| Summary: | A specially designed, highly restrained speciamen test (60º groove angle) was successfully used to study root run weld metal solidification cracking origins and tendencies. Thirty six submerged arc welds were made on two High Yield Strangth Quenched and Tempered base steels (HY100 and Q2N0, with two low alloy wires (LINDE 95 and OERLIKON S3 NiMo 1), two basic fluxes (OP41TT and OP121TT), and six welding parameters combination at three energy inputs, viz: 1.9, 2.8 and 4.1 KJ/mm. The weld metal solidification cracking phenomenon showed up to be extremely complex and basically dependent on chemical composition, solidification (primary) structures, microsegregation, flux type, dendrites size and orientation, and base metals stress state. The most resistant welds to this defect were those made using Q2N base steel, OP121TT flux and OERLIKON S3 NiMo 1 wire. A combination of the following properties are thought to have induced this result: high carbide-to-ferrite elements forming in the base metal; flux promoting weld centerline equiaxed structure and less microsegregation; wire chemical composition, with high Mn/Si ratio, inducing also less microsegragation at the weld centerline; lower carbon and nickel combined contetens in presence of also lower phosphorus and sulphur contents. The Q2N greater ability to relieve the strain around the weld pool at high temperature is another possibility, albeit not practically demonstrated. The welding paramenters main influence on solidification cracking tendency was found to be through change in centreline solidification macrostructure and dendrite size, but the stress fields within and around the weld pool are also thought to play a significant role. The dendrite size holds a strong relationship with weld bead geometric factors, such as weld bead height-to-width ration and, principally, external area-to-perimeter ratio. The dendrite size measured on the weld bead longitudinal section must be corrected, for it depends on the angle between the dendrite growth direction and the weld bead symmetry line in a transverse section. Thus, the actual dendrite size rate of change with welding parameters is not that observed through metallographic analysis. Four types of centreline solidification macrosturctures were identified and associated with solidification cracking tendency, viz: stray, competitive columnar, centreline and equiaxed. The former three macrostructures were found to be dependent on welding paramenters, whilst the latter was promoted by the OP121TT flux. Experiments have shown that this flux releases more than twice the amount of gas(es) released by the OP41TT flux. The time available for reactions between the weld pool liquid metal and the surrounding atmosphere was evaluated through an (approximate) weld pool retention time, given as the weld ripple lag-to-welding speed amount of fused slag and deoxidants recovery. The root run welds have shown a secondary microstructure principally composed of ferrite with aligned M-A-C, acicular ferrite and martensite. No differences were detected between weld metals deposited by both wires of fluxes, being noticed the change in microstructure size only. A good correlation was found between transformation temperature and cooling time. Evidence was found of ‘cold’ or high temperature ( not solidification) cracking propagation from existing solidification cracking, and carbides segregation along the solidification cracking path. Change in the groove angle to 0º and 120º modified the general weld bead geometry and reduced the solidification cracking tendency. Cold wire addition reduced the centreline solidification cracking tendency, with no action on transverse cracks. A device was designed to make exploratory welds with wire oscillation keeping the welding head still. Low oscillation frequencies ( below 10 Hz) in the welging direction have shown to be very effective to overcome both, transverse and centreline solidification cracks. Stress relieving the base metals considerably reduced the solidification cracking tendency. The all-weld metal mechanical properties were assessed through the use of a specimen desing which allowed the obtention of practically nil diluted wels, made with 1.9 and 4.1 KJ/mm energy inputs and the two wires and fluxes combinations. The results have shown that even at the highest energy input it is possible to obtain weld metals with relatively high yield strength and toughness, using any one of the available flux/wire combinations. The results have shown that even at the highest energy input it is possible to obtain weld metals with relatively high yield strength and toughness, using any one of the available fluz/wire combinations. The results yielded not less than 600 N/mm² Yield Strength and average Charpy V-notch energy absorbed of 80 J at -40º, with transition temperature to quasicleavage fracture mode occurring below -40ºC and at or above – 80ºC. The basic techniques utilized included metallographic analysis using light microscope, transmission Electron Microscope and Scanning Electron Microscope; weld metals ‘in situ’ thermal analysis; all-weld metal mechanical tests; hardeness.
|
|---|
