INVESTIGATION INTO POSITIONAL WELDING OF STRUCTURAL STEEL USING PULSE CURRENT GMAW PROCESS

Abstract

Construction and maintenance of many conventional and critical steel structures like off-shore drilling rigs, submarine hull, storage tanks, transmission towers, reactor etc. often necessitates the use of reliable positional welding in vertical and overhead positions. The positional welding is generally carried out using shielded metal arc welding (SMAW) or GMAW process with careful selection of welding parameters to control the fluidity and size of weld pool permitting its proper retention in desired position. The growing need of quality, economy and automation in welding has been found in number of cases to replace the versatile SMAW process by continuous current GMAW process. The practice of positional welding using continuous current GMAW process is possible only in short circuit mode of metal transfer allowing a favourable low heat input. But, this mode of the process often found to produce weld defects like lack of penetration, cold lapping and spattering alongwith low process economy. Other mode of metal transfer of this process such as globular and spray transfer cannot be employed for positional welding due to directional incapability and high heat input respectively. All these difficulties and limitations can be addressed by employing pulsed current GMAW process, which offers desired spray mode of metal transfer for positional welding at lower current (as mean current), giving low heat input than that which otherwise produces short-circuit/globular mode of metal transfer in the continuous current GMAW process. In recent past, number of investigations have been reported claiming the benefit of pulsed current GMAW process in improving the weld characteristics and properties of various ferrous and non-ferrous metals. Moreover, this process is more economical and having less health hazards with immense possibility of mechanization in positional welding. However, acceptable pulse current weld ii at the time of detachment is primarily governed by the amplitude of peak current and surface tension of molten droplet at the tip of the electrode. After detachment, the droplet is accelerated along the arc column due to drag force created by high velocity of electromagnetically induced plasma at a given current density. The axial arc force towards the weld pool caused by the conical configuration of arc plasma alongwith the force due to impact of droplets on the weld pool governs the bead geometry. During pulsed or continuous current GMAW at a given mean current or welding current respectively the temperature of droplet has been reported to be comparatively lower in case of the former one. The lowering of the droplet temperature under the pulsed current in largely governed by the higher burn off rate (wire feed rate) than that of continuous current GMAW process. At a given mean current, the burn off rate is marked to be significantly influenced by the pulse parameters, which is reported to be maximum at equal pulse and pulse-off durations. The Chapter III presents the experimental procedure and scheme of welding parameters adopted in this work. The geometrical, metallurgical and mechanical testing methods for welds are also included in this chapter. The vertical-up single pass square butt positional welding of 6 mm thick structural steel plate was carried out by employing comparatively higher mean currents using 1.2 mm diameter mild steel filler wire of specification SFA-5.18-89ER70S-6 under a shielding of commercially available argon + 5% CO2 gas mixture at a flow rate of 15 Lmin-1. Prior to the preparation of weld joint, the basic relationships of (1) with the characteristics of the weld bead was studied by vertical-up bead on plate weld deposition permitting a simpler and straightforward diagnosis of thermal behaviour of depositing weld metal and its subsequent influence on weld characteristics primarily in absence of many complex functions, such as joint gap, weld groove, root fusion and heat distribution in two plates. The vertical-up bead on plate deposition was performed at the mean currents of 100, 130 and 160 A. At each level of mean current, the pulse frequency was varied to 25, 50 and 100 Hz and at each frequency level, the pulse duration was varied to 3.5, 5.5, 6.5 and 7.5 ms. The arc voltage, producing acceptable bead deposition was always kept constant at 21± 0.5 V. At any mean current, the inductance of the power source was kept at maximum during 25 and 50 Hz of pulse frequencies, but at minimum during 100 Hz of pulse frequency to control the arc stability by governing the rate of response of the power source to achieve appropriate nature of base and peak currents. At each level of mean current the welding speed was adjusted to maintain the energy input of the order of 1.5 kJ/mm. However, during bead deposition at a given mean current using any pulse frequency and pulse duration, the welding speed was kept constant. The vertical-up single pass square butt weld deposition was studied at the welding parameters same as those of bead on plate deposition, except the energy input which was kept as of the order of 2.3 kJ/mm resulting in a full penetration and filling of the groove in single pass at different pulse parameters. The welds were also produced by continuous current (0 Hz) GMAW process in short-circuit mode of metal transfer, using welding current same as the mean currents used in pulsed current process, to carry out a comparative study of the weld characteristics with those of pulsed current weld. Chapter IV deals with the analytical studies on basic characteristics of depositing weld metal such as thermal behaviour of droplet and nature of metal (droplet) transfer under different conditions of pulsation, affecting the weld geometry, microstructure and mechanical properties of a weld joint. The estimation of thermal