Escolar Documentos
Profissional Documentos
Cultura Documentos
WELDING RESEARCH
The behaviors of the keyhole and the weld pool in plasma arc welding
can be used to indicate weld quality
G. K. ZHANG, J. CHEN (chenji@sdu.edu.cn), and C. S. WU are with the Key Laboratory for LiquidSolid Structural Evolution and Processing of Materials, Ministry
of Education, and Institute of Material Joining, Shandong University, China. G. K. ZHANG is also with the Leibniz Institute for Plasma Science and Technology,
Greifswald, Germany.
WELDING RESEARCH
Fig. 1 — Schematic of controlled pulse keyholing PAW. Fig. 2 — Schematic of the experimental system for controlled
pulse keyholing PAW.
WELDING RESEARCH
Observation Strategy
Imaging Principle
Plasma arc, welding pool, and base Fig. 5 — The backside of weld pool and keyhole images in different periods (Test 1). A —
metal have different radiation intensi- Open keyhole period; B — blind keyhole period.
ties because of their different emissivi-
ties and temperatures during the con-
trolled-pulse keyholing PAW process. Table 1 — Parameters of Welding Current Waveform
The purpose of this vision system is to
detect these three kinds of radiation Test Ip Tp Tf1 Tf2 Ib Tb Welding Plate
information at one observation win- No. (A) (ms) (ms) (ms) (A) (ms) Speed Thickness
dow, which is emitted from the back- (mm/min) (mm)
side of the workpiece. For the stainless
steel, there are two important temper- 1 Varied Varied 250 100 60 100 120 8
atures at the backside of the work- 2–5 140–170 Varied 40 60 60 100 120 6
piece. Melting point (1800 K) stands 6–9 150 Varied 40 60 60 100 100–130 6
for the boundary between weld pool 10–13 150 Varied 20–80 60 60 100 120 6
WELDING RESEARCH
A B
Fig. 6 — Weld pool gray level distribution. A — Open keyhole; B — blind keyhole.
A B
Fig. 7 — Weld pool and keyhole boundary. A — Open keyhole; B — blind keyhole.
and base metal. Boiling point (3000 K) B1800K(660 nm)/B3000K(660 nm) = at observation window of 1080 nm
means the boundary between plasma 1/127, was proper to identify the weld pool
arc and weld pool, respectively. If the boundary and the radiation intensity
emissivity of the weld pool is assumed B1800K (1080 nm)/ B3000K (1080 nm) = of weld pool was still strong enough to
to be a constant value at a different 1/19. be detected without illumination.
temperature, the range of workpiece Hence, an observation window of
radiation at the backside can be calcu- It is clear that there is a significant 1080 nm was also a good choice for
lated by Plank’s law (Ref. 33), difference (1:127) in the weld pool ra- the detection of whole weld pool dur-
diation between 1800 and 3000 K at ing the blind keyhole period. Mean-
2hc 2 1 the observation window of 660 nm, while, exposure time and lens aperture
BT ( ) = 5 hc/ kT
(1)
e 1 which means the radiation difference of the camera should be optimized to
between the unmelted and melted capture the keyhole and weld pool evo-
Under the action of Ip, the work- zone was low for an 8-bit gray scale lution continuously in a complete
piece was fully penetrated. Once an image. So it was hard for the camera to pulse period with infrared filter (cen-
open keyhole was established first in identify the weld pool boundary. On tral wavelength was 1080 nm, band-
the weld pool, the plasma efflux the contrary, the weld pool radiation width was 20 nm and transparency
flowed out from the keyhole channel, difference (1:19) between 1800 and was 90%). Figure 5 was captured when
which induced a strong interference in 3000 K at an observation window of the exposure time was 1 ms and aper-
the visual signals of the camera. Get- 1080 nm was beneficial to making the ture was F1/8. It shows that not only
ting clear images of the weld pool dur- weld pool boundary and keyhole can both keyhole and weld pool be de-
ing the open keyhole period was a pri- boundary detected simultaneously by tected during the open keyhole period,
mary consideration. a single CCD camera sensor as shown but also that the weld pool image can
To get rid of the influence of Ar ra- in Fig. 4. So the observation window be captured during the blind keyhole
diation on CCD signals, two observa- of 1080 nm was a good choice for de- period.
