Effect of Gas Pressure on the Formation Mechanism of Welds Based on Local Dry Underwater Welding

Abstract

During the local dry underwater welding process, the gas pressure in the micro drainage cover plays a key role in the formation of welds, but the relationship among the stability of the underwater arc, the formation of welds and the gas pressure has not been clarified. In this study, a micro drainage cover with dual-gas curtain based on the Lafar tube was used in the local dry underwater MIG welding experiments at a depth of 20 cm to explore the influence of gas pressure on the formation of underwater welds. The results showed that when the arc shielding gas was only 0.1 MPa in the micro drainage cover, the welding process was interrupted and the width of welds was not uniform; while the drainage gas was added, both the welding process and formation effect were improved. However, when both the shielding gas and the drainage gas pressure reached 0.4 MPa, the process and formation effect performed worse. The photos of arc combustion and metal transfer under different gas pressure were shot using high-speed photography. It turned out that the most ideal condition was both 0.2 MPa of the shielding gas pressure and drainage gas pressure at the water depth of 20 cm.

1. Introduction

With the urgent requirements for technological advancement in marine engineering and nuclear power construction, local dry underwater welding suitable for automation applications is drawing more and more attention. Comparing to the direct wet welding method, the local dry underwater method not only brings better arc stability and high quality of welds, but also has a lower cost. The drainage cover, an important device for local dry underwater welding, conducts gas or liquid to flow from externally to internally, so that a stable area similar to the onshore welding zone is formed around the welding arc, which protects arc burning, droplet transfer, and weld formation during the welding. Because the structural design and the protection mode of the drainage cover are closely related to the weld formation, scholars have done a lot of research work on the development and optimization of the drainage cover. Hamasaki et al. proposed a water curtain¹⁾ and steel brush type micro drainage cover featuring automated welding and successfully applied to butt and fillet joints at a depth of 0.3 m. Although the spatter, porosity and crack in the welding process were effectively reduced, the robustness of the system still needed to be improved. Kielczynski et al. developed a mobile gas-chamber type of drainage device,²⁾ which could separate the parts to be operated, and completed a weld to move the gas-chamber until the whole weld was completed, improving the quality of underwater welding between the terminal and the line. Rogalski et al. investigated the effects of the combination of heat input and drainage gas of the local dry underwater welding on the microstructure and hardness of the fusion zone , and established a formula to estimate the maximum hardness of HAZ (Heat affected zone) based on heat input and gas flow.³⁾

Harbin Welding Research Institute conducted a research of local dry underwater welding in the 1970s, and developed the CO₂ semi-automatic all-position welding process. Liu et al. developed a local dry underwater welding method for flux-cored wire micro-drainage cover.⁴⁾ The welds obtained from the welding experiment was free of pores, cracks and slag inclusion, and the content of diffused hydrogen in the welded joint and the highest hardness value at HAZ of welds were lower compared with wet welding. Guo et al. applied a dual-gas curtain cover to underwater laser splicing welding.⁵⁾ They studied the formation effect, porosity and mechanical properties of the welded joint under different welding parameters, and invented a welded joint with similar welding properties onshore through optimization of welding parameters. Huang et al. developed a square drainage cover.⁶⁾ They tested the drainage performance with different air intake methods, and simulated the drainage process with Fluent. Gao et al. developed a micro drainage cover with porous media baffles,⁷⁾ and studied the influence of wind field in the drainage hood on temperature field, velocity field and pressure field of arc plasma by means of numerical simulation. It was found that the lateral wind field had a greater influence on the area outside the central zone of arc column; That is, the larger the blowing angle of arc column, the greater the lateral wind speed. Finally the welding test results were proven literally consistent with the simulation results. After that, the drainage cover was redesigned to be used for underwater dry lateral welding,⁸⁾ and high-speed photography was used to study the characteristics of droplet transfer under lateral wind. However, when the existing drainage covers are being welded based on the local dry underwater welding process, the stability in the drainage covers is far from strong according to the structural design. In addition, the effect of drainage covers is determined to a large extent by the modes of gas flowing and the size of gas pressure, but there are very few researches on how gas pressure in the drainage covers are working on arc, and hence its effect on formulation of welds and microstructure evolution mechanism.

