Legacy high-strength steel welding filler metals based on the Fe-Ni-Cr-Mo alloy system, such as MIL-120S-1 and MIL-140S-1 electrodes, have well-characterized yield strength responses to cooling rate. At slow cooling rates (e.g., noted here as less than 40-50 °F/sec at 1000°F), weld metal yield strengths sharply diminish. This effect becomes amplified as electrode classification strength increases, and has primarily been attributed to the formation of coarse bainitic microstructures during slow cooling. Sensitivity of mechanical properties to minor carbon variations in this alloys class exacerbate the difficulties with their use. As a result, achieving specification minimum (or maximum) yield strength in production welds requires narrow welding procedures and tight manufacturing controls. Implementing such restrictive measures over a range of variables that influence cooling rate common in naval construction (e.g., material thicknesses, variable preheat and interpass temperatures, heat inputs limits) has proven costly.
Over the past decade, researchers at the Naval Surface Warfare Center, Carderock Division and Carpenter Technologies have collaborated to develop a high-strength steel solid wire welding electrode for gas metal arc welding (GMAW) that is cooling rate insensitive. The design has centered on composition optimization of a martensitic Fe-10wt. % Ni steel that exhibits a stable continuous cooling transformation (CCT) behavior over a broad range of welding cooling rates.
The current work presents Fe-10Ni steel GMA weld metal strength, hardness, and toughness data as a function of weld metal cooling rate and carbon content. Charpy V-notch (Cv) impact and dynamic tear (DT) testing were used to characterize the absorbed energy-temperature transition behavior. Typical welding procedure modifications relating to position, metal transfer mode, and arc weaving were also investigated to interrogate their influence on weld metal strength and toughness. Lastly, motivated by experimental observation of low distortion in the study’s test weldments, the influence of the electrode’s martensite transformation temperature on residual stress is discussed.
A series of five weldments were fabricated using 0.75-in. thick HY-130 plate and a low carbon variant of the Fe-10Ni steel 0.045-in. diameter solid wire electrode (nominal composition, Fe-0.02C-9.5Ni-0.75Mn-0.45Si-0.65Mo-0.15V). Test assemblies featured a single-vee, 45° included angle butt joint with 0.5-in. root gap and a 0.25-in. thick HY-130 backing strap. Three of the five weldments were fabricated at select heat inputs (HI) to produce a range of slow, intermediate, and fast cooling rates (10 – 80 °F/sec. at 1000°F). All weld metal tensile, Cv, Vickers microhardness, and weld metal chemistries were measured. The fourth weldment, a duplicate of the intermediate cooling rate condition, was fabricated to characterize the weld metal’s DT energy ductile-to-brittle transition temperature (DBTT) behavior. These four welds were fabricated in the flat position using an automated GMAW system operating in spray metal transfer mode under M2 shielding gas (98% Ar / 2% O2) with a 250-300°F preheat and interpass (P/I) temperature.
The final test assembly in the series was welded to induce a fast cooling rate using a 25 kJ/in. HI and 150-175°F P/I temperature. These conditions were selected to closely mimic the MIL-140S-1 electrode high cooling rate conformance test. Welding was performed via a mechanized process using M2 shielding gas in the vertical-up position, employing a GMA pulsed spray metal transfer mode and weaving technique typical of out-of-position welding. All weld metal tensile, DT, hardness, and weld metal chemistry properties were evaluated, and microstructure analysis was performed using optical and scanning electron microscopy.
Previous work demonstrated that a high carbon (0.10 wt. %) variant of Fe-10Ni steel exhibits a stable CCT from austenite to martensite over a range of cooling rates. To characterize the influence of carbon on the filler metals phase transformation stability and hardenability, a new CCT diagram was developed using the lower carbon (0.02 wt. % carbon) electrode employed in this study. A Gleeble 3500 thermomechanical simulator was used to perform the testing. The measured martensite start temperatures were then compared to those predicted by ThermoCalc using the TCFE9 steels database.
Results and Discussion:
Experimental testing has demonstrated Fe-10Ni steel GMA weld metal deposits in HY-130 plate are capable of obtaining a consistent yield strength (125-130 ksi) and average hardness (330 Hv) over a broad range of cooling rates (10 – 80 °F/sec. at 1000°F) while maintaining good impact toughness (average Cv greater than 45 ft-lb at -60°F). This insensitivity to cooling rate, particularly with respect to yield strength, is quite different from the application-limiting behavior observed in MIL-120S-1 and MIL-140S-1 weld metals. The strength and toughness properties place the Fe-10Ni steel electrode in a currently unoccupied design space with respect to MIL-120S-1 and MIL-140S-1 classifications, making it a unique alternative for overmatched 100 ksi or undermatched 130 ksi minimum yield strength welded steel designs.
Weld metal DT testing revealed lower upper shelf energies but more gradual DBTT behavior as compared to similar strength electrodes. Fe-10Ni DT specimens exhibited high percentages of ductile shearing at lower temperatures even when the absorbed energy was low. Local brittle zones on DT fracture surfaces in spray GMA welds have been associated with as-deposited microstructure in the deep penetrating papilla regions. Arc oscillation and penetration-reducing pulsed GMAW were employed to increase the amount of weld metal transformation and reheating to improve weld metal toughness.
In addition, low distortion was observed during test weldment fabrication over a range of HI, suggesting the filler metal’s martensitic phase transformation may be inducing compressive residual stresses in the weld area. Martensitic steel filler metals based on Fe-10Ni-10Cr alloy systems have demonstrated similar behavior. Results from shallow hole drilling residual stress and all weld metal Gleeble dilatometry experiments will be presented,
Low carbon Fe-10Ni steel GMA weld metal deposits are capable of obtaining a consistent yield strength of 125-130 ksi over a broad range of cooling rates while maintaining good impact toughness. Weld metal DT testing revealed lower upper shelf energies, more gradual DBTT behavior, and higher percentages of ductile shearing at low temperature when compared to similar strength Fe-Ni-Cr-Mo electrodes. Low distortion was observed during test weldment fabrication over a range of HI, suggesting the filler metal’s martensitic phase transformation may be inducing compressive residual stresses in the weld area.