t i o n measure the l o w e r peak v a l u e of the arc
results i n a s m a l l o s c i l l a t i o n o f t h e arc v o l t a g e s i g n a l . M o r e o v e r , f r o m the v i e w
v o l t a g e s i g n a l , w h i c h is c o m b i n e d w i t h point of signal stability, the lower peak K,(0 = d )VJi-l) + -x(i)
the 3 6 0 H z ripples of the rectified three- v a l u e is v e r y s u i t a b l e as a process c o n - w w
p h a s e p o w e r s o u r c e Fig. 3. C o n s e - trol parameter. Thus, the electrode tip- i = 1, 2,..., n
q u e n t l y t h e a n o d e spot is c o n s i d e r e d to t o - w o r k p i e c e d i s t a n c e c o u l d be d e t e r -
o s c i l l a t e in a s m a l l a m p l i t u d e at a b o u t m i n e d by d e t e c t i n g the l o w e r peak v a l u e
w h e r e V m ( i - 1 ) is t h e m o v i n g a v e r a g e d
19 k H z , w h i l e the path of the a n o d e spot o f t h e arc v o l t a g e s i g n a l s a m p l e d in a
v o l t a g e signal at t = ( i - 1 ) T p , n t h e n u m -
m o t i o n m o v e s o n t h e c u t t i n g s u r f a c e at s a m p l i n g t i m e l o n g e r t h a n 'A s, w h i c h
ber o f m o v i n g a v e r a g i n g segments, a n d
360 Hz. is r e l a t e d t o t h e 3 6 0 H z r i p p l e s o f t h e
w the w e i g h t i n g factor. T h e o u t p u t w a v e -
p o w e r source.
W h e n the a n o d e spot is o n the u p p e r f o r m of t h e signal s h o w n i n Fig. 9 f r o m

60 100

3 4 5 6 7 b 10 3 4 5 6 7 8 10

Tip-to-workpiece distance, Ls Tip-to-workpiece distance, Ls

Fig. 8 Variation of raw arc voltage signal for a linearly varying Fig. 9 Lower peak values of sampled raw arc voltage for linearly
electrode tip-to-workpiece distance. varying electrode tip-to-workpiece distance (At = 40 |is, N = 1000,
Tp = 40 ms).

280-s I A U G U S T 1992
the moving averaging filler with w = 3,
T p = 40 ms was plotted in Fig. 10. As
shown in the figure, the arc voltage sig-
nal rises almost linearly w i t h the in-
creasing electrode tip-to-workpiece dis-
tance from 2 to 10 m m , w h i c h can be
used to realize a height control system
for keeping a constant electrode tip-to-
workpiece distance during cutting. The
signal processing procedure, such as the Fig. 10 Output
lower peak detecting and moving aver- waveform processed
aging, is presented concisely in the 10 from arc voltage of
flowchart of Fig. 11. Fig. 9 by using mov-
Tip-to-workpiece distance, Ls ing averaging filter
Figure 1 2A and B show the appear-
with w = 3.
ance of the resultant cuts of the curved
plates manufactured by using a CNC
plasma arc cutting system equipped with
the developed height control system. It
can be demonstrated that the cutting arc
voltage signal effectively controls the ( start J
electrode tip-to-workpiece distance in
real time.

Conclusions

Based on the lower peak detecting

and moving averaging method, the arc
voltage signal was processed from the
measured raw values to control the elec-
trode tip-to-workpiece distance during
plasma arc cutting. The conclusions can
be summarized as follows:
1) Analyzing the frequency power
spectrum of the cutting arc voltage sig-
nal measured w i t h a FFT analyzer, the
first dominant frequency component
was found to be 360 Hz, due to the rip-
ples of the rectified three-phase cutting
power source used, and the second
dominant frequency was found to be 19
Fig. 11 Flowchart
kHz, due to the anode spot motion.
of signal processing
2) The distance between the elec- procedure of lower
trode tip and top surface of the work- peak detecting and
piece can be determined by detecting moving averaging
the lower peak value caused by the method.
anode spot motion on the upper part of
the cut surface, where the sampling time
should be longer than &o s.
3) The arc voltage signal detected
from the lower peak value and filtered
by using the moving averaging method
( stop )
rises almost linearly when increasing
electrode distance from 2 to 10 mm. This
voltage signal can be effectively used to

Fig. 12 Appearance of curved plate cuts.