behaviour of droplet and nature of metal transfer, has been carried out using a duly modified analytical model reported earlier. The modification has made the model more vi versatile by introducing the functions as latent heat of fusion, degree of superheat of depositing droplet and acceleration of a droplet in the electromagnetically induced plasma after its detachment from electrode tip. The estimation of thermal behaviour and nature of metal transfer has been made more realistic with respect to the earlier observations by modification of the resistive heating component of total heat generated at the tip of electrode using effective current (root mean square value of the pulse wave) instead of mean current. The analytical studies have also been carried out to develop a basic understanding over the role of pulse parameters on the mechanism of bead penetration in consideration of their influence on the arc force affecting the impact of droplets on the molten weld pool situated in vertical position. The trajectory of a droplet during its travel from tip of the electrode to weld pool in arc column has been estimated by considering its initial velocity at the time of detachment and acceleration during flight due to plasma drag force. Thus, the magnitude and direction of final velocity of a droplet at the time of its deposition has been evaluated to derive the impact of droplet using impulse momentum equation. Chapter V deals with results and discussion consists of observations on verification of estimated droplet characteristics with the reported experimental results, influence of the factor 4 on the nature of metal transfer and thermal behaviour of depositing droplet, correlation of the factor (I) with the characteristics of bead on plate weld deposit and the characteristics of the weld joint, effect of (I) on the microstructure and mechanical properties of weld joint and identification of working zone for vertical-up pulsed current GMA weld deposition as bead on plate and in producing square butt weld joint. Prior to using the mathematical expressions to characterize the depositing weld metal in this investigation, their suitability and versatility has been verified and justified by comparing the estimated results with some experimental data reported vii earlier on characteristics of various depositing ferrous and non-ferrous metals. A smooth bead on plate or square butt weld deposition has been marked to be largely dependent on the selection of pulse parameters such as Im, Ip, Ib, tp and f at a given mean current. The variation in pulse parameters affecting the amount of deposition of weld metal per unit length as well as temperature and heat content of droplets alongwith their impact on weld pool has been found to govern the surface and geometrical characteristics of weld deposit known as its penetration, height and width of reinforcement, toe angle and dilution. The heat content and degree of superheating temperature of a droplet at the time of detachment are primarily governed by a competitive role in between the heat generated at tip of the filler wire and heat consumed in melting the volume of filler wire fed to the arc. At a given mean current, both of these aspects vary with the pulse parameters. The heat generation has been found to be a function of effective current of the pulsed current waveform whereas burn-off rate varies with the effective utilization of arc heat primarily dictated by the pinch effect at tip of the filler wire to detach a droplet. Under these circumstances, the degree of superheating temperature of the droplet has been found to decrease with the increase of both the factor 4) and the mean current. The behaviour of pulsed arc, nature of transportation and thermal behaviour of droplets and weld geometry bears a good correlation with the factor 4). In case of bead on plate deposition, it has been observed that the penetration of weld bead decreases with the increase of 4, which is in agreement to the proposed mechanism of bead penetration under the influence of decreasing arc force and force due to impact of droplets on the weld pool with the increase of 4). The bead height and dilution of weld metal has been found to be primarily influenced by the amount of deposition of weld metal per unit length. It is revealed that with an increase of 4, the amount of deposition of weld metal decreases resulting into decrease in bead height and increase in dilution. viii The bead width and toe angle are found to be markedly influenced by predominant vertical downflow characteristics of weld metal under the influence of gravity. The high temperature and subsequent high fluidity of weld pool at relatively low values of 4) has been found to enhance the tendency in hump formation due to downflow of weld metal. This flow behaviour of weld metal restricts its lateral spread resulting into reduction in the bead width as well as in bead toe angle (external angle). Due to variation in height, width and depth of penetration, the geometrical characteristics defined as bead form factorand bead shape factor have been found to be enhanced with the increase of 4. The successful operation of pulsed current GMAW for deposition of bead on plate weld in vertical-up position has been analysed and found feasible within a suitable range of pulse parameters, identified as the working zone, at a narrow range of arc voltage of the order of 21 ± 0.5 V. The pulse parameters satisfying the value of [(4)x) / Ind, where x is the pulse duty cycle (tp/t) and t is equal to (tp + tb), lying in the range of about 2.81 ± 0.14 and 1.77 ± 0.07, in case of welding with and without inductance respectively, are found suitable for the bead on plate deposition in vertical-up position. In case of square butt weld deposition also, the characteristics of depositing weld metal and consequently their influence on geometry, microstructure and mechanical properties of