1 Introduction

Fabrication industries are facing cut throat competition from similar organizations. In order to sustain themselves against such tough competition, the fabrication industries are continuously working to improve productivity without compromising quality. The improvement in productivity may reduce the cost of assembled products thereby helping the industries to remain competitive. Along with others they are involved in fabrication of different products of thin sheets to be used in pharmaceutical, petrochemical, chemical, pulp and paper industries and in architectural work [1]. Austenitic stainless steel in the form of thin sheets is one of the most widely used materials by these industries. These require the application of assembling technique. The general assembling methods utilized are fusion welding together with gas tungsten arc (GTA) welding. But the welding of thin sheets of austenitic steel poses many problems because of comparatively higher heat input by fusion welding. The problems encountered may be melt through, distortion of joint, formation of porosity, buckling, warping and twisting, grain coarsening, evaporation of some of useful elements of coating of the sheets, variation of gap between sheets and fume generation [1, 2]. The difficulties experienced because of application of high heat input of fusion welding of GTA welding process may be reduced to minimum through newer welding process of gas tungsten arc with pulsed current (GTA-PC) where, better control over energy distribution and transfer of it to the weld pool is possible. In this process the current is continuously reciprocated between high and low value at a given frequency [3, 4]. The higher value of current designated as pulse current is to be properly set to achieve satisfactory penetration and bead profile as transfer of heat to generate the weld pool and welding primarily take place during maintenance of it. The lower value of current named as base current is kept minimum so that arc should not get extinguished and permit the weld pool to get cooled to achieve favourable microstructure and properties of weld joint [5, 6]. Additionally, due to cooling of weld during the maintenance of base current, heat is getting dissipated to base metal resulting in comparatively lower width of heat affected zone (HAZ) [7, 8]. All these may result in better profile of bead, relatively lesser heat input, association of weld joint with lower residual stresses and distortion and production of refine weld and HAZ microstructure [9]. The use gas tungsten arc welding with pulsed current for preparation of weld joint will generate further welding parameters pulse current (I p), base current (I b), pulse frequency (f) and pulse on time (t p). These extra parameters generated are interrelated to each other during welding of joint [10, 11]. The association of these extra parameters creates extra complexity in selection and application of it. This problem to some extent can be minimized by taking into consideration a summarized influence of pulse parameter termed by dimensionless factor ɸ = (I b/I p)ft b recommended earlier [12], where, t b can be obtained by expression [(1/f) − t p].

Productivity of the welding process of GTA-PC welding can be increased by enhancement of welding speed, but the enhancement of welding speed can result in decrease of heat input. Such reduction of energy input in the form of heat may not result in required melting of base material producing weld joint with the problem of insufficient penetration and lack of fusion. There may be decrease in cross section of weld and no melting of base metal will take place.

The heat input per unit length in GTA-PC may be found out with the help of the following expression [13]:

$${\text{HI}} = \frac{{\eta VI_{\text{m}} }}{S},$$
(1)

where, η, V, I m and S are arc efficiency, arc voltage, mean current and welding speed, respectively.

The value of I m may be found as [14, 15]:

$$I_{\text{m}} = \frac{{I_{\text{p}} t_{\text{p}} + I_{\text{b}} t_{\text{b}} }}{{t_{\text{p}} + t_{\text{b}} }},$$
(2)

with the help of Eq. 1, it can be said that the heat input per unit length may be enhanced by increasing the welding current in high speed welding as there are relatively less variations in η and V. Consequently, to maintain the required energy input per unit length, the welding current is increased so that proper geometry of weld bead and properties of the weld joint can be obtained. Nevertheless, the application of comparatively larger welding speed can give rise to weld joint with defects like humping, tunneling, split bead, parallel humping and undercutting that would limit the enhancement of productivity. The humping is the undulation of the weld bead which consists of hump and valley along the length of the weld [16,17,18]. Tunneling is a type of welding defect in which an open channel is formed that remains unfilled with the molten metal at the root of the weld. Split bead is the case of welding at higher speed in which the bead is split into two parallel seams which are separated by an empty channel [17, 18], whereas parallel humping is the case of split bead where the parallel seams are associated with humps and valleys [18]. In case of undercutting, the weld bead will have parallel grooves along the side. The humping can be due to premature freezing of the gouging region in welding of thin sheets [19]. It is also influenced by arc pressure and gas shear. The arc pressure and gas shear may be managed by application of suitable combination of parameters in high speed pulse current gas tungsten arc welding. Arc pressure may be also be decreased by means of tilt of the electrode through an angle [20, 21]. Pointed and fresh electrodes are associated with larger arc pressure resulting in generation of comparatively higher humping at high speed and high current suggesting the use of blunt electrode [22]. The arc pressure and humping are also governed by the type of shielding gas and its flow rate. The reduction in flow rate of shielding gas decreases the arc pressure and thereby diminishes the chances of humping at high welding speed but it may lead to inadequate shielding [19]. Further, humping may be avoided by keeping gap among plates to be joined [17]. Taking into account of all these, it can be said that the setting of proper combination of parameters in GTA-PC welding of thin sheets of austenitic stainless steel is critical in welding with high speed to achieve higher productivity where welding current is increased with speed. Thus, selection of a combination of parameters of GTA-PC has been discussed which will be able to make sound weld joint out of thin sheets of stainless steel using high welding speed and mean current without the problem of humping. The effects of welding speed and mean current on mechanical properties of the weld joints are also discussed. Dislocation density within the grains was also studied using TEM micrographs in 304 stainless steel weld joints.