Based on the principle of Rafael tube, a dual-gas curtain micro drainage cover for local dry underwater welding was developed in this study. The outer casing of the cover was made of high-temperature-resistant flexible materials. The air intake mode of the shielding gas was the same as that of the onshore welding. The drainage gas was introduced tangentially from the intake manifold, flowed along the pipe wall to the outlet end, and formed an air curtain with a certain “robustness” at the outlet end to isolate the external water environment and carve a stable gas phase environment for the combustion of the arc underwater. As it was impossible to observe the internal environment directly, the combustion process of the onshore welding arc under the same gas flow parameters was taken as the subjects for high-speed photography to figure out the combustion state of the underwater indirectly. The research results were conductive to understanding the arc combustion stability and weld formation mechanism at different air pressure of the micro drainage cover.

2. Experiment

2.1. Experiment System

In the experiment, the 600 A all-digital inverter underwater MIG welding power supply independently developed by the laboratory was adopted to cooperate with light weight all-sealed wire feeder. During the welding process, the protective out wall of the drainage cover was attached to the workpiece. The welding gun was installed on a six-degree-of-freedom mechanical arm. The shielding gas and the drainage gas were both 98% Ar+2% CO₂ mixed gas, and were separately supplied. The planned water depth of the welding experiment was 20 cm, and the overall schematic diagram of the experiment system was shown in Fig. 1. The method of surfacing welding was adopted in the welding experiment.

Fig. 1.

Schematic diagram of local dry underwater MIG welding experiment system. (Online version in color.)

2.2. Design of Micro Drainage Cover

Drainage cover is an important device for dry underwater welding. However, there are still some problems needed to be solved in the application of drainage covers, one of which is that the welding efficiency and the protection effect cannot be effectively balanced, and hence the scope of application was greatly limited to auxiliary repairs.⁹⁾

In order to improve the efficiency and the protective effect of drainage device, a set of local dry underwater welding experiment system was developed and the drainage cover was specially designed in this study. Based on the principle of Laval tube, the dual-gas curtain structure was adopted, which means the arc shielding gas in the inner layer and drainage gas in the outer layer. The drainage gas flowed from the tangential direction into the cavity, and the sectional area was continuously decreasing along the direction leading to the outlet. According to Bernoulli’s principle, a decrease in fluid pressure came after an increase in flow velocity, so it was seen that fluid flow velocity would increase when the restricted flow passed through the tapering flow cross section. According to Han et al. the cross-sectional radius equation of the outer drainage channel was determined by the Vickers equation.¹⁰⁾ The formula is as follows:   

$${\left(\frac{d_e}{r}\right)}^2=1-\left(1-\frac{1}{c}\right)\times \left[1-{\left(\frac{x}{l}\right)}^2\right]/{\left[1+\frac{1}{3}\left(\frac{x}{l}\right)\right]}^2$$(1)

Where: d_e is the diameter of the nozzle throat, which is determined by the size of the torch; C is the shrinkage ratio; C = (d₀/d_e)²; d₀ is the outlet diameter; and l is the vertical height of the curve. The schematic cross-sectional view of the micro drainage cover is shown in Fig. 2.

Fig. 2.

Schematic diagram of the micro drainage cover. (Online version in color.)

2.3. Experiment Materials and Parameters

The material used in the experiment was AISI 304 austenitic stainless steel plate with a size of 300 × 100 × 5 mm³, using ER 308Lsi stainless steel wire with higher composition of Cr and Ni. The wire diameter was Ø1.2 mm. The chemical compositions of the base material and wire were shown in Table 1.^11,12) In order to systematically study the influence of gas pressure on droplet transfer and weld formation, the experiment parameters of local dry underwater welding were determined according to the existing welding parameters and the practical experience of the applications. To be specific, five sets of the value of shielding gas pressure and drainage gas pressure for local dry welding were designed and additionally, a set of the value for onshore welding with shielding gas of 0.1 MPa was taken as a comparison. As shown in Table 2, where Group A was the onshore welding parameters.

Table 1. Chemical composition of AISI 304 stainless steel (wt.%).

Material	C	Mn	S	Si	Cr	Ni	Mo	Fe
AISI 304	0.071	1.53	0.027	0.57	17.51	8.02	–	Bla.
ER 308Lsi	≤0.030	1.0–2.5	–	0.65–1.0	19.5–22.0	9.0–11.0	≤0.75	Bla.

