STUDIES ON ACTIVATED FLUX GAS TUNGSTEN ARC WELDED JOINTS OF STEELS

Abstract

Almost all engineering structures need welding at least once in life time either to join the individual parts to form one as a whole or to repair during the service period. Gas tungsten arc welding (GTAW) process is widely used for the steel structures to provide the weld joints part with equal or even better structural integrity. However, constraints like low single pass penetration, uncontrolled weld penetration due to minor variation in alloying elements, and requirement of several number of passes for thick gauge sections make it less productive process. Gauge sections thicker than 3 mm need grooves, filler wire to fill the grove and multiple welding passes using GTAW process. Making of groves and multiple passes involves time and effort, and ultimately increase the production cost. So, to increase the productivity of the gas tungsten arc welding process; its limitation of the single pass depth of penetration should be increased. Flux assisted or activating flux tungsten inert gas (A-TIG) welding process is a novel approach to weld the thick gauge sections in single pass without using filler wire. In A-TIG process, a thin layer of activating flux is coated over the metal surface prior to welding using TIG welding process. Currently, A-TIG welding process is in developing stage; weldability of only few metal systems has been investigated using the A-TIG welding process. Dissimilar metal system response towards the A-TIG welding is almost unknown. Three different metal systems namely 409 ferritic stainless steel, P91 martensitic steel, and 316L austenitic stainless steel were selected as base metal for the current work. The thickness of the plate to be welded was kept 8 mm during whole work. The prime objective of the current work is to develop the A-TIG welding procedure for full penetration during similar metal welding (409 FSS and P91 steel) and dissimilar metal (P91-316L) welding. Silicon dioxide (SiO2), Titanium dioxide (TiO2), Chromium(III) oxide (Cr2O3), Molybdenum trioxide (MoO3), Cupric oxide (CuO), Nickel(II) oxide (NiO), Zinc oxide (ZnO), and Cerium dioxide (CeO2) were used as activating flux components. The work presented in the thesis cover different types of the weld joints such as (1) autogenous bead-on-plate weld, (2) autogenous butt weld joint, (3) multi-pass butt weld joint using filler wire, and (4) autogenous dissimilar butt weld joint. A water-cooled TIG torch with a standard 2% thoriated tungsten electrode rod of 2.9 mm diameter and electrode tip angle of 60° was used. A welding tractor was used to maintain the uniform welding speed and arc length throughout the weld length. A constant current power source in direct current electrode negative (DCEN) mode was used to supply the power to the v welding torch. Welding setup was same for all types of welding during the present thesis work. Intertest weld camera (EM19585) and National instruments made data acquisition system (DAQ) were used to study the arc profile and arc voltage during welding. The different aspects of the weld bead geometry such as depth of penetration (DOP), bead width (BW), depth to width ratio (D/W) and weld fusion zone area (WA) were analyzed using stereomicroscope to characterize the bead geometry. Optical microscope and field emission scanning electron microscope (FESEM) equipped with Energy-dispersive X-ray spectroscope (EDS) were used metallographic study including phase analysis and secondary particle characterizations. Vickers microhardness measurements were carried out on the transverse cross-section specimen comprising all the zones (base, HAZ, and weld metal) of the weldments using the microhardness tester. Sub-size tensile specimens according to ASTM E8M-04 were prepared to evaluate the tensile properties of the as received base metal and the A-TIG weldments. However non-standard samples were also fabricated to evaluate the tensile properties of the A-TIG weld fusion zone. Sub-size Charpy toughness test samples were prepared according to ASTM A370. Creep tests for the base metal, A-TIG and M-TIG weldments were carried out in an open air condition furnace. The creep tests were conducted in three different conditions namely Case 1 (as welded samples tested at 650 °C and 100 MPa), Case 2 (heat treated samples tested at 650 °C and 100 MPa), and Case 3 (as welded samples tested at 600 °C and 50 MPa). The potentiodynamic polarization test was performed on the welded samples at room