2 Experimentation

2 mm thin sheets of austenitic stainless steel, ASS (304/1.4301, X5CrNi18-9) were joined in butt welds with no filler wire (autogenous welding) by using GTA-PC welding (Fig. 1). The figure also shows the welding machine and automated trolley used for moving the welding torch during preparation of weld joints. The chemical composition of 304, ASS sheets is shown in Table 1. The joints are made applying different combinations of GTA-PC pulse parameters keeping the heat input (Ω) and pulsing factor ɸ of constant value of the order of 0.71 ± 0.10 kJ/cm and 0.10 ± 0.04 keeping gap between the plates as 0 mm as depicted in Table 2. During preparation of weld joints, the sheets were kept fixed using pneumatic clamping unit to prevent distortion because of application of heat for welding and subsequent cooling. Plasma-cut copper sheet of 4 mm thickness was mounted over the grooves of laser table being used for carrying the electrical current. For welding, a power source EWM Tetrix 521 DC/AC and Tetrix 1002 DC for high current and speed capable of providing pulsed mode with direct current electrode negative polarity was used for welding with commercial software of EWM PC300.net V1.27 for the selection and use of pulse parameters. The welding torch with nozzle diameter of 8 mm was moved during production of weld joints automatically with the help of automated trolley as shown in Fig. 1. GTA-PC welding was done utilizing different parameters under shielding with a combination of argon and 25% nitrogen at a flow rate of 16 l/min. 6.4 mm diameter tungsten electrodes (with copper head) with 2% CeO2 and a cone angle of 30° and tip diameter of 0.25 mm was used.

Fig. 1
figure 1

Experimental setup for preparation of weld joints

Full size image
Table 1 Chemical composition of 304 steel obtained by optical emission spectrometer
Full size table
Table 2 Pulsed parameters used for producing the welds in 304 stainless steel sheets
Full size table

For each condition, the pulsing effect was kept same using hypothetically derived dimensionless factor ϕ = [(I b/I p)ft b] [3], which can be used to summarize the influence of pulse parameters, in GTA-PC welding process. The Ω, arc length and ɸ were maintained constant for preparation of weld joints through control of pulse parameters. In order to purge the hoses and torch and protect the electrode from oxidation a pre-flow and post-flow are provided as they cool from higher temperature of welding.

The transverse section of weld joints was polished with the use of standard metallographic procedure cut from central part of the joint. Etching was done using a combination of 7 ml HNO3, 10 ml acetic acid, 15 ml HCL and two drops of glycerol. Microstructures were examined with a LEICA DMI 5000 M optical microscope under polarized light. Studies of the characteristics of heat affected zone (HAZ) of the base metal adjoining the fusion line were done on the metallographically prepared and etched specimens with the optical microscope.

The porosity and the inclusion content of the weld were measured by image analyzer software on the polished transverse section of the weld joints before etching. The measurement was done on at least nine randomly captured images at various locations at 50× magnification.

The characterization of the microstructure of the weld at high magnification was carried out using a Phillips CM12 Transmission Electron Microscope (TEM) operating at 120 kV. For TEM studies, the specimen preparation requires mechanical grinding and twin-jet electropolishing with 5% perchloric acid, 25% glycerol and 70% ethanol mixture at −30 °C temperature at 45 V.

Macro-hardness measurement was carried out by using FEI Macro Vickers hardness testing machine. During measurement of hardness a load of 10 kg was applied for dwell time of 10 s in all three zones of weld joints of fusion zone, heat affected zone and base metal. Three weld joints were prepared for each combination of parameters. The average of at least three readings was taken to obtain a hardness value for a given combination of parameters.

