| Author(s) |
Tailin Ren, Mohsen Mohammadijoo, J Barry Wiskel, Douglas G Ivey, Muhammad Rashid, Eric Willett, Laurie Collins, Leijun Li, Hani Henein |
| Abstract Scope |
For thick-walled microalloyed steel linepipe (> 17 mm), tandem submerged-arc welding (TSAW) is used due to its deeper penetration and greater productivity compared with conventional submerged-arc welding. The use of tandem electrodes may, however, lead to an excess in overall heat input, which can adversely impact the properties of the coarse-grained heat-affected zone (CGHAZ). Cold-wire TSAW (CWTSAW) is an improved welding process consisting of three electrodes, a lead electrode, a trail electrode, and a cold electrode without applying a current. The primary objective of using the additional cold electrode is to increase the productivity without increasing heat input. In this study, a series of CWTSAW trials were conducted on the thick-walled (19.1 mm) X70 microalloyed steel to determine the effect of cold-wire feed-rate (CWFR) on microstructure and Charpy impact energy.
Two CWTSAW welds were made using cold-wire feed-rates of 16.9 mm/s and 33.9 mm/s, and one conventional TSAW weld was made without cold-wire. All welding trails were conducted using the same nominal heat input. Following welding, sub-sized Charpy impact samples were extracted from the ¼ skelp thickness along the transverse direction to the welds. Charpy testing was conducted at temperatures ranging from -60 °C to 28 °C. Vickers micro-hardness was measured in both the weld metal and HAZ. Optical microscopy (OM) analysis of etched (4% Nital) samples was used to determine and compare the phases present in the CGHAZ. The prior austenite grain (PAG) size and the morphology (maximum length, width, area and aspect ratio) of the martensite-austenite (MA) constituents in the CGHAZ of each sample were measured using OM micrographs. The microscopic features of PAGs and MA morphology were characterized using scanning electron microscopy. Modified Picral solution (4 wt.% picric acid in ethanol mixed with a 1 ml of HCl acid) and LePera’s solution (4 wt.% picric acid in ethanol mixed with a 1wt.% sodium metabisulfite in distilled water in a 1:1 volume ratio) were used for PAG and MA analysis, respectively. OM micrographs at the same magnifications were used to generate size and number distribution plots of the MA phase for each welding condition.
The Charpy energy at -30 °C increased from 11 ± 4 J to 263 ± 16 J when the CWFR was increased from 0 mm/s to 16.9 mm/s. This was followed by a reduction in Charpy energy to 12 ± 4 J at a CWFR = 33.9 mm/s. The average Vickers micro-hardness (HV0.5) immediately adjacent to the fusion line was 237 ± 4.9 and 236 ± 1.8 at CWFR for 0 mm/s and 33.9 mm/s, respectively. A relatively low hardness (HV0.5 = 227 ± 2.8) was measured for the 16.9 mm/s CWFR weld. The microstructure in the CGHAZ was a bainite/ferrite phase mixture and was similar in all the weld samples. The average PAG size was reduced from 139 ± 9 µm to 83 ± 3 µm as the CWFR was increased from 0 mm/s to 33.9 mm/s. The individual MA area and aspect ratio exhibited complex distributions for all welds. For the 16.9 mm/s CWFR, the number of MA constituents per 40,000 µm2 was 162 and comprised an area fraction of 1.4%. Conversely, the number of MA constituents per 40,000 µm2 was 648 (7.7 % of area fraction) and 443 (4.4 % of area fraction) for the 0 mm/s and 33.9 mm/s CWFR welds, respectively. Individual MA constituents were classified as four types: island-like (aspect ratio < 4 and area < 5 µm2), coarse stringer-like (aspect ratio > 4 and area > 5 µm2), fine stringer-like (aspect ratio > 4 and area < 5 µm2) and massive (aspect ratio < 4 and area > 5 µm2). Both the 0 mm/s and 33.9 mm/s CWFR samples had a larger number of both coarse stringer-like and massive MA constituents compared with the 16.9 mm/s CWFR sample.
The lower amount of coarse stringer-like and massive types of MA regions for the 16.9 mm/s CWFR weld resulted in improved Charpy toughness at -30 °C. Two key concepts can explain the effect of CWFR on MA morphology variation and the resulting Charpy impact values. At a CWFR = 0 mm/s, there is no effect of cold wire on the nominal heat input, leading to the longer time duration above AC1 and a larger PAG size. A larger PAG size is correlated with a higher martensite start temperature, which leads to a greater likelihood for MA formation. For a CWFR = 33.9 mm/s, the PAG is the finest as the fusion of cold wire addition decreased the nominal heat input and the time spent above AC1 is the lowest. However, at the high CWFR, the faster cooling rate promotes the formation of MA. This study illustrates that to achieve good impact toughness in the CGHAZ of a CWTSAW weld, a balance is needed between the duration above AC1 and cooling rate.
Keywords: Cold-Wire Tandem Submerged-Arc Welding, Martensite-Austenite Morphology, Prior Austenite Grain, Charpy Toughness, Micro-hardness |