ACTIVATED FLUX TUNGSTEN INERT GAS WELDING OF DISSIMILAR STEELS

Abstract

In this work, the activated flux-tungsten inert gas (A-TIG) welding of dissimilar martensitic steel and austenitic stainless steel (ASS) was studied. This dissimilar steel combination is of interest for thermal power plant. Initially, the selection/development of activated flux and welding process parameters were established in order to achieve through thickness (8 mm) depth of penetration in single pass using A-TIG welding. The weld zone of this dissimilar steel A-TIG joint resulted in the high hardness, low ductility, and poor impact toughness; primarily owing to the presence of hard martensitic structure in the weld zone. The reasons of formation of martensite in the weld zone were identified as: (a) chemical composition limitation due to autogenous nature of welding and (b) high weld cooling rate to achieve through thickness penetration in single pass. To mitigate the issues related to: (a) formation of problematic martensite phase, (b) weld zone’s embrittlement and (c) high cooling rate; attempts have been made to further improve the A-TIG welding process to join this dissimilar steel combination by modifying the chemical composition of weld zone, reducing the welding heat input, retarding the cooling rate. The desired metallurgical changes were realized using three approaches namely ‘A-TIG welding with wire feed, ‘induction heat assisted A-TIG welding’ and ‘pulse current A-TIG welding’. Initially, the A-TIG welding of P92 steel and 304H ASS plates of thickness 8 mm was carried out with aim to select a proper flux for through thickness penetration. Nine different single component oxide fluxes were used; out of which TiO2 and SiO2 resulted in through thickness penetration (due to the dominance of reversal of Marangoni’s convection) and narrowest weld bead (due to the dominance of arc constriction), respectively. The integrity of weld joint developed using TiO2 flux was assessed using microscopy, hardness test, tensile test and impact toughness test. Microstructure study of the weld joint showed the formation of untempered martensite, polygonal austenite grains and ferrite stringers in different zones of weld joint. In tensile testing, the specimens were failed from 304H austenitic stainless steel base metal in as-welded condition. The V notch Charpy impact toughness of the weld joint was improved after post weld heat treatment. To further improve the weld bead geometry (increase in depth of penetration (d) and reduction in weld width(w)) of dissimilar P92 steel and 304H ASS A-TIG joint, SiO2 and TiO2 fluxes were mixed together in various proportions to exploite the advantage of both arc constriction and reversal of Marangoni’s convection. The experimental results indicated that binary flux (flux B) with composition 90% SiO2+10% TiO2 yields the most pronounced effect on weld bead geometry with 289.9% improvement in d value and 58.48% reduction in w in comparison to TIG (no flux) weld joint. The addition of a little quantity of TiO2 (10%) in SiO2 (90%) flux helped in achieving the narrowest arc (arc diameter: 5.40 mm), the maximum increase in arc voltage (∆V=2.14 V), and an increase in nascent oxygen amount (151 ppm) in the weld pool as compared to TIG (no flux) welding, which suggested the occurrence of both arc constriction and reversal of Marangoni’s convection. The average hardness of the weld zone was obtained as 415 HV. The developed weld joint qualified the tensile test as fracture took place from AISI 304H ASS base metal far away from weld zone. The impact toughness of weld zone was quite low. A comparative study of multipass-TIG welding and A-TIG welding of dissimilar P92 steel and 304H ASS plates was also carried out. The multipass-TIG weld joint was developed using Ni based filler wire ErNiCr-3/Inconel 82 in nine passes, whereas, in A TIG welding, TiO2 powder is used as flux to obtain through thickness penetration in single pass. The A-TIG welding process considerably lowered the angular distortion of the welded plates owing to the uniform thermal cycle in single pass. The weld zone’s microstructure of A-TIG joint was martensitic whereas, the weld zone of multipass-TIG joint was completely austenitic with solidified grain boundaries and migrated grain boundaries. The tensile test revealed better enactment of A-TIG weld joint as compared to multipass-TIG weld joint. The A-TIG weld joint failed from 304H ASS base metal whereas, the multipass-TIG weld joint failed from weld zone. Under impact loading, the multipass-TIG weld joint performed better. The cost incurred in A-TIG welding is only the 3.39% of the total cost in multipass-TIG welding for the development of same length of weld joint. A pathway is proposed to improve the strength-ductility synergy and impact toughness of dissimilar P92 steel-AISI 304H ASS A-TIG weld joint by feeding the ErNiCrMo-3 wire into the weld pool during welding. Three wire feeding configurations were studied to ensure proper melting of filler wire, continuous transfer of molten metal from filler wire to weld pool and consistent mixing of molten filler wire in the weld pool. The metal from filler wire is transferred into the weld pool in the form of ‘interrupted liquid bridge’ and ‘uninterrupted liquid bridge’. The ‘uninterrupted liquid bridge’ melting resulted in homogeneous mixing of filler wire into the weld pool. The weld joint produced using the best wire feeding configuration was characterized and compared with the weld joint developed without wire feed. A-TIG welding ‘with wire feed’ successfully modified the weld zone’s chemical composition (Creq: 15.01 and Nieq: 28.16), which altered its microstructure from hard and brittle martensite to Fe-Ni-Cr rich soft austenite. The microstructural transformation led to the improvement of ductility and impact toughness of weld joint without substantial reduction in tensile strength. The efficacy of ‘induction post-heating’ on weld thermal cycle, microstructure and mechanical properties (micro-hardness, tensile behavior and impact toughness) of dissimilar P92 steel-304H austenitic stainless steel (ASS) activated flux-tungsten inert gas (A-TIG) weld joint is discussed. Four induction coil configurations (IC25, IC50, IC75 and IC100) based on the horizontal distance (25 mm, 50 mm, 75 mm and 100 mm, respectively) between induction coil and welding torch was evaluated. The weld cooling rate reduces with increase in the distance between induction coil and welding torch which in turn resulted in weld zone’s microstructural alteration from martensite to dual phase (martensite+ferrite) and ferrite. The CCT diagram constructed for weld zone also confirmed that there were martensite and ferrite regions only. The favorable microstructural transformations lead to the reduction in micro-hardness, improvement in ductility and impact toughness of weld joint. In IC25, IC50, IC75 and IC100, the average hardness of weld zone was obtained as 317 HV, 270 HV, 228 HV and 166 HV, respectively. With increasing the distance from 25 mm to 50 mm, 75 mm and 100 mm between induction coil and welding torch, the ductility and impact toughness were also found to be increased. Pulse current activated flux-tungsten inert gas (Pulse current A-TIG) welding of dissimilar P92 steel and 316L ASS was carried out with an aim to reduce the welding heat input requires to achieve through thickness penetration in single pass. Pulse current A TIG welding successfully achieved through thickness penetration in 8 mm thick plates of P92 steel using less heat input (1.28 kJ/mm) as compared to constant current A-TIG welding (2.78 kJ/mm). During the pulse cycles, with the variation in welding arc profile the repetitive and higher arc pressure pushes the weld pool surface in downward direction, which makes the heat source and interface of force downward. This phenomenon could enhance the stirring effect driven by electromagnetic force that leads to the increase in depth of penetration. The integrity of developed weld joint was assessed in terms of metallurgical and mechanical behavior and compared with that obtained using constant current A-TIG welding. The mechanical properties were comparable to constant current A-TIG welding.