Tensile testing was done on S-Series, H25K-S materials testing machine operated at constant cross-head speed with an initial strain rate of 5 × 10−4 s−1 to evaluate the strength and ductility of the weld at room temperature. The tensile specimens were machined as per ASTM E-8 specifications using wire-cut electro-discharge machining. Prior to testing, polishing was done on gauge length of the specimens to maintain a uniform width throughout the length and obtain a scratch-free surface. For each combination of parameters, six specimens were tested and the average was considered as the tensile strength.

3 Results and discussions

The weld joints were prepared using thin sheets of 304 austenitic stainless steel (ASS) for studying mechanical and metallurgical properties.

3.1 Studies on arc pressure during preparation of weld joints

The mean current, I m obtained using various settings of pulse parameters of GTA-PC welding at different welding speeds (S) for production of joints at constant Ω and ɸ of 0.71 ± 0.10 kJ/cm and 0.1 ± 0.04, respectively, are shown in Fig. 2. Similarly, the effect of I m on maximum arc pressure during welding is depicted in Fig. 3. It is understood that, I m has been enhanced with increase of S through proper setting and control of pulse parameters in order to maintain required heat input to get sound weld joint by proper fusion of base metals. It has been also observed that at lower values of S, there is sharp increase in I m and at moderate value of S, I m has been enhanced relatively, gradually. Again at higher value of S, I m enhances steeply because arc pressure increases exponentially with I m as shown in Fig. 3 and it also affects the weld pool dynamics and the geometry of weld. Therefore, at relatively higher S and I m, even with little variation in I m may result in generation of keyhole and various other defects. Considering all these, it may be said that at higher S and I m, the selection and setting of pulse parameters of GTA-PC welding process to produce weld joints without weld defects like humping, tunneling, split bead, parallel humping is complicated and difficult. The combination of various pulse parameters giving rise to suitable S and I m to produce weld joints with no such defects is shown in Table 3. It is observed that the Ω and ɸ are maintained at the order of 0.68 ± 0.14 and 0.143 ± 0.04, respectively, during welding. It is also observed that the pulse current, I p and mean current, I m are also increased with enhancement of S. These are increased to maintain the required heat input and heat transfer to weld for the fusion of the base plate.

Fig. 2
figure 2

Variation of I m with welding speed, S to produce weld joints using GTA-PC welding

Full size image
Fig. 3
figure 3

Effect of I m on maximum arc pressure during welding by GTA-PC

Full size image
Table 3 Parameters, heat input and pulsing factor used for preparation of weld joints of 304 stainless steel
Full size table

Figure 4 shows the weld joints welded using various combinations of pulse parameters of GTA-PC welding process giving rise to different I m and S as given in Table 3. It is found that all the weld joints do not have any defects of humping, tunneling, split bead, parallel humping, and the bead is showing continuity with respect to its appearance. This advantage has been achieved due to the application of the GTA-PC welding process since humping has been found to be a defect while using either high speed GTAW or GMAW [15,16,17]. Such benefits due to application of GTA-PC welding process has been achieved because of unique characteristics of it to control more precisely the weld thermal cycle and heat transfer to weld through manipulation of pulse parameters. The S and I m affects the geometrical aspects of weld joints of undercut size, reinforcement height, area of fusion zone and inclusion and porosity content. Therefore, undercut size, reinforcement height, area of fusion zone and inclusion and porosity content as a function of S and I m have been studied further for characterization of weld joint.

Fig. 4
figure 4

Weld joints welded using different combinations of pulse parameters of GTA-PC giving rise to various mean currents and welding speeds of a I m = 204 A, 2 cm/s, b I m = 400 A, 6.3 cm/s and c I m = 490 A, 9.6 cm/s

Full size image

3.2 Geometrical aspects of weld joints

The geometrical aspects of area of fusion zone, under cut size, reinforcement height, inclusion and porosity content and width of heat-affected zone (HAZ) of the weld joints prepared under varing S and I m of 2-9.6 cm/s and 204-490A at Ω and ɸ of the order of 0.68 kJ/cm and 0.143 have been depicted in Table 4. It is found that the area of fusion does not vary appreciably with S or I m averaging of the order of 0.07 ± 0.004 cm2 which might have occurred due to increment of I m with S giving rise to approximately same heat input to the weld. It is also observed that the undercut size increases and reinforcement height decreases with the enhancement of S. This has happened because with enhancement of S, I m increases and with increase of I m, the maximum arc pressure increases leading to higher under cut size and lower reinforcement height (Fig. 3). It is further observed that inclusions and porosity content of the weld joints are increased with enhancement of S. This has occured primarily due to increase of turbulences in the shielding environment and association of more entrapment of the atmospheric gases with higher S [22]. However, the width of the heat-affected zone adjacent to the fusion zone has been found to decrease with enhancement of S. This may be thought to have happened due to enhancement of I m with S and the relatively higher I m is associated with lowering of heat transfer efficiency. Therefore, in spite of keeping heat input constant, the increase of S and I m leads to lower temperature of weld pool giving rise to relatively lower HAZ.

