Key words: A-TIG, Activated flux, TIG, Cryogenic, Underfill phenomena, Cold wire feeding process, Slag

1. Introduction

Tungsten inert gas (TIG) welding is a welding method that melts and bonds the base metal and welding consumables with the arc heat generated between the tungsten electrode, a non-consumable electrode, and the base metal while using inert gas, such as argon (Ar) and helium (He), as shielding gas¹⁾. This welding method can be applied to all welding positions, and it can obtain high-quality welds with high tensile strength and impact toughness because the arc is very stable and the oxygen content in the welds is lower compared to the welds generated by other welding methods²⁾. Since the TIG welds of cryogenic steel, such as stainless steel (STS) and invar alloys, have high impact toughness even at cryogenic temperatures, they have been widely applied to cryogenic structures, such as LNG carrier cargo containment systems^3-5). They, however, have low productivity due to the low welding rate caused by the low deposition rate compared to gas metal arc welding (GMAW), which generates an arc at the front end of the continuously fed welding consumables and melts them with the arc heat¹⁾.

To complement the shortcomings of this TIG welding method, studies have been actively conducted on plasma arc welding (PAW) and keyhole TIG welding processes that increase the penetration depth and the amount of weld deposition by increasing the arc temperature through the high density of the arc and increased current, hot wire and TOPTIG techniques that increase only the amount of deposited welding consumables, and activated TIG (A-TIG) welding that increases only the penetration depth^6-9).

Among them, A-TIG welding can improve the penetration depth simply by applying the flux to the weld surface without further improvement and investment in the existing TIG welding power source and device. The increase in the penetration of A-TIG has been explained with the arc contraction mechanism in which the increase in arc density due to the contraction of the welding arc caused by the fluorine-based flux increases the arc temperature and penetration and the Maragoni effect in which penetration is improved as the sulfur (S) and oxygen (O)-based flux changes the convection phenomenon inside the weld pool from the edge of the weld pool to the center by changing the surface tension of the weld pool¹⁰⁾.

The penetration improvement effect of A-TIG can modify the V-groove welding of conventional TIG welding to I-groove and Y-groove. In the case of STS plate butt welding, welds with thicknesses of 8 and 12 mm require approximately 6 and 12 weld passes under the application of conventional TIG welding, respectively. When A-TIG welding is applied, however, welding is completed with 2 and 5 passes, respectively, thereby reducing weld passes by two to three times. The decrease in the number of passes can also reduce welding deformation. In terms of welding workability, flux coating can be easily performed with a marking pen, a spray, or a brush, and simple automation can reduce the fatigue of welders.

As penetration increases to the back surface of the weld, however, the underfill phenomenon in which the surface of the weld is lower than the surface of the base metal tends to occur. The underfill phenomenon further accelerates as the gap and height difference increase, and additional welding to fill it is required. Therefore, it is necessary to examine ways to reduce additional welding by preventing underfill¹¹⁾.

There is also a problem that slag occurs on the surface and back surface of A-TIG welds. Since conventional TIG welding does not use fluxes and active gas, slag is not generated. When slag occurs in TIG welds, it is recognized as a welding defect caused by weld surface contamination and poor welding gas purity, and it is not good in appearance. Therefore, working hours after welding are significantly increased by performing grinding and modifying welding during the inspection process. It is necessary to examine the cause of the slag in A-TIG welds and its impact on the welds, but there is still no related study.

As such, wire feeding conditions that do not cause underfill were examined using a semi-automatic cold wire device that automatically feeds welding consumables during welding. Based on this, the applicability of A- TIG welding to cryogenic STS pipes was investigated. In addition, the cause of the slag generated on the weld surface, its impact on the weld geometry, and its corrosivity were examined.

2. Experimental Method

Fig. 1 shows the schematic diagram of the welding test. A 4mm-thick STS 316L plate and a pipe (chemical components: Table 1) were used as base metal, and I-groove butt welding was performed. The 0.9mm-diameter STS 308L wire (chemical components: Table 1) was used as a welding consumable. LFX-SS7 (chemical components: Table 2), a commercial product for STS, was used as A-TIG flux. The A-TIG flux was mixed with alcohol and applied using a brush so that the weld surface could not be seen before welding. Table 3 shows the welding conditions used in the experiment of this study. For the welding current, pulse welding was applied to prevent burn-through and reduce welding deformation. To examine the effects of the welding consumable input direction and position on welding, the stick-out, which is the distance from the tungsten rod to the base metal surface, varied from 0.5 to 2.0 mm. Argon and nitrogen were used as shielding gas and back surface purging gas, respectively.

