Professional Documents
Culture Documents
BY Y. H. KANG AND S. J. NA
ABSTRACT. Narrow groove welding is an tal ones. The periodic signal was adopted sideways with respect to the welding di-
important technique for increasing pro- to develop an automatic joint tracking sys- rection could be used for strip cladding
ductivity in the manufacture of thick- tem in narrow groove GMAW. and welding of materials that are sensitive
walled components. The nature of the to hot cracking because this gives a wide
process demands a n a u t o m a t e d approach Introduction bead and uniform and shallow penetra-
and precise control to ensure consistently tion (Refs. 2, 3). Subjecting the welding
high weld quality. The most important ob- A magnetic field externally applied to arc to transverse magnetic fields increases
jective of narrow groove welding is to a welding arc deflects the arc by electro- the welding speed several times, at which
maintain uniform and sufficient penetra- magnetic force (Lorentz force) in the rate undercut-free and no-porosity welds
tion at both groove faces. Several different plane normal to the field lines. The mag- can be made (Ref. 4). Also, magnetic arc
approaches, such as wire bending tech- netic field exerts the force on the elec- oscillation could be applied to high-speed
nique and wire rotating method, have trons and ions within the arc, which joint tracking because the magnetically
been adopted in an attempt to minimize causes the arc to be deflected away from oscillating arc possesses virtually no iner-
the incomplete fusion in the narrow the normal arc path. The welding arc can tia (Ref. 5).
groove G M A W process. In this study, a be deflected forward, backward, or side- Narrow groove welding is an important
welding system using electromagnetic arc ways with respect to electrode and weld- technique for increasing productivity in
oscillation was developed for narrow ing direction, depending upon the direc- the manufacture of thick-walled compo-
groove welding. The electromagnet for tion of the external magnetic field. A nents. Narrow groove welding has many
applying a magnetic field to the welding transverse magnetic field deflects the arc advantages such as high productivity and
arc was designed from the electromag- in the welding direction, whereas a longi- quality, minimal distortion, and all-posi-
netic analysis results. This paper shows the tudinal magnetic field deflects the arc tion capability. But incomplete fusion into
arc and bead characteristics in narrow perpendicular to the bead. If a unidirec- groove faces is the most frequent defect in
groove G M A W using electromagnetic arc tional magnetic field is applied to an AC narrow groove G M A W due to the low
oscillation. Based on the results, the ap- arc, or an alternating field is applied to a heat input and small molten weld pool. In
propriate welding and oscillation condi- DC arc, then the arc can be oscillated in order to improve the weld quality, an arc
tions were selected to satisfy high weld the position normal to the direction of weaving technique has to be used. Arc
quality. Consequently, magnetic arc oscil- welding, and this has been used to im- length control and weld joint tracking are
lation resulted in uniform and sufficient prove the arcs with both gas tungsten arc also needed because the weld quality is
penetration to both groove faces. Arc sig- welding (GTAW) and gas metal arc weld- sensitive to any disturbance of the arc.
nal characteristics for automatic joint ing (GMAW) (Ref. 1). Oscillating the arc Therefore, the nature of the process de-
tracking were also investigated. The peri- mands an automated approach and pre-
odic change of welding current in electro- cise control to ensure consistently high
magnetic arc oscillation was examined by weld quality. The most important objec-
KEY W O R D S
experiments and numerical analysis. To tive of narrow groove welding is to main-
establish the mathematical model of the tain uniform and sufficient penetration at
Narrow Groove Welding
arc sensor, some assumptions were both groove faces. Several different ap-
G M A Welding
needed to calculate the arc length. Ana- proaches have been adopted in an attempt
Electromagnetic Arc
lytical results using these assumptions to minimize the incomplete fusion in the
Oscillation
showed good agreement with experimen- narrow groove G M A W process (Refs.
Automated Welding
6-9). To improve the groove face fusion,
Arc Sensor
Y H. KANG is with Biz4M Co., Ltd., Seongnam, the electrode may be oscillated by adopt-
Korea. S. J. NA is with Dept. of Mechanical Eng., Joint Tracking
ing a wire bending technique in which the
Korea Advanced Institute of Science and Tech- bending direction is periodically changed,
nology (KAIST). Taejon, Korea.