tion windows (660 and 1080 nm) were tection of the whole weld pool during
selected as shown in Fig. 3, which was the keyhole period. Image Analysis
a LTE spectrum for Ar (Ref. 34). The When the welding current was
black body infrared radiation ratios at falling, the keyhole would close and According to the principles of cam-
different temperatures (1800 and leave the molten weld pool alone. As era sensors, pixel gray values of images
3000 K) and different wavelengths mentioned above, the radiation differ- are the reflection of objective infrared
(660 and 1080 nm) are listed as, ence (1:19) between 1800 and 3000 K intensity. For a constant emissivity,
WELDING RESEARCH
A B
Fig. 8 — The sampled electric signals in one pulse. A — Welding current and plasma arc voltage; B — welding current and efflux plasma
voltage.
WELDING RESEARCH
A B
Fig. 10 — The dimensions of weld pool and keyhole in one pulse. A — Keyhole; B — weld pool.
Fig. 11 — The keyhole center and weld pool front edge position Fig. 12 — The keyhole center and weld pool at the widest
in one pulse. position.
in Fig. 8A and efflux plasma voltage is This peak current period can be divid- lidified and the length of the total weld
shown in Fig. 8B. The images of both ed into two parts: blind keyhole period pool decreased sharply. However, the
keyhole and weld pool in a whole pulse and open keyhole period. At the first new weld pool continued growing from
cycle are illustrated simultaneously in part (15.7–16.6 s), the depth of the 16.7 to 17.1 s.
Fig. 9. The length and width of key- keyhole increased under high welding Once keyhole size was high enough,
hole and weld pool are defined as geo- current and plasma arc force, while the the backside keyhole closed to avoid
metrical dimensions and perpendicu- geometry of the backside weld pool re- melt-through defects during the weld-
lar to the welding direction, respec- mained the same, which means that ing current falling period (17.1–17.27
tively. Figure 10 is the dynamic varia- major heat input mainly acted on the s). As plasma arc force is mainly decid-
tion of keyhole and weld pool dimen- internal weld pool instead of the back- ed by welding current, a slowly de-
sions in one whole cycle. side of the workpiece. creasing welding current (slope f1) was
During the base current period At the second part of the peak cur- used first to make sure that the size of
(15.5–15.7 s), backside keyhole was rent period (16.6–17.1 s), once an open the keyhole decreased smoothly with-
closed and the temperature of the keyhole was established after welding out keyhole collapse. Then a steeper
weld pool dropped. However, the tem- current was exerted for enough time, slope f2 was utilized to completely
perature of the weld pool at the back- the plasma arc heated and melted more close the backside keyhole. Figures 9
side is still higher than the melting metal. As the plasma arc force was big- and 10 also show the size of the back-
point as heat energy was continuously ger than gravity and surface tension, side weld pool reached its highest val-
transferred from plasma arc and melt- the size of the keyhole expanded gradu- ue at 17.27 s when the backside key-
ed metal in the weld pool. As a result, ally to make a new weld pool generate hole closed completely.
the dimension of the backside weld and grow in the old weld pool that
pool could keep constant. formed in the last pulse. Meanwhile, Keyhole and Weld Pool Dynamic
During the peak current period the temperature of the old weld pool Position
(15.6–17.1 s), high welding current decreased step by step because its posi-
provided sufficient heat input and tion moved far away from the heat Figure 9 shows that the relative po-
heat accumulation to the weld pool. source. At 16.7 s, the old weld pool so- sition of keyhole and weld pool
WELDING RESEARCH
A B
Fig. 13 — Data under different peak current (tests 2–5): A — Weld pool sizes at the end of the pulse; B — keyhole area in one pulse.
A B
Fig. 14 — Data under different welding speeds (tests 6–9). A — Weld pool sizes at the end of the pulse; B — keyhole area in one pulse.
changed over time under controlled- formed in the last pulse. Under the made the metal transfer from the
pulse keyholing PAW. The dynamic be- heat function of plasma arc generated workpiece to the weld pool. Therefore,
haviors of the keyhole center and weld by peak welding current, the melting the movement of the keyhole was the
pool front edge on the backside work- rate of base metal in the front of key- direct reason why the weld pool ex-
piece in one pulse cycle are illustrated hole was greater than the welding panded along welding direction. When
in Fig. 11. Position 0 is the intersec- speed. Hence, the front edge of the the open keyhole stopped moving and
tion of the welding torch axis and weld pool moved forward and the po- closed at the base current period, the
backside workpiece. sition of the keyhole center became weld pool could not move forward
The weld pool movement in the close to 0 position (the axis of the anymore. On the contrary, the weld
blind keyhole period determined the welding torch) at the keyhole period in pool would move backward with the
position where the keyhole appears in Fig. 11. The plasma arc in keyhole, workpiece, as shown in Fig. 11.