temperature in 3.5 % NaCl solution, using a Gamry make potentiostat. The A-TIG welding of AISI 409 steel using SiO2 flux showed more depth of penetration than that of the Cr2O3. Response surface methodology (RSM) was used to study the main effects of the welding process parameters (welding current, welding speed, and flux coating density) and their interactional effects on the weld bead geometry. The depth of penetration increases with increase in welding current and flux coating density while it decreases with increase in welding speed. The bead width increases with welding current while it decreases with the increase of welding speed and flux coating density. The depth to width ratio increases with the welding current and flux coating density and it decreases with the increase of welding speed. The weld fusion zone area was also increases with welding current and decreases with welding speed. The flux coating density does not show significant effect on the weld fusion zone area within the selected range. Corrosion study of the A-TIG weld fusion zone suggested that a deteriorating effect of the activating flux which could be minimized using optimal combination of process parameters. The vi corrosion sensitivity in the A-TIG weld fusion zone reduces at high welding current and low welding speed. A-TIG weldments showed lower angular distortion than that of the multi-pass TIG (M-TIG) weldments during welding of 8 mm thick AISI 409 FSS. Fusion zone, coarse grain heat affected zone (CGHAZ), and fine grain heat affected zone (FGHAZ) were clearly differentiable in the M-TIG weldments. However, in A-TIG weldments weld fusion zone and CGHAZ were difficult to distinguish. The dendritic austenite and delta ferrite at the grain boundary were observed in the M-TIG weld fusion zone. In A-TIG weld fusion zone, delta ferrite and needle like martensite were observed. The presence of different phases such as ferrite, martensite, and austenite in the weld fusion zone was estimated with the Schaeffler diagram and the same was studied through normalized intensity ratio (NIR) technique also. The average microhardness of the A-TIG weld fusion zone was more than that of the M-TIG weld fusion zone. The distribution of the microhardness was found more uniform across A-TIG weld fusion zone as compared to M-TIG weld fusion zone. Tensile specimens of A-TIG and M-TIG weldments were fractured from the unaffected base metal. Charpy toughness of the M-TIG weldmetal (65 J), M-TIG HAZ (72 J), and A-TIG HAZ (72 J) was almost comparable with respect to the as-received base metal (76 J). However, A-TIG weld fusion zone showed significantly lower toughness (4 J) than that of the as received base metal. Post weld heat treatment of the A-TIG weldments at 760 °C for 2 h to 4 h of tempering time restored the toughness of the 409 FSS A-TIG weld fusion zone (≈ 116 J). During the creep test of the as welded samples tested at 650 °C and 100 MPa, the tertiary stage of creep deformation started first in the base metal and then in the M-TIG joint and in last A-TIG joints. However, during the creep test of the as welded samples tested at 600 °C and 50 MPa the tertiary stage of creep started first in the M-TIG weld joint and then A-TIG joint and then in the base metal. The creep tests performed at 650 °C and 100 MPa were of short duration creep tests, so it was expected that the tensile strength of the samples was played more effective role in deciding the fracture life rather than the critical creep mechanisms such as dislocation slip, grain boundary sliding and diffusional flow. For A-TIG welding of P91 steel, multicomponent activating flux comprising TiO2, SiO2, NiO, CuO, and CeO2 was prepared. A basic multicomponent flux of 35 % TiO2 + 40 % SiO2 + 15 % NiO + 10 % CuO was made first. Then CeO2 was mixed in the basic multicomponent flux in varying proportions. Five different compositions were developed by adding 0 %, 12.5 %, 25 %, 37.5 %, 50 % CeO2 in the basic multicomponent flux. It was vii observed that oxygen content in the weld fusion zone first increased with increasing ceria fraction in the flux. The maximum oxygen content [O] was observed 180 ppm at 25 % of CeO2, after that further addition of the CeO2 leads to decrease of oxygen content [O] in the weld fusion zone. Depth of penetration also initially increased with CeO2 up to 12.5 % but, after that further increase of CeO2 fraction in the activating flux reduced the depth of penetration. Moreover, the bead width was continuously increased with