Table 2. Welding experiment parameters.

Group	Background current (A)	Peak current (A)	Duty ratio	Shielding pressure (MPa)	Drainage pressure (MPa)	Welding speed (mm/s)	Remarks
A	70	300	30%	0.1	0	7	Onshore
B	70	300	30%	0.1	0	7	Underwater
C	70	300	30%	0.1	0.1	7	Underwater
D	70	300	30%	0.2	0.2	7	Underwater
E	70	300	30%	0.3	0.3	7	Underwater
F	70	300	30%	0.4	0.4	7	Underwater

The tensile test sample was designed comparing to the standard G/B T 228.1-2010, and the welding position of metallographic sample and tensile dimension parameters were shown in Fig. 3. The test samples were obtained by EDM, and the impurities were removed by ultrasonic cleaning machine. The metallographic samples were inlaid, grinded and polished by standard instruments. The corrosion solution was picric acid + nitric acid + ethanol. The macrostructure and microstructure of the welds were observed under Lycra stereomicroscope and metallographic microscope.

Fig. 3.

Welding direction and tensile dimension of metallographic samples and tensile samples. (Online version in color.)

3. Experiment Results and Discussions

3.1. Drainage Experiments and Results

The degree of dryness of the local dry cavity was difficult to be quantitatively evaluated. In order to study the underwater drainage condition of the micro drainage cover with dual-gas curtain, a special underwater camera was set up to observe the drainage effect on five groups of local dry underwater welding experiments, which was taken as the evaluation standard. The micro drainage cover was clung to the transparent glass plate fixed on the tank, and the underwater camera was aligned with the axis of the welding gun. When the gas was conducted into the drainage cover, gas pressure was adjusted to the corresponding parameters in the five groups’ and the underwater drainage effect was shown in Fig. 4. The drainage effect under the welding parameters of group B was shown in Fig. 4(a). When only 0.1 MPa of shielding gas was in the micro drainage cover, it can be seen that the water remaining in the micro drainage cover could still be captured by the underwater camera during a long time with the surrounding water continuously permeating through the micro drainage cover, which seriously affected the process of underwater welding. The drainage effect on the micro drainage cover of the Group C is shown in Fig. 4(b). When the drainage gas was added, it could be clearly observed that the water spiraled out from the cavity under the action of the drainage gas at the beginning of the drainage work and then disappeared quickly, indicating the better effect of the drainage work for the reason that the pressure in the shielding gas and the drainage gas was equal. Consequently, the surrounding arc combustion area was not affected by the drainage gas. Although there were still sporadic water droplets attached to the glass plate, the water seepage in the micro drainage cover was significantly enhanced. As the pressure of both the shielding gas and the drainage gas were increased to 0.2 MPa, as shown in Fig. 4(c), the time required to completely drain the water in the micro drainage cover was greatly shortened, effectively avoiding the interference of water on arc combustion. Because of the total disappearance of the water seepage phenomenon in the micro drainage cover, the drying degree of the combustion space was much higher, ensuring the smooth progress of the welding. Thereafter, the pressure in the shielding gas and the drainage gas was continuously increased, but there was no significant difference compared with to 0.2 MPa except for higher efficiency of the drainage. The details were shown in Figs. 4(d) and 4(e).

Fig. 4.

Internal drainage condition of micro drainage cover: (a), (b), (c), (d) and (e), corresponding to underwater welding parameters Group B, C, D, E and F, respectively. (Online version in color.)

3.2. Effect of Weld Formation

Corresponding welding experiments were conducted according to the parameters in Table 2, and the weld formation results were shown in Fig. 5.

Fig. 5.

The weld formation results were shown in Fig. (a), (b), (c), (d), (e) and (f), under the welding experiment parameters of Group A, B, C, D, E, F, respectively. (Online version in color.)