controlled-pulse PAW. During the peak which is affected by current, gradually During PAW, the keyhole was locat-
welding current period, the plasma became weak if welding current de- ed in the front of the weld pool. The
arc, which flowed to the wokpiece creased. If metal melting rate in the maximum temperature of the backside
along the welding torch axis, penetrat- front of the keyhole was less than or workpiece was located at the position
ed the workpiece and appeared on the equal to the workpiece movement of the keyhole. However, the keyhole
backside. However, the keyhole posi- speed, the keyhole would stop moving was always located between the solid
tion was far away from the welding forward (welding current falling peri- metal and the molten pool due to its
torch axis when the backside keyhole od), which is shown in Figs. 9 and 11. movement. As shown in Fig. 12, the
began to appear. Compared with the The behavior of the keyhole affect- widest position of weld pool and the
low-temperature solid metal, flow re- ed weld pool behavior and position in location of keyhole have a certain lag
sistance in a high-temperature melting the keyhole period. The keyhole and because of thermal hysteresis.
weld pool was much smaller. As a re- weld pool were coupled together and
sult, the plasma arc flowed and pene- the front edge of the keyhole formed Influences of Welding Parameters
trated the workpiece along the front the weld pool front edge. Plasma arc in
edge of the weld pool, which was the keyhole heated the base metal and During controlled-pulse keyholing
WELDING RESEARCH
A B
Fig. 15 — Data under different current falling time (tests 10–13). A — Keyhole area in one pulse; B — weld pool sizes at the end of pulse.
PAW, the current waveform and weld- the welding process, only one parame- ing speed.
ing speed are critical to keyhole and ter was changed in each test based on 3) Influence of welding current
weld pool behaviors. It is essential to experiment design. falling time (Tf1)
analyze the influence of welding pa- 1) Influence of peak current In controlled-pulse keyholing PAW,
rameters on controlled-pulse keyhol- In this part, only peak current was welding current had two falling slopes.
ing PAW. It is clear that the function of changed to observe the behaviors of The purpose of the first slope (f1) was to
Ib was to keep a proper weld pool at the keyhole and weld pool. In tests 2–5, Ip drive the keyhole to close smoothly. The
backside during Tb and to protect it was set from 140 to 170 A. Figure 13 duration of the keyhole could be con-
from overheating. Meanwhile, Tb is the illustrates the keyhole area and weld trolled by Tf1. In tests 10–13, Tf1 was set
major parameter to guarantee the con- pool size at the backside workpiece in as 20, 40, 60, and 80 ms, respectively.
tinuity of backside reinforcement be- different tests. For a fixed thickness workpiece,
tween neighboring pulses during the With the decrease of peak current, it there exists the minimum welding cur-
welding process. In this research, both became difficult to produce open key- rent that makes arc force equal to
Ib and Tb are fixed to constant values. hole as shown in Fig. 13A. Especially gravity and surface tension to keep the
Ip, Tf1, and welding speed were selected when peak welding current decreased to keyhole open. During Tf1, the welding
to study the relationship between 140 A, the plasma arc could not pene- current was still higher than the mini-
welding current parameters and key- trate the workpiece because of low heat mum welding current to keep the key-
hole/weld pool behaviors. To decrease input. As shown in Fig. 13B, keyhole hole open. However, Tf1 was much less
the influence of keyhole forming re- size increased with the growth of peak than Tp, so the heat input during Tf1
sistance (gravity) and highlight the ef- current. However, the duration of key- was much smaller than the total heat
fect of welding parameters, Experi- hole was almost the same at different input in one cycle. Increasing Tf1 main-
ments 2–13 were conducted on 6-mm- peak currents as it was controlled by Sp ly increased the duration of keyhole
thick plate instead of 8-mm-thick and Tf1. As a result, the size of the weld time instead of keyhole size as shown
plate. The plasma gas flow rate was 2.8 pool was mainly determined by the heat in Fig. 15A. Similarly, though a longer
L/min. The welding parameters are input during the open keyhole period. A Tf1 would give the plasma arc more of a
listed in Table 1. higher peak current will lead to a larger chance to increase weld pool size, both
According to previous analysis, key- weld pool, which should be controlled in the length and width of the weld pool
hole is the heat source and the force case of burn down. increased a little, which is shown in
source to widen weld pool size and 2) Influence of welding speed Fig. 15B.