Fig. 1

Schematic diagram of welding test

Table 1

Chemical composition of base metal and weld consumable

Chemical composition (wt%)	C	Si	Mn	P	S	Cr	Ni	Mo	Fe
STS 316L (Base metal)	0.019º	0.49	1.76	0.002	0.002	17.11	10.10	2.04	Bal.
STS 308L (Weld consumable)	0.02	0.49	1.62	-	-	19.77	9.61	0.16	Bal.

Table 2

Chemical composition of A-TIG flux

Chemical composition(wt%)	TiO	Cr2O3	SiO2	Other
LFX-SS7	49.9	40.5	9.5	< 0.1

Table 3

Welding condition of automatic TIG welding

Welding current (Peak - Base)	Electrode angle	Stick-out	Welding speed	Wire feeding speed	Shielding gas
60A-200A	60º	0.5~2.0mm	20cm/min	80cm/min	20L/min

To observe the weld, stereo-microscopy (SM, Leica S8 APO), optical microscopy (OM, Zeiss Axioskop2 MAT), and scanning electron microscopy (SEM, Jeol JSM-500) were used. For the qualitative analysis of slag that occurred on the weld surface, energy dispersive X-ray spectroscopy (EDX, Oxford Link ISIS) and X- ray diffractometer (XRD, BRUKER D8 Advance A25) were used.

To examine the properties and corrosion of A-TIG welds for cryogenic STS pipes, the room-temperature tensile test, cryogenic impact test, and corrosion test were conducted and compared with manual TIG welds.

3. Experiment Results and Discussion

3.1 Review of A-TIG semi-automatic cold wire feeding conditions

3.1 Review of A-TIG semi-automatic cold wire feeding conditions

Fig. 2 shows the molten metal transfer types in semi- automatic TIG welding according to the welding wire feeding rate and wire feeding angle, which are the research results of Homma¹²⁾. Fig. 3. shows the results of observing the back bead and cross-section of the weld according to the wire feeding direction in the A-TIG process.

Fig. 2

Three types of molten metal transfer in semi-automatic TIG welding¹²⁾

Fig. 3

Weld appearance and weld profile formation of STS316L A-TIG weld by wire feeding direction and molten metal transfer

As can be seen from Fig. 2, semi-automatic TIG welding exhibits three molten metal transfer modes according to the welding wire feeding rate and wire feeding angle. When the feeding rate is high and the wire feeding angle is low, contacting transfer of Fig. 2(a) occurs, which is known to increase the probability of convex bead geometry. When the feeding rate is low and the wire feeding angle is high, the drop transfer mode of Fig. 2(c) occurs, which increases the probability of discontinuous beads. When the feeding rate and wire feeding angle are appropriate, discontinuous contacting transfer of Fig. 2(b) occurs.

When the molten metal transfer of the A-TIG semi- automatic cold wire feeding technique in Fig. 3 was observed based on this, contacting transfer and drop transfer occurred regardless of the wire feeding direction. When the back surface was observed according to the feeding direction, back beads were formed in specimens with no feeding and front feeding, but specimens with rear feeding had many areas where back beads were not formed. In other words, back beads were formed regardless of the transfer phenomenon for front feeding, but they were not formed in the case of contacting transfer for rear feeding. This appears to be because the wire affects the generation of back beads for front feeding, but the wire affects only the generation of the reinforcement of the surface bead regardless of the generation of back beads for rear feeding. For rear feeding, there were many cases where the wire did not melt when the wire feeding angle slightly deviated to the left and right from the arc center. In such cases, back beads were not generated either. In the case of front feeding, continuous welding was possible in most cases because the wire melted and entered the melt pool as long as the wire feeding angle did not significantly deviate to the left and right from the arc center. This indicates that front feeding is required to obtain stable back beads.

Fig. 4 shows the result of examining the effect of the distance between the TIG tungsten rod front end and the wire feeding point on welding stability. The distance was defined as Dw. When Dw was 1.0 mm or less, the wire came into contact with the tungsten rod during welding, causing the extinction and weakening of the arc and unmelting of the wire and base metal. This caused welding to be interrupted and the tungsten rod to be replaced.

Fig. 4

Effect of the distance between TIG tungsten tip and wire feeding point on welding stability

Therefore, it is necessary to feed the wire at a point more than 1 mm away from the tungsten rod front end in the front direction for stable A-TIG welding without interruption under automatic wire feeding.

3.2 Evaluation of the properties of STS pipe A-TIG welds

3.2 Evaluation of the properties of STS pipe A-TIG welds

STS 316L pipes (thickness: 4 mm, outer diameter: 318.5 mm) were welded using the A-TIG semi-automatic cold wire welding (Table 3) and feeding conditions (welding wire front feeding within 1 mm below the tungsten rod front end) secured through the previous plate experiment. The results are shown in Fig. 5 and 6 in comparison with the pipe welds performed through conventional manual TIG welding.