As shown in Fig. 5, although the welding current, voltage and welding speed were set the same, weld formation results varied due to different welding environments and different gas pressures. In local dry underwater welding experiment of Group B, when there was only 0.1 MPa shielding gas in the micro drainage cover using the same parameters as the normal onshore welding of Group A, issues emerged from the welding process such as difficulties in large-area arc initiation and arc breaking. At the same time, uneven weld width, internal porosity and poor formation appeared as the results shown in Fig. 5(b). Comparing with the onshore welding of the same parameters as shown in Fig. 5(a), the formation effect was quite distinctive. When 0.1 MPa drainage gas and 0.1 MPa shielding gas were both added, the arc interruption phenomenon repeated in the welding experiments for many times almost disappeared, and the uniformity of weld width was significantly improved. However, there were still humps in the welds and short arc interruption and re-ignition were observed during the welding process as shown in Fig. 4(c). When the shielding gas pressure and the drainage gas pressure were further increased to 0.2 MPa, the welding arc was smooth and there was no arc breaking during the welding process. As a result, the weld was uniform in width without any defect, which was shown in Fig. 5(d). When the pressure on both the shielding gas and the drainage gas was increased to 0.3 MPa, the welding process was the same as that at the gas pressure on 0.2 MPa, no abnormal phenomena such as arc breaking occurred, but the average width of the welds was significantly smaller than that of the welds in Fig. 5(d). This was in accordance with the trend as the width of the onshore welds under the same gas pressure parameters. When both the shielding gas pressure and the drainage gas pressure were set as 0.4 MPa, the arc initiation could be achieved. However, the process only last for 3 s. Although there were some arc breaking issues within about 3 s, the success rate of arc re-initiation was high and the welding process was smooth. However, after 3 s, arc extinction occurred and the welding process was interrupted. The results of the repeated experiments using the same parameters were almost the same, which the irregular appearance of welds with bead and pores as well as black oxidation could be clearly seen. In addition, more experiments with the pressure in both the shielding gas and the drainage gas exceeding 0.4 MPa were carried out within the withstand voltage threshold of the solenoid valve. The results showed that although there was an arcing signal during welding, the arc was extinguished immediately, and the welding wire was not completely melted, so that the welding work could not be completed.

With the change of gas pressure parameters, the average width and reinforcement of the welds also changed. Figure 6 showed weld penetration, weld reinforcement and their variation trend corresponding to the sets of experiment parameters.

Fig. 6.

Weld penetration, weld reinforcement and their variation trend corresponding to the series of experiment parameters. (Online version in color.)

In Fig. 6, reinforcement and penetration of the welds increased with higher gas pressure, but they were not identical. The welding results of three groups using 0.1 MPa shielding gas showed that the penetration and reinforcement under Group C welding parameters were less than those under group B, but they were almost the same as those under Group A welding parameters. The penetration and reinforcement of Group D increased obviously, making it the largest group in all groups of weld penetration and reinforcement.

As shown in Fig. 7, the width of underwater weld decreased with the increase of gas pressure. Moreover, although Group A for onshore and Group B for underwater both used 0.1 MPa shielding gas, the weld width obtained in the former was greater than that in the latter. Only after 0.1 MPa drainage gas was added in Group C, the obtained weld width was almost the same as that in Group A.

Fig. 7.

Weld width and changing trend corresponding to each group of experiment parameters.

3.3. Effect of Gas Pressure on Weld Formation

During the welding, all parameters were the same except the gas pressure, which led to the different cross-section size values of welds. It was generally believed that due to the influence of Marangoni convection, changes in surface tension caused changes in the shape of the welds, as reported by Tanaka et al.¹³⁾ and Li et al.¹⁴⁾ The typical feature of local dry underwater welding was the realization of local drainage function, which provided a stable gas phase space similar to the onshore environment for underwater welding. Under the action of micro drainage cover based on Laval tube, the gas around the arc area forms a spiral gas wall in the welding process, which can effectively isolate the external water environment. At the same time, part of the gas can also discharge the water in the slag area so as to improve the stability of the local dry underwater welding process. The drainage experiment results in Fig. 4 also illustrated this, but high quality weld formation required proper gas pressure parameters to match.