drive the weld pool moving forward. If Welding speed was one of the most
the duration of the keyhole is very important parameters to control weld- Conclusion
short, weld discontinuities will appear ing heat input. In tests 6–9, only weld-
because of inadequate melting on the ing speed was changed from 100 to 1) A vision system combined with
backside. However, if the duration of 130 mm/min to observe the behaviors an image processing strategy was de-
keyhole is very long, the weld pool will of keyhole and weld pool. Figure 14 il- veloped to detect both keyhole and
burn down because of overheating. At lustrates the keyhole area and weld weld pool information during a whole
the end of each pulse, the weld pool pool size at different welding speeds. controlled-pulse keyholing period
reaches the largest size, which reflects The keyhole area is large at a low without illumination.
the result of heat and force effects of welding speed because of high heat in- 2) During the blind keyhole period,
plasma arc. To investigate the influ- put. A bigger keyhole brings more heat heat input mainly acted on the inter-
ence of welding current waveform pa- to melt base metal and increase the nal weld pool to increase keyhole vol-
rameters Ip, Tf1, and welding speed on weld pool size under decreasing weld- ume and the weld pool varied little at
WELDING RESEARCH
the backside. The keyhole position and 1975. Keyhole stability in plasma arc weld- laser welding. Journal of Physics D: Applied
the maximum width position of the ing. Welding Journal 54(11): 401–404. Physics 38(12): 1881–1887.
backside weld pool were located be- 9. Jia, C. B., Wu, C. S., and Zhang, Y. M. 22. Saeed, G., and Zhang, Y. M. 2007.
hind the welding torch axis because of 2009. Sensing controlled pulse key-holing Weld pool surface depth measurement us-
condition in plasma arc welding. Transac- ing a calibrated camera and structured
thermal hysteresis.
tions of Nonferrous Metals Society of China light. Measurement Science and Technology
3) The location of backside weld 19(2): 341–346. 18(8): 2570–2578.
pool was at rest relative to the work- 10. Lu, W., Zhang, Y. M., and Lin, W.-Y. 23. Zhang, W. J., Liu, Y. K., and Zhang,
piece during the blind keyhole period. 2004. Nonlinear interval model control of Y. M. 2012. Real-time measurement of
It only moved forward driven by key- quasi-keyhole arc welding process. Auto- three dimensional weld pool surface in
hole under open keyhole status. The matica 40(5): 805–813. GTAW. Welding Process, ISBN 979-953-307-
sizes of weld pool and keyhole were 11. Zhang, Y. M., and Liu, Y. C. 2003. 1089-4.
more susceptible to peak current and Modeling and control of quasi-keyhole arc 24. Zhang, W., Liu, Y. K., Wang, X., and
weld speed than welding current welding process. Control Engineering Prac- Zhang, Y. M. 2012. Characterization of
falling time(Tf1). tice 11(12): 1401–1411. three-dimensional weld pool surface in gas
12. Zhang, Y. M., and Liu, Y. C. 2007. tungsten arc welding. Welding Journal
Control of dynamic keyhole process. Auto- 91(7): 195-s to 203-s.
matica 43(5): 876–884. 25. Liu, Y. K., and Zhang, Y. M. 2013.
Acknowledgments
13. Wu, C. S., Jia, C. B., and Chen, M. A. Control of 3D weld pool surface. Control
2010. A control system for keyhole plasma Engineering Practice 21(11): 1469–1480.
arc welding of stainless steel plates with 26. Liu, Y. K., and Zhang, Y. M. 2015.
The authors are grateful to the fi- medium thickness. Welding Journal 89(11): Controlling 3D weld pool surface by adjust-
nancial support for this project from 225-s to 231-s. ing welding speed. Welding Journal 94(4):
the National Natural Science Founda- 14. Steffens, H. D., and Kayser, H. 125-s to 134-s.
tion of China (Key Program Grant No. 1972. Automatic control for plasma arc 27. Liu, Z., Wu, C. S., and Gao, J. 2013.