W E L D I N G RESEARCH SUPPLEMENT I 281-s

keep a constant electrode tip-to-work- N. 1987. Automatic real-time bead height 8. Severance, W . S., and Anderson, D. G.
piece distance during plasma arc cut- control with arc sensor in TIG welding. Trans, 1984. H o w plasma arc cutting gases affect
ting, resulting in good cut quality on of JWS, Vol. 18, No. 2, pp. 3 5 ^ 2 . productivity. Welding Journal 57(2) 3 5 - 3 9 .
curved or thin plates, which easily de- 4. ibid. pp. 4 3 - 5 0 . 9. Nishiguchi, K., et al. 1972. Fundamen-
5. Boehme,)., Odrich, D., Wenzel, B. D., tal studies on cutting phenomena in plasma
form during plasma arc cutting.
and Jagieniak, P. 1984. Plasma torch stand- arc cutting (1st Report). Quarterly J. of JWS,
off distance control for precision control of V o l . 4 1 , No. 7, pp. 7 8 1 - 7 9 1 .
References plasma cutting (patent). WPI85-0322306/06. 10. Bendat, J. S., and Piersol, A. G. 1986.
6. Frohlich, H., Berger, A., Roloff, H. ]., Random Data (Analysis and Measurement
1. Recommended Practices for Plasma Arc and Telzer, H. ). 1985. Automatic nozzle Procedures). John W i l e y and Sons, pp.
Cutting C5.2-83, pp. 1-5. 1983. American stand-off distance control for plasma arc cut- 361-422.
Welding Society. ting torch (patent). WPI 85-249671/41. 11. Braun, S. 1986. Mechanical Signature
2. Fujimura, H., Ide, E., and Inoue, H. 7. Cook, G. E. 1 983. Robotic arc w e l d - Analysis (Theory and Applications). Aca-
1987. Joint tracking control sensor of G M A W . ing: research in sensory feedback c o n t r o l . demic Press, pp. 1 4 1 - 1 7 1 .
Trans, of JWS, Vol. 18, No. 1, pp. 32^10. IEEE, Trans, of Industrial Electronics, V o l . IE-
3. Nomura, H., Sugitani, Y., and Tamaoki, 3, No. 3, pp. 252-268.

Research on Modern High-Strength Low-Alloy Steel Welding

WRC Bulletin 373

June 1992
(1) Influences of Steel Composition and Welding Procedure on the HAZ Tough-
ness of Thick-Section Structural Steels
By P.L. Harrison and P. H. M. Hart

(2) Heat-Affected Zone Properties of Thick-Section Microalloyed Steels

A Perspective
By F. Heisterkamp, K. Hulka and A. D. Batte

(3) Experience in Fabricating New Types of Offshore Plate and Linepipe

By P. Tuvnes and I. Harneshaug

(4) Influence of Local Brittle Zone on HAZ Toughness of TMCP Steels

By S. Aihara and K. Okamoto

The four papers contained in this Bulletin were presented at the Conference on "Metallurgy, Welding and Qualifica-
tion of Microalloyed (HSLA) Steel Weldments," held at Houston, Tex., November 6-8, 1990. The American Welding So-
ciety holds the copyrights and is the source of these papers. Publication of this document was sponsored by the Weld-
ing Research Council, Inc. The price of WRC Bulletin 373 is $40.00 per copy, plus $5.00 for U.S. and $10.00 for over-
seas, postage and handling. Orders should be sent with payment to the Welding Research Council, Room 1301, 345
E. 47th St., New York, NY 10017.

282-s I AUGUST 1992