Fig. 5

Pipe girth welding test and weld profile using A-TIG process with semi-auto cold feeding

Fig. 6

Mechanical test results of manual TIG weld and A-TIG weld with semi-auto cold feeding process (YS: Yield Strength, TS: Tensile Strength, EL: Elongation, Impact: Charpy impact test, subsize W 2.5 × T 10 × L55mm)

A-TIG semi-automatic cold wire welding was performed by fixing the welding torch at the 12 o’clock direction for the pipe with the A-TIG flux (welding groove: I-groove) and rotating the pipe. It was completed within five minutes with a welding speed of 20 cm/min and one pass. On the other hand, conventional manual TIG welding was performed while the welder held the welding torch in one hand and fed the welding wire in the other hand for the pipe without A-TIG flux (welding groove: V-groove, angle: 60 degrees) rotated automatically. It was completed with a welding speed of 10 cm/min and three passes. The total welding time was 40 minutes, including the weld surface blushing time after each pass. The fact that 35 minutes were additionally required compared to A-TIG semi-automatic cold wire welding indicates the significant productivity improvement effect of A-TIG semi-automatic cold wire welding.

Fig. 5 shows the scene of A-TIG semi-automatic cold wire welding and the cross-sectional geometry of the weld. The cross-sectional geometry shows that a sound weld without underfill was formed despite the occurrence of the height difference.

Fig. 6 shows the average results of the mechanical property test for STS 316L pipe welding specimens subjected to A-TIG semi-automatic cold wire welding and manual TIG welding. For the test, three tensile specimens, three bending specimens, five impact specimens, and five corrosion specimens were prepared at each pipe weld based on ASME Section IX. The tensile and bending tests were conducted at room temperature. The impact toughness test was conducted at a cryogenic temperature of -196°C by preparing a subsize with a thickness of 2.5 mm because the base metal thickness was 4 mm.

As for the bending specimens, no crack occurred in both the A-TIG and manual TIG welding specimens. In the case of the tensile test, the A-TIG semi-automatic cold wire welding specimen showed a strength of more than 580 MPa, which was similar to the strength of the manual TIG welding specimen. For cryogenic impact toughness, both the A-TIG and manual TIG welding specimens also showed an average of 26 J or higher based on the subsize. When it was converted into the toughness value of the existing impact specimen with a thickness of 10 mm based on ASTM A370 Table 9 “Charpy V-Notch Test Acceptance Criteria for Various Sub-Size Specimens”, more than 60 J was obtained, which met the classification society standards.

Therefore, it can be said that the STS pipe A-TIG weld completed with one-pass welding under the application of semi-automatic cold wire feeding has similar mechanical properties to the specimen subjected to three-pass welding with manual TIG welding.

3.3 Analysis of the slag in the STS pipe A-TIG weld with semi-automatic cold wire feeding process

3.3 Analysis of the slag in the STS pipe A-TIG weld with semi-automatic cold wire feeding process

The heat tint caused by the welding heat occurs on the heat-affected zone surface of the STS weld. Since the tinted area is vulnerable to corrosion, it must be removed through pickling¹³⁾. As such, the A-TIG welding specimen was subjected to pickling with a mixture solution of nitric acid and hydrofluoric acid, which is commonly used in shipyards. Five specimens were prepared for both the A-TIG semi-automatic cold wire welding specimens and manual TIG welding specimens for the experiment.

Fig. 7 shows the weld surface beads before and after pickling. It can be seen that slag in black was present on the surface and back surface even after pickling. Qualitative analysis was conducted to identify the cause of the slag. Fig. 8 shows the results of analyzing components using SEM and EDX. It can be seen that the Ti component had the largest content. Fig. 9 shows the results of conducting XRD analysis for the slag. It was identified as TiO₂ or Ti₂O₃. These results indicate that the slag generated on the STS pipe A-TIG weld surface is the residue of the A-TIG flux.

Fig. 7

Pickling results of A-TIG weld with semi-auto cold wire feeding process

Fig. 8

Qualitative analysis results of slag on A-TIG weld bead with semi-auto cold feeding process

Fig. 9

XRD analysis results of slag on A-TIG weld bead with semi-auto cold feeding process

Fig. 10 shows the results of examining the bonding status and thickness to identify the slag generation status. The slag had a thickness of approximately 60 ㎛, and it tended to adhere to the surface without growing into the weld. If it had grown into the weld, it could have adversely affected the fatigue strength of the weld. Since it just adhered to the weld, it is judged that there is no impact on the fatigue strength.