50936003). welding with constant keyhole diameter. Vision-based observation of keyhole geom-
Welding Journal 51(6): 408–418. etry in plasma arc welding. International
15. Zhang, S. B., and Zhang, Y. M. 2001. Journal of Thermal Sciences 63(1): 38–45.
References Efflux plasma charge-based sensing and 28. Liu, Z. M., Wu, C. S., and Chen, J.
control of joint penetration during keyhole 2013. Sensing dynamic keyhole behaviors
plasma arc welding. Welding Journal 80(7): in controlled-pulse keyholing plasma arc
1. Wu, C. S., Wang, L., Ren, W. J., and 157-s to 162-s. welding. Welding Journal 92(12): 381-s to
Zhang, X. Y. 2014. Plasma arc welding: 16. Zhang, Y. M., Zhang, S. B., and Liu, 389-s.
Process, sensing, control and modeling. Y. C. 2001. A plasma cloud charge sensor 29. Liu, Z. M., Wu, C. S., Liu, Y. K., and
Journal of Manufacturing Processes 16(1): for pulse keyhole process control. Measure- Luo, Z. 2015. Keyhole behaviors influence
74–85. ment Science and Technology 12(8): weld defects in plasma arc welding process.
2. Wu, C. S. 2011. Welding thermal 1365–1370. Welding Journal 94(9): 281-s to 290-s.
processes and weld pool behaviors. CRC 17. Zhang, Y. M., and Ma, Y. 2001. Sto- 30. Liu, Z. M., Liu, Y. K., Wu, C. S., and
Press/China Machine Press. chastic modelling of plasma reflection dur- Luo, Z. 2015. Control of keyhole exit posi-
3. Fortain, J. M. 2008. Plasma welding ing keyhole arc welding. Measurement Sci- tion in plasma arc welding process. Welding
evolution & challenges. IIW Doc 12: ence and Technology 12(11): 1964–1975. Journal 94(6): 196-s to 202-s.
1948–08. 18. Saad, E., Wang, H., and Kovacevic, 31. Zhang, G. K., Wu, C. S., and Liu, Z.
4. Zhang, Y. M., and Zhang, S. B. 1998. R. 2006. Classification of molten pool M. 2014. Experimental observation of
Double-sided arc welding increases weld modes in variable polarity plasma arc weld- both keyhole and its surrounding thermal
joint penetration. Welding Journal 77(6): ing based on acoustic signature. Journal of field in plasma arc welding. International
57–61. Materials Processing Technology 174(1): Journal of Heat and Mass Transfer 70(3):
5. Martikainen, J. K., and Moisio, T. J. 127–136. 439–448.
I. 1993. Investigation of the effect of 19. Seto, N., Katayama, S., and Mat- 32. Zhang, G. K., Wu, C. S., and Liu, X.
welding parameters on weld quality of sunawa, A. 2000. High-speed simultaneous F. 2015. Single vision system for simulta-
plasma arc keyhole welding of structural observation of plasma and keyhole behav- neous observation of keyhole and weld
steels. Welding Journal 72(7): 329-s to ior during high power CO2 laser welding: pool in plasma arc welding. Journal of Ma-
340-s. Effect of shielding gas on porosity forma- terials Processing Technology 215(1): 71–78.
6. Irving, B. 1992. Plasma arc welding tion. Journal of Laser Applications 12(6): 33. Landsberg, P. T. 1978. Thermody-
takes on the advanced solid rocket motor. 245–250. namics and Statistical Mechanics. Courier
Welding Journal 71(12): 49–50. 20. Wu, C. S., Zhong, L. M., and Gao, J. Corp.
7. Zhang, Y. M., and Zhang, S. B. 1999. Q. 2009. Visualization of hump formation in 34. Ralchenko, Y., Kramida, A. E., Read-
Observation of the keyhole during plasma high-speed gas metal arc welding. Measure- er, J., and NIST ASD Team. 2007. NIST
arc welding. Welding Journal 78(2): 53-s to ment Science and Technology 20(11): 115702. Atomic Spectra Database (version 3.1.3),
58-s. 21. Fabbro, R., Slimani, S., Coste, F., physics.nist.gov/PhysRefData. National In-
8. Metcalfe, J. C., and Quigley, M. B. C. and Briand, F. 2005. Study of keyhole be- stitute of Standards and Technology,
haviour for full penetration Nd–Yag CW Gaithersburg, MD.