I WELDINGJOURNAL I~gl~"9
Yoke
i _~ ~ / ~ /
/ Core
Coil
m SUB =1
TIME=I
SUE =I
TIME=I
AS ESUM (AVG)
RSYS=0 RSYS=0
SMN =-.004217 PowerGraphlcs
SMM =.~37E-03 EZACET=I
-,004130 AVRES=MaT
00397 SHN =.023856
Fig. 3 - - Magnetic flux lines and distribution of magnetic flux density in the welding area.
Fig. 4 - - Narrow groo~,e GMA wehling head. Fig. 5 - - Arc column images for various oscillation widths during narrow groove weMing.
creasing flux density. The oscillation width Formation of Weld Bead to 75 gauss by changing current applied to
was small at a density of 25 gauss, and os- the electromagnet. The oscillation fre-
cillation effect could not be expected. At The influence of welding parameters quency was fixed at 30 Hz, and the other
75 gauss, the width was too large and un- on weld bead formation using electromag- welding conditions were the same as in
dercut at the groove face could be ex- netic are oscillation was investigated. The previous experiments. Macrosections ob-
pected. The deflected arc is directed to- examined parameters were oscillation fre- tained from this experiment are shown in
ward the corner between the weld quency and magnetic flux density (or os- Fig. 8. Increased density caused increased
material and the backing plate at a density cillation width) applied in the welding oscillation width, as mentioned previ-
of 50 gauss. area. Macrosections of weld beads under ously. The penetration to groove faces,
Figure 6 shows images of the arc col- various oscillation frequencies from 0 to therefore, increased with increased mag-
umn for root openings of 8 and 10 mm. 50 Hz are shown in Fig. 7. The figure netic flux density. In conventional straight
Despite operating under the same welding clearly indicates the side penetration P welding without magnetic arc oscillation,
conditions and with the same magnetic with arc oscillation is deeper than without the side penetration P was especially
flux density, the arc shape and oscillation oscillation. Side penetration P slightly in- small, while the penetration H was very
width are very different from each other if creased and the penetration H decreased large. But at a density of 75 gauss, under-
the root opening decreases from 10 to 8 with increased arc oscillation frequency. cut occurred at both groove faces. The
mm. In the 8-mm root opening, the de- The relationship between magnetic maximum magnetic flux density would be
flected arc occurred between wire tip and flux density and bead formation was in- limited to 50 gauss under the welding con-
groove wall. vestigated. The density was varied from 0 ditions in this experiment with a 10-mm
W E L D I N G J O U R N A L Ii,~I;B,.'!
root opening for a square groove weld.
Gap The same results could be obtained by in-
width LEFT CENTER RIGHT vestigating the images of the arc column
(mm) - - Fig. 5.
Figure 9 shows a macrosection of a
thick plate. The thickness of the plate is 22
!0 mm with a 10-mm root opening. A deposi-
tion of 4-5 mm per pass was obtained at
280 A , 31 V, and 22-cm/min welding speed.
An average depth of side penetration was
about 1.7 mm. A stable uniform penetra-
tion into the groove faces was secured by
8 using magnetic arc oscillation.
"o
3o~ dt Ls +Lc
E 2.5-
K,+Rc+n.+Re I
m 13. 2.0-
L s + L¢
ff 1.5-
1.0-
Eal + E a i l La
a
0.5- L s + Lc (1)
0.0 , . , - , - , . , . , - , . , .
LtI
t
Je f
t-zS/ (3)
As shown in Fig. 10, the arc length L a
is expressed by the geometrical relation-
ship of the deflection angle 0, the contact Fig. 9 - - L~antple qf macroseetion o['thick plate
tip to workpiece distance Lp, the wire ex- welded by N G W using magnetic arc" oscillation W
tension length L e, and the distance be- (thickness 22 mm, root opening 10 mm).
tween the electrode wire and groove face Fig. 10--Arc length in narrow groove welding.
W. Thus, the arc length L a can be written
as follows:
(Lt-Le)tanO<_W-+La -Lt-Le
cosO (4)
4004
0'2
,
04
.