Table 4 Geometrical data and porosity of weld joints of 304 stainless steel prepared under different combinations of pulse parameters of GTA-PC welding
Full size table

3.3 Microstructural studies of weld joints

The microstructures of different zones of weld joints of HAZ and weld prepared using various S and I m of 2-9.6 cm/s and 204-490A keeping constant Ω and ɸ of the order of 0.68 kJ/cm and 0.143 have been shown in Fig. 5. It has been observed that the microstructure is dendritic near fusion line due to high constitutional super cooling whereas, at the weld centre, it is equiaxed grains. Increasing the S leads to higher solidification rate at the fusion line leading to increase in susceptibility of dendrites formation and dendrites growth. Competitive growth of dendrites is also shown in figure below which shows the growth of favorable dendrite orientation due to higher temperature gradient perpendicular to the interface or in easy growth direction as per crystal structure of the metal.

Fig. 5
figure 5

Microstructure of different zones of weld joints of 304 stainless steel prepared with various combinations of pulse parameters of PC-GTAW using diverse I m and S of a I m = 204 A, S = 2 cm/s, b I m = 225 A, S = 2.49 cm/s, c I m = 400 A, S = 6.3 cm/s, d I m = 445 A, S = 8.77 cm/s and e I m = 490 A, S = 9.6 cm/s, respectively

Full size image

3.4 Hardness testing of weld joints

The variation of hardness of weld joints of 304 stainless steel with different S and I m of 2–9.6 cm/s and 204–490 A giving rise to approximately constant Ω and ɸ of the order of 0.68 kJ/cm and 0.143, respectively, is shown in Fig. 6. In line with earlier observation on low carbon steel weld joints, here also the hardness of weld increases with the enhancement of S and I m in spite of keeping Ω and ɸ constant. Such an increase in hardness at comparatively higher S might have occurred due to higher solidification and cooling rate resulting in generation of harder phases. It is found that the hardness at a location of weld centre is enhanced by 10–12% with increase of S from 2 to 9.6 cm/s. It is further understood that hardness increases with the distance from the weld center towards the fusion line. This may be because of changes in microstructure due to differential cooling with the distance from the weld center and availability of better heat sink.

Fig. 6
figure 6

Plot of variation in hardness with increase in welding speed and mean current from weld centre to base metal in 304 SS

Full size image

3.5 Tensile testing of weld joints

The tensile strength of weld joints of 304 stainless steel made using varying S and I m of 2–9.6 cm/s and 204–490 A with constant Ω and ɸ of the order of 0.68 kJ/cm and 0.143 is shown in Fig. 7. It is found that the failure of all weld joints takes place at the fusion line, and in those cases also where there was no undercut at the fusion line. The undercut problem is comparatively more severe in 304 stainless steel weld joints as compared to that of low carbon steel weld. It is also observed that with increase in S and I m, the tensile strength enhances from 508 to 571 MPa (12.5%) showing stronger weld. The enhancement of tensile strength with S and I m can be attributed to relatively higher solidification rate at larger S due to faster cooling and changes in the grain morphology and phase content in the weld. At the same time there can be increase in dislocation density at higher S because of rapid solidification of the weld. To further justify this, a TEM study was done on weld joints prepared at two welding speeds of 2 and 9.6 cm/s as shown in Fig. 8. It is found that the density of dislocation is relatively higher for the weld joint produced at higher S due to rapid solidification rate. Therefore, it can be said that increase of dislocation density leads to higher tensile strength.

Fig. 7
figure 7

Showing the variation in tensile strength with welding speed at various mean currents of optimized weld samples

Full size image
Fig. 8
figure 8

Transmission electron micrograph a I m = 204, S = 2 cm/s, b I m = 490, S = 9.6 cm/s showing the dislocation density variation with increased welding speed in weld metal

Full size image

4 Conclusions

Based on studies of weld joints prepared by using the GTA-PC welding process, the following conclusions can be drawn:

  1. 1.

    Pulsing of current in the GTA welding process for assembling thin sheets with high welding speeds produces weld joints without any defects of humping and other defects generally occurring at higher speeds.

  2. 2.

    The weld joints assembled with GTA-PC have superior properties. The superiority has been observed in the form of generation of weld joints with refined microstructure of weld and reduced HAZ width.

  3. 3.

    It is also possible to get enhanced mechanical properties such as high hardness and tensile strength in comparison to weld joints made at lower speed and mean current. The tensile strength is enhanced by 12.5% and hardness by 10–12% for higher speed and higher mean current.