When only 0.1 MPa shielding gas was used, the results of weld formation onshore and underwater varied as shown in Figs. 5(a) and 5(b), indicating that the surrounding water had an adverse impact on the underwater welding process and weld formation. According to Fu et al.¹⁵⁾ and Zhang et al.,¹⁶⁾ as Group B only used 0.1 MPa shielding gas, the surrounding water environment directly affected the welding process the same method as described for wet welding. Compressed by the surrounding water environment, the weld penetration and reinforcement increased and the width of welds was narrowed. Therefore, the penetration and reinforcement as well as the width of the welds under Group B parameters of underwater welding were greater than those under Group A parameters of onshore welding. In addition, as the stainless steel welding wire used for underwater welding lacks a self-protecting function and hence the quality of the weld forming was not good as shown in Figs. 6 and 7. However, when both in the process of underwater welding, although both Group B and Group C used 0.1 MPa shielding gas as parameters, the weld formation showed obvious difference, which indicated that the drainage gas of 0.1 MPa under Group C parameters had played a key role in isolating the surrounding water environment as shown in Figs. 5(b) and 5(c). It was observed that the weld penetration, the reinforcement and the weld width of Group C for underwater welding were have parallels with those in Group A for onshore welding, indicating that the welding environment created by the two sets of welding parameters was almost the same. It can be seen that the drainage gas of 0.1 MPa under the parameters of Group C not only achieved better in drainage work but also caused least effect to the welding process, which was consistent with the results from the drainage experiment. From above, when the internal protection capacity of the micro drainage cover is lacked, the external water will cause detrimental effects to the stable combustion of the arc, and the surrounding water environment will lead to the emergence of pores and hydrogen induced cracks inside the welds, which deteriorated the weld formation quality.

Due to the special conditions of the underwater welding process, it was impossible to describe the underwater arc combustion situation and the droplet transition process in a direct image form. Through the drainage experiment and the data of the weld formation results of each group, it can be found that the stability of local dry cavity formed by the underwater welding process can be remained, and the drainage gas has little influence on the arc combustion region. Thus, the process of underwater arc combustion including arc combustion and droplet transition, can be indirectly reflected by high-speed photography under the same parameters from onshore welding. The lens used for high-speed photography was Nikon AF-S with an exposure time of 100 μs and an acquisition frame rate of 500 fps. To achieve sufficient depth of field, the aperture was adjusted to f32. The arc combustion and droplet transition under the corresponding parameters of onshore welding were shown in Fig. 8.

Fig. 8.

Arc combustion effect and droplet transition under various gas parameters: The internal drainage condition of the micro drainage cover under various gas parameters: Fig. (a) and (e) showed the arc combustion effect and droplet transition under the 0.1 MPa shielding gas, respectively, Fig. (b) and (f) showed the arc combustion effect and droplet transition under the 0.2 MPa shielding gas, respectively, Fig. (c) and (g) showed the arc combustion effect and droplet transition under the 0.3 MPa shielding gas, respectively, Fig. (d) and (h) showed the arc combustion effect and droplet transition under the 0.4 MPa shielding gas, respectively.

The arc combustion is a strong and lasting gas discharge phenomenon between two electrodes. From the point of the end of the welding wire to the surface of the workpiece, the temperature distribution gradually increases from the outer layer to the inner arc core. Jiang et al.¹⁷⁾ studied the characteristics of GTAW (gas tungsten arc welding) arc under high-pressure air condition, and revealed that the arc had compression phenomenon under the action of the gradually increased pressure of the shielding gas. As shown in Figs. 8(a), 8(b), 8(c) and 8(d), the temperatures of the center of the arc rises with the increase of the compression on arc combustion. And the temperatures on the workpiece also rises as a result.

It can be seen from the high-speed photography that the progress of the droplet transition also changed by altering the shielding gas pressure parameter. Figures 8(e) and 8(g) showed the progress of droplet transition produced two droplets per pulse, specifically a large droplet fell with a small droplet. Figure 8(f) showed the process of the droplet transition produced multiple droplets per pulse, and the jet transition mode was relatively distinctive. Figure 8(h) showed that the droplet transition mode was One Droplet Per Pulse, indicating the typical droplet transition characteristics, and the droplet transition frequency was relatively low. During the jet transition, the droplets dropped to the molten pool at a high speed along the axial direction of the welding wire, impacting the molten pool and increasing the weld penetration. Therefore, the penetration and the reinforcement of the welds under the Group D parameters were the largest among all of the experiment groups. Due to a smaller size of droplets than the diameter of the welding wire, the welds formation remained good with a little spatter in spite of the high speed of the droplets.