Fig. 10

Thickness measurement results of slag on A-TIG weld bead with semi-auto cold feeding process

3.4 Corrosion test of the STS pipe A-TIG weld with the semi-automatic cold wire feeding process

3.4 Corrosion test of the STS pipe A-TIG weld with the semi-automatic cold wire feeding process

Fig. 11 shows the pitting corrosion test results for the A-TIG weld with the semi-automatic cold wire feeding process and the manual TIG weld. Five corrosion specimens were prepared for both welds. The corrosion test was conducted at a temperature of 10°C for 24 hours based on ASTM G48 A after pickling. When the corrosion test was conducted for the specimens with slag on the back surface of the A-TIG weld, it was found that corrosion did not occur at the position of the slag. When the weight reduction before and after the corrosion test was examined, it was found that the weight reduction was similar for both the A-TIG semi-automatic cold wire welding specimen and the manual TIG welding specimen. This indicates that the A-TIG semi-automatic cold wire welding technique does not affect corrosivity.

Fig. 11

Pitting corrosion test results of manual TIG weld and A-TIG weld with semi-auto cold feeding process

4. Conclusion

The applicability of activated tungsten inert gas (A- TIG) welding technology with a semi-automatic cold wire feeding technique to cryogenic stainless steel (STS) pipes was examined, and the following conclusions were drawn.

• 1) It is necessary to feed the wire at a point more than 1 mm away from the tungsten rod front end in the front direction for stable A-TIG welding without interruption under automatic wire feeding.

• 2) The slag generated on the A-TIG weld surface was the residue of the A-TIG flux, which had no impact on the weld geometry and pitting corrosion characteristics.

• 3) Since the STS pipe A-TIG weld that applied semi-automatic cold wire feeding showed similar tensile strength (more than 580 MPa) and cryogenic impact toughness (more than 60 J) to the manual TIG welding specimen as well as the same corrosion characteristics, it meets the classification society standards and can be applied in the field.

Acknowledgments

This work was supported by The Korea Institute for Advancement of Technology(KIAT) grant funded by the Korea Government(MOTIE).(P0020284, The Project of Smart Production Innovation for Small and Medium Shipbuilding)

REFERENCES

References

1. Korean Welding and Joining Society. Welding and Joining Handbook, III:Welding and Related Processes. Korean Welding and Joining Society; Daejeon, Korea: (2007), p. 34–35

2. S. Kou. Welding Metallurgy, 2nd Edition. Wiley Interscience; New Jersey, USA: (2003), p. 67

3. D. Y. Ryoo, K. Y. Kim, Y. D. Lee, and J. Y. Kang, Effect of Ni-content on the tensile properties, formability and weldability of type 304 stainless steel sheets used for the membrane of LNG storage tank, Korean J. Met. Mater. 39(11) (2001) 1233–1241.

4. H. K. Lee, Analysis of white spots formation on autogenous TIG welds of cryogenic Fe-36% Ni steels, Korean J. Met. Mater. 57(12) (2019) 771–777. http://dx.doi.org/10.3365/KJMM.2019.57.12.771
[Article]

5. N. H. Kang, Development of alloy design and welding technology for austenitic stainless steel, J. Korean Weld. Join. Soc. 28(1) (2010) 10–14. https://doi.org/10.5781/KWJS.2010.28.1.010
[Article]

6. C. Rosellini and L. Jarvis, The keyhole TIG welding process:a valid alternative for valuable metal joints, Proceedings of the 4^th National Welding Day, Workshop : Welding International. Genova, Switzerland(2007) 616–621. https://doi.org/10.1080/09507110802543237

7. B. Silwal and M. Santangelo, Effect of vibration and hot-wire gas tungsten arc (GTA) on the geometric shape, J. Mater. Process. Tech. 251 (2018) 138–145. https://doi.org/10.1016/j.jmatprotec.2017.08.010
[Article]

8. T. Opderbecke and S. Guiheux, TOPTIG:robotic TIG welding with integrated wire feeder, Weld. Int. 23(7) (2009) 523–529. https://doi.org/10.1080/09507110802543146
[Article]

9. B. Bonnefois, L. Coudreuse, and J. Charles, A-TIG welding of high nitrogen alloyed stainless steels:a metallurgically high-performance welding process, Weld. Int. 18(3) (2004) 208–212. https://doi.org/10.1533/wint.2004.3226
[Article]

10. K. H. Tseng and C. Y. Hsu, Performance of activated TIG process in austenitic stainless steel welds, J. Mater. Process. Tech. 211(3) (2011) 503–512. https://doi.org/10.1016/j.jmatprotec.2010.11.003
[Article]

11. I. K. Park, H. S. Ham, and S. M. Cho, The Eeffect of Arc Length and Shield Gas on Penetration Aspect Ratio in A-TIG Welding, J. Korean Weld. Join. Soc. 26(6) (2008) 42–47.
[Article]

12. S. Honma and K. Yasuda, Study of semi-automatic TIG welding, Weld. Int. 18(6) (2004) 450–455. https://doi.org/10.1533/wint.2004.3280
[Article]

13. L. H. Boulton and R. E. Avery, Heat tinted stainless steel welds-guidelines for acceptance, Stainless Steel World. (2004) 42–49.