05
, . ,
08
.
a. W < Wc
1. La > W - - , Arc length = W 250.
BpN
.~ 20,3 ,
Equations 4 or 5
Experiment
15o. -- Simulation
b. W > Wc --+ Arc length =
1004
Equations 4 or 5 oo o2 o4 06 08 10 D0 02 O4 O6 O8
Welding time [sec] Welding time [ s e c ]
~g lOO4
2£ 200,
g
specimen with both groove faces, as shown
Experiment
-- Simulation in Fig. 14. The waveforms of the welding
current were monitored under magnetic
c ~
O0 0.2 04 06 08 I0 flux density of 50 gauss and arc oscillation
Welding time [sec] frequency of 5 Hz with a 10-mm root
opening for a square-groove weld. Figure
Fig. 12 - - Welding current waveforrns under various distances between welding wire and groove face at 14A shows a welding current waveform for
a magnetic flux density o f 75 gauss. no deviation between the welding head
and groove center. Figure 14B shows a
waveform that indicates a deviation of the
welding head to the right side by about 1
mm. A waveform frequency of 10 Hz
Wire [-----] showed no deviation. Two similar wave-
electrod ~ T o r c h l forms of welding current were obtained in
one period of arc oscillation. The wave-
form frequency was 5 Hz for the 1-mm de-
Arc._ I ~ I Base viation. This periodic change of welding
~',i,Iil .. J material current can be used as the output signal
for an arc sensor for automatic joint track-
ing in narrow groove GMA welding.
Backing plate
Conclusions
350 ,
I~I:!I.,.1 MAY2003
the h i g h - f r e q u e n c y noise of the welding Institute for providing m a t c h i n g funds for 6. Malin, V. Y. 1983. The state-of-the-art of
power source, the welding c u r r e n t fluctu- this project. narrow gap welding - - part I. WeldingJournal
a t i o n - - t h e basic signal c o m p o n e n t - - 62 (4): 22-30.
p r o d u c e d by m a g n e t i c arc oscillation was References 7. Sugitani, Y., Kobayashi, Y., and Mu-
clearly shown. Basic signal c o m p o n e n t s rayama, M. 1991. Development and application
can be used as the o u t p u t signal of the arc 1. Serdyuk, D. B., and Kornienko, A. N. of automatic high-speed rotation arc welding.
sensor for a u t o m a t i c joint tracking in nar- 1963. The welding arc in an alternating trans- Welding International 5(7): 577-583.
row groove G M A welding. verse magnetic field. Automatic Welding (10): 8. Kimura, S., Ichihara, I., and Nagai, Y.
N u m e r i c a l analysis of the arc s e n s o r 7-13. 1979. Narrow gap, gas metal arc welding
was also carried o u t for t h e o r e t i c a l pre- 2. Mallya, U. D., and Srinivas, H. S. 1993. process in flat position. WeldingJournal 58(7):
d i c t i o n . T h e arc s h a p e a n d o s c i l l a t i o n Magnetic steering of arc and bead characteris- 44-52.
width varied d e p e n d i n g o n the root o p e n - tics in submerged arc strip cladding. Welding 9. Ostrovskii, O. E., Kulik, V. I., and Kaza-
ing width. Some assumptions in establish- Journal 72 ( 11): 517-s to 522-s. kov, V. A. 1995. Narrow gap welding large thick-
ing the m a t h e m a t i c a l model of the arc sen- 3. Tseng, C. E, and Savage, W. F. 1971. The nesses with a deformed electrode. Welding In-
sor w e r e n e e d e d to c a l c u l a t e t h e arc effect of arc oscillation. WeldingJournal 50 (11): ternational 9(9): 731-734.
length. Analytical results using these as- 777-s to 786-s. 10. Ecer, G. M. 1980. Magnetic deflection of
s u m p t i o n s showed fairly good a g r e e m e n t 4. Hicken, G. K,. and Jackson, C. E. 1966. the pulsed current welding arc. WeldingJournal
with e x p e r i m e n t a l ones. The effect of applied magnetic fields on weld- 59(6): 183-s to 191-s.
ing arcs. WeldingJournal 45( 11): 515-s to 524-s. I 1. Ushio, M., and Mao, W. 1996. Modeling
Acknowledgment 5. Hughes, R. V., and Walduck, R. P. 1987. of an arc sensor for DC MIG/MAG welding in
Electromagnetic arc path control in robot open arc mode: study of improvement of sensi-
This work was s u p p o r t e d in part by the plasma welding. J. D. Lane, ed., Robotic WeM- tivity and reliability of arc sensors in GMA
B r a i n K o r e a 21 Project. T h e a u t h o r s ing, pp. 243-263. IFS Publication and Springer- welding (lst Report). Q. J. Japan. Weld Soc.
would like to t h a n k the P O S C O R e s e a r c h Verlag. 14(I): 99-107.