In the process of local dry underwater welding, although the double effects of drainage gas and shielding gas realized a higher drying degree, the “water film” on the workpiece cannot be completely avoided. The droplet frequency of spray transfer was high and the continuity was excellent, showing resistance against cleaning the “water film” on the workpiece, so this avoided the impact of water on the molten pool. When the droplets were large and the transition frequency was low, the “water film” on the workpiece cannot be effectively removed, and the splashes form when droplets fell down, causing interference to the arc combustion process, or even leading to arc breaking. Therefore, the higher shielding gas pressure and drainage gas pressure did not effectively improve the quality of weld formation under the Group F parameters, and on the contrary the welding process could not be carried out smoothly. In addition, when the internal pressure in the micro drainage cover infinitely increased, it not only caused the waste of gas, but also reduced the forming coefficient of the welds. The reduction in the forming coefficient led to the segregation of weld center area and consequently the accumulation of impurities, and led to the deterioration of the thermal crack resistance, which negatively affected the mechanical properties of the welds.

3.4. Mechanical Test Results of Welds

The tensile strength and variation trend of weld joints in six groups were shown in Fig. 9. The tensile strength of Group A in onshore welding was much lower than that of Group B in underwater welding. The tensile strength of Group B and Group C was almost the same, and the fracture position was not located in the welded joints. From Group D, the tensile strength of the joints began to decrease. The fracture position of most of the specimens of Group D located on the welded joints, but some specimens were close to the HAZ, which was similar to the specimens of Group A in onshore welding, and the fracture positions of Group E and Group F were located on the welded joints.

Fig. 9.

Tensile strength of welded joints in 6 groups in tension.

3.5. Microstructure of Welds

Figure 10 showed the metallography of the weld under the optical microscope with the magnification of 100.

Fig. 10.

The metallographic structure of the cross section: (a), (b), (c), (d), (e) and (f), corresponding to the welding parameters of Group A, B, C, D, E and F, respectively. (Online version in color.)

The measurement of the volume fraction of austenite and ferrite in the welds of each group also reflected the characteristics of the tensile strength of the welded joints in each group to a certain extent. Fu et al. studied the phase of AISI 304 stainless steel welds and held the belief that in the solidification structure of AISI 304 stainless steel weld, the black was δ ferrite with strip distribution, and the light was austenite matrix.¹⁸⁾ According to Di et al.,¹⁹⁾ as there were only ferrite phase and austenite phase in the weld microstructure, Image Pro was used to analyze the volume fraction of ferrite phase and austenite phase in the metallographic samples. The volume fraction of each group was shown in Table 3.

Table 3. Volume fraction of austenite and ferrite of welds microstructure in each group.

Group	δ ferrite	Austenite	Bal.
A	15.76	83.78	0.46
B	9.07	90.69	0.24
C	9.19	90.35	0.46
D	12.25	87.36	0.39
E	14.32	85.16	0.52
F	15.98	83.44	0.58

3.6. Weld Microstructure Evolution

Under different gas pressure parameters, the microstructure and mechanical properties of the welds changed significantly. The microstructure of AISI 304 stainless steel was affected by both the composition and cooling rate. However, Hu et al.²⁰⁾ thought that, in the environment with fast cooling rate, the cooling rate was the main factor affecting the microstructure formation. And the solidified phase in the cooling process was largely determined by the overall chemical composition of the alloy.

AISI 304 stainless steel is a typical multi-component alloy containing a variety of alloying elements to improve the mechanical properties and corrosion resistance of the material. According to Hunter et al.,²¹⁾ the equivalent of Cr and Ni can be used to simplify multi-component austenitic stainless steel into Fe–Cr–Ni ternary alloy, based on the ratio of Cr equivalent Cr_eq and that of Ni equivalent Ni_eq (Cr_eq/Ni_eq). The Eqs. (2) and (3) were shown as follows:   

$$\mathrm{N}{\mathrm{i}}_{\mathrm{e}\mathrm{q}}=[\mathrm{N}\mathrm{i}]+30[\mathrm{C}]+30[\mathrm{N}]+0.5[\mathrm{M}\mathrm{n}]$$(2)

  

$$\mathrm{C}{\mathrm{r}}_{\mathrm{e}\mathrm{q}}=[\mathrm{C}\mathrm{r}]+[\mathrm{M}\mathrm{o}]+1.5[\mathrm{S}\mathrm{i}]+0.5[\mathrm{N}\mathrm{b}]$$(3)

Where the concentration unit was the mass fraction, wt.%.