weld joint under different conditions of pulsation have been studied and found to be well correlated with the factor 4). It has been observed that the increase in 4) significantly reduces the height and marginally reduces the width of top reinforcement of weld metal alongwith an increase in area of root reinforcement and dilution of weld metal. The variation in temperature of droplet and amount of deposition of weld metal with a change in 4) or mean current has been found to govern the weld geometry. It has also been marked that during bead on plate weld deposition, ix the smooth and acceptable bead geometry has not been achieved beyond 100 A mean current. But in case of square butt weld deposition, the useful range of mean current giving acceptable weld geometry has been found to be extended upto 130 A mean current due to effective retention of liquid weld metal supported by the plates at the joint gap. Unlike that observed in case of pulsed current welding, the geometry of weld deposited by short-circuiting arc GMAW process has resulted relatively rougher surface appearance and like undercut due to incomplete fusion of the side wall at comparatively low heat of this mode of metal transfer. The microstructure of weld joint produced by short-circuiting arc GMAW process at a given energy input reveals that the variation in welding current does not significantly affect the microstructure of weld metal and heat affected zone. But in case of pulsed current welding at a given energy input, the increase in temperature of droplet with the decrease of I), has been found to coarsen the microstructure of weld metal, but at the same time, suitable interruption in weld deposition at appropriate (l) may have refined the microstructure. At a given (1) of certain order of interruption in metal deposition, the comparative increase in temperature of droplet with the decrease in mean current from 130 to 100 A has not been found to affect the microstructure of weld metal significantly, which depicts a predominant role of interruption in metal deposition on refining of microstructure under the pulsed current. The notch tensile strength of weld metal has been found to be marginally increased with the increase of (1) under a competitive effect of resulted refinement of microstructure and increased amount of proeutectoid ferrite at the grain boundaries. But the notch tensile strength of weld metal has not been significantly affected by the variation in mean current due to insignificant change of its microstructure. In the case of pulsed current welding, the microstructure and width of HAZ has been found to be coarsened and widened respectively due to increase in temperature of droplet with the decrease of (I) or mean current at a given mean current or 4) respectively. The notch tensile strength of HAZ has been found to decrease with the increase of 4) under the competitive influence of refined microstructure, increase in the amount of grain boundary ferrite and increase in hardness enhancing the notch sensitivity. The Cy—toughness of weld metal and HAZ has been found to enhance with the increase of 4) due to cumulative effect of both the refined microstructure and consequent increase in the amount of grain boundary ferrite. The Cy—toughness of weld metal deposited by short-circuiting are GMAW process has been found comparatively less than that of pulsed current welds. At a given mean current, the variation in 4) did not affect the hardness of weld deposit significantly, but the same of HAZ up to certain extent. Due to insignificant change in microstructure, the hardness of short-circuiting arc weld has not been found to markedly vary with the change in welding current. The toe angle and its microstructure has been found to be significantly influenced by the pulse parameters. The toe angle has been found to enhance with the increase of 4) under the competitive influence of the downward flow, lateral spread and amount of deposition of weld metal. The fatigue life of the joint has been estimated to reveal the influence of weld toe angle and weld form factor giving notch effect under dynamic bending load imparting tensile stress at the toe of the weld reinforcement. It has been found that the fatigue life of the joint significantly enhances with the increase of 4), weld form factor and weld toe angle. The improvement in fatigue life is highly encouraging with respect to that observed in case of short-circuiting arc weld which has been found markedly reduced due to the presence of undercut at the weld toe. In Chapter VI, the key observations of the present work has been concluded. It is revealed that the use of pulsed current GMAW process in vertical-up position welding is beneficial over the use of short-circuiting arc GMAW process due to its xi favourable nature of metal transfer and thermal behaviour. The pulsed current GMAW process offers more flexibility to control the overall behaviour of the process to achieve improved geometrical, metallurgical and mechanical properties of a weld joint. The criticality in selection of appropriate pulse parameters for the vertical-up welding has been resolved by developing the correlations of weld characteristics with the summarized influence of pulse parameter defined by the factor 4). These correlations are also analysed in the light of estimated nature of metal transfer, thermal behaviour of droplet, arc force and impact of droplets on the weld pool situated in vertical-up position. The useful working zone giving smooth and stable metal transfer with acceptable bead/weld geometry has been identified and explicitly explained. The analysis on useful working zone has revealed that the criticality in the selection of appropriate pulse parameters for smooth and acceptable weld deposit in terms of 4) may be resolved by a constant defined as [(4) x ) / Ind, where x is pulse duty cycle (tp / t) and t is equal to (tp + tb). xii