increase of CeO2 fraction. The feasibility of the nitrogen gas application in the shielding gas mixture and its effect on the weld bead geometry was also investigated. The depth of penetration increased by introducing nitrogen (Ar + N2) in the shielding gas as compared to that of the argon (Ar). But increase in bead width was also noticed. The increase in depth of penetration and bead width can attributed to the high heat input owing high arc voltage during Ar+N2 shielding. Unstable arc and spattering during Ar+N2 shielding were expected to disturb the molten metal flow pattern in the weld pool. Activating flux was not found to be effective to control the weld pool molten metal flow direction. Hence, wider bead width was obtained in flux coated region as compared to the uncoated region. On the basis of the outcome it could be stated that neither arc constriction nor reversal in Marangoni convection was effective in A-TIG welding under Ar+N2 gas mixture shielding. As it was found that the implementation of the nitrogen gas was not successful to achieve the full penetration with good weld quality. Additionally the shielding gas composition variation make the welding process more complicated. So, it was considered that in spite of shielding gas variation; development of new activating flux composition could be more beneficial for higher weld penetration and weld quality point of view. Therefore, efforts also have been made to investigate the effect of CeO2 and MoO3 on the P91 weld joints characteristics. Two different compositions Flux 1 and Flux 2 were prepared by adding 17.5 % CeO2 and 17.5 % MoO3 in the basic activating flux (35 % TiO2, 40 % SiO2, 15 % NiO, and 10 % CuO), respectively. Full penetration in 8 mm thick P91 plate was achieved in single pass using MoO3 modified flux. An increment of 200% and 300 % of depth of penetration was achieved using CeO2 and MoO3 modified flux as compared to without flux conventional TIG welding. The increase in depth of penetration is attributed to the (a) high heat input during A-TIG welding, (b) arc constriction and (c) reversal of Marangoni convection. Increase in arc voltage is attributed to the electrical resistivity of the activating fluxes. Lower melting and boiling points, lower Gibbs free energy, and more oxygen [O] atom per molecule of the MoO3 as compared CeO2 were viii found to be helpful to achieve more arc constriction and stronger centripetal convection (reversal in Marangoni convection) in the weld pool which results in high depth of penetration and narrow bead. Significant reduction in angular distortion was observed in Flux 1 weldments (0.78°) and Flux 2 weldments (0.12°) as compared to the conventional TIG weldments (1.96°). Asymmetric weld bead is a common problem during the dissimilar steel welding. In the P91-316L welding, shifting of the weld bead towards the P91 steel side was noticed. A new coating pattern was developed to tackle this problem. Almost symmetric weld bead was obtained across the joint center of the P91-316 weld joint by applying a newly developed flux coating pattern. Delta ferrite of vermicular shape, lath martensite and twins were witnessed in the HAZ of the 316L side. However, lath martensite along with delta ferrite was observed in the P91 side interface and in the HAZ. The average hardness of the weld fusion zone (450±12.6 HV) was more than that of the as-received P91 steel (253.6±4.8 HV) and 316L SS (213.4±4.9 HV). The microhardness in the weld fusion zone was found to increases towards the P91 side. The P91-316L A-TIG weldments fractured from the 316L side during tensile testing. Maximum impact energy was absorbed by the 316L SS side HAZ (60±3 J) followed by weld fusion zone (45±4 J) and minimum by P91 steel side HAZ (18±2 J) during the Charpy toughness test. Corrosion properties of the P91-316L A-TIG weldments were evaluated through potentiodynamic polarization test at three different locations namely 316L SS side weld fusion boundary (EC1), weld fusion zone center (EC2), and the P91 steel side weld fusion boundary (EC3). The EC1 had shown maximum corrosion potential (Ecorr= -352 mV), followed by EC2 (Ecorr= -449 mV), and then minimum by EC3 (Ecorr = -501 mV). The corrosion rate of the EC1 (CR = 2.07 mpy), EC2 (CR = 3.35 mpy) and EC3 (CR = 75.24 mpy) were fully aligned with the corrosion potentials. However, pitting potentials did not show the same trend. Maximum pitting potential (Epitt =390 mV) was shown by EC2 followed by the EC1 (Epitt =200 mV) and then EC3 (Epitt =111 mV).