According to the EPMA-1600 electron probe and chemical method, the mass fraction (wt.%) of each component was 0.048C, 0.52Si, 17.26Cr, 2.36Mn, 7.82Ni, and the balance was Fe. The Ni_eq was calculated to be 10.44 wt.%, Cr_eq was 18.04 wt.%, and the ratio (Cr_eq/Ni_eq) was 1.728, located in the range of 1.48 < Cr_eq/Ni_eq < 1.95. Fu et al. investigate the effect of cooling rates on its microstructural evolution and solidification sequences between δ ferrite and austenite phases and thought that when the weld was cooled under a slow equilibrium condition,²²⁾ the solidification mode was L→L+δ→L+δ+γ→δ+γ→γ. Therefore, the initial microstructure of the weld was δ ferrite. However, the cooling process of the molten pool was accompanied by localized diffusion and non-equilibrium cooling, such as quenching during the underwater weld formation. Due to the chemical composition of the austenitic stainless steel itself, the cooling mode of the AISI 304 stainless steel weld was FA under low speed and balanced cooling conditions, which was not conducive to the formation of austenite. The austenitic stainless steel of ternary alloy was sensitive to temperature conditions. As the cooling rate increased, the solidification process accelerated. Due to the increase of initial heat extraction rate and the kinetic advantage of austenite itself, it was more favorable for the metastable austenite component to be precipitated out of the molten pool. Even when the degree of undercooling increased to a certain extent, the cooling mode determined by the chemical composition changed. The experiment results showed that the underwater welding parameters of Group B were the same as those of Group A, but the volume fraction of austenite in the weld formation of the former was significantly higher than that of the latter, and the heat conduction rate of water was 26 times that of air. The fast cooling rate during underwater welding regulated the austenite and ferrite content, strengthened the austenite precipitation, and refined the grain, so as to improve the mechanical properties of the weld joints.

According to Jiang et al., in addition to the chemical composition and cooling rate, the heat input provided by welding power also had an important influence on the microstructure evolution of the welds.²³⁾ In the same cooling environment, the volume fraction of austenite in the welds of Group B, C, D, E and F in underwater welding decreased in line with the increase of pressure of shielding gas and drainage gas, and the ferrite composition increased, as shown in Table 3. The mechanical properties of the welds tended to decrease, and the reduction in the weld formation coefficient further exacerbated the decline in mechanical properties, as shown in Fig. 8. After the arc was compressed by gas, the heat input of the base metal involved in the melting increased. Yakubtsov et al. reported that,²⁴⁾ the increase of heat input prolonged the high-temperature retention time of the welds, which resulted in the prolonged recrystallization time for austenite grains near the HAZ, and caused the formation of coarse austenite grains in the overheated zone, the increase in the HAZ, and the deterioration of mechanical properties of the welded joints. The increase of heat input ultimately affected the evolution of weld microstructure through the change of cooling rate.

4. Conclusions

The characteristics of arc combustion and droplet transition of local dry underwater welding under different drainage gas pressure and shielding gas pressure were researched. The weld structure and mechanical properties of AISI 304 stainless steel were analyzed. The results were summarized as follows:

(1) Six groups of welds under different gas pressure were experimented. It is confirmed that the gas pressure parameters played a significant role in the local dry underwater welding process. The micro drainage cover must be combined with the appropriate pressure of the shielding gas and the drainage gas to obtain a high quality weld formation. The pressure range should be 0.2–0.3 MPa, and the pressure of the shielding gas should be not less than the drainage gas pressure.

(2) The micro drainage cover is based on the Lafar tube design, and the protection and drainage functions are implemented by the dual-gas curtain structure. The drainage gas forms a spiral gas wall around the arc combustion area, but there is no interference to the arc combustion.

(3) When the gas pressure in the micro drainage cover gradually becomes larger, the arc is compressed under the action of the shielding gas and hence the heat input of the welds increases, turning the droplet transfer into large droplet transition. Accordingly, the spatter increases, which not only affects the smooth welding process, but also gradually worsens the mechanical properties of the joints.

Acknowledgments

The authors gratefully acknowledge the financial support for this work from National Natural Science Foundation of China (Grant No. 51875212), Science and Technology Planning Project of Guangdong Province (Grant No. 2018A030313192), and International Cooperation Project of Guangzhou City (Nos. 201807010035, 2019070110006).

References

• 1)   M.  Hamasaki,  J.  Sakakibara and  Y.  Arata: Met. Constr., 8 (1976), 108.
• 2)   W.  Kielczynski,  K.  Lesinski and  T.  Piatkowski: Weld. Int., 8 (1994), 257.
• 3)   G.  Rogalski,  J.  Łabanowski,  D.  Fydrych and  J.  Tomków: Pol. Marit. Res., 21 (2014), 58.
• 4)   S.  Liu,  J. G.  Zhong,  T.  Zhang,  X.  Liao and  G. R.  Wang: J. Nanchang Univ., Eng. Technol., 22 (2000), No. 2, 11 (in Chinese).
• 5)   N.  Guo,  Y. L.  Fu,  X.  Xing,  Y. K.  Liu,  S. X.  Zhao and  J. C.  Feng: J. Mater. Process. Technol., 260 (2018), 146.
• 6)   J. F.  Huang,  J. Q.  Huang,  Y. L.  Sun,  L.  Xue and  Y.  Zou: J. Shanghai Jiaotong Univ., 50 (2016), No. 10, 1622 (in Chinese).
• 7)   Y. F.  Gao and  A.  Hu: Trans. China Weld. Inst., 38 (2017), No. 8, 59 (in Chinese).
• 8)   Y. F.  Gao,  D.  Wu and  L. R.  Huang: Trans. China Weld. Inst., 40 (2019), No. 9, 82 (in Chinese).
• 9)   L. L.  Wang,  F. X.  Xie,  Y. L.  Feng and  Z. M.  Wang: ISIJ Int., 57 (2017), 203.
• 10)   L. G.  Han,  X. M.  Wu,  G. D.  Chen,  Z. M.  Wang and  W. Y.  Fan: J. Mater. Process. Technol., 268 (2019), 47.
• 11)   J.  Wen,  H. D.  Jia and  C. S.  Wang: ISIJ Int., 59 (2019), 2073.
• 12)   K. D.  Ramkumar,  P.  Mithilesh,  D.  Varun,  A. R. G.  Reddy,  N.  Arivazhagan,  S.  Narayanan and  K. G.  Kumar: ISIJ Int., 54 (2014), 900.
• 13)   M.  Tanaka and  J. J.  Lowke: J. Phys. D: Appl. Phys., 40 (2006), R1.
• 14)   Y.  Li,  Y. H.  Feng,  Y. F.  Li,  X. X.  Zhang and  C. S.  Wu: Int. J. Heat Mass Transf., 92 (2016), 628.
• 15)   Y. L.  Fu,  N.  Guo,  Y. P.  Du,  H.  Chen,  C. S.  Xu and  J. C.  Feng: J. Manuf. Process., 35 (2018), 161.
• 16)   Y.  Zhang,  C. B.  Jia,  J. X. Wang,  B.  Zhao and  C. S. Wu: Proc. Inst. Mech. Eng., Part B, 233 (2019), 1808.
• 17)   M.  Liang,  G. R.  Wang and  J. G.  Zhong: J. Mech. Eng., 43 (2007), 148.
• 18)   J. W.  Fu and  Y. S.  Yang: J. Alloy. Compd., 580 (2013), 191.
• 19)   X. J.  Di,  S. X.  Ji,  F. J.  Cheng,  D. P.  Wang and  J.  Cao: Mater. Des., 88 (2015), 505.
• 20)   Y.  Hu,  Y. H.  Shi,  K.  Sun and  X. Q.  Shen: J. Mater. Process. Technol., 264 (2019), 366.
• 21)   A.  Hunter and  M.  Ferry: Scr. Mater., 46 (2002), 253.
• 22)   J. W.  Fu,  Y. S.  Yang,  J. J.  Guo and  W. H.  Tong: Mater. Sci. Technol., 24 (2008), 941.
• 23)   L. P.  Jiang,  Z. H.  Wang,  X. D.  Jiao,  C. F.  Zhou,  X. M.  Fang and  H. X.  Ma: Trans. China Weld. Inst., 28 (2007), No. 6, 1 (in Chinese).
• 24)   I. A.  Yakubtsov,  P.  Poruks and  J. D.  Boyd: Mater. Sci. Eng. A, 480 (2008), 109.