Effect of reaustenization on grain refinement and mechanical properties in multipass Fe-10Ni weld metal

Completely reversed, strain-controlled, low-cycle fatigue behavior at room temperature is investigated for steels and their weld metals. Weld metal specimens were taken from multi-pass weld metal deposited by shield metal arc welding (SMAW) and gas metal arc welding (GMAW), such that their gage length consisted entirely of the weld metal. Results indicate that there is a trend toward reduction in the low-cycle fatigue life of weld metals as compared with the base metals. In low carbon steel weld metals, the tendency described above is explained in terms of local plastic strain concentration by lack of uniformity of the multi-pass weld metals. The weld metals do not have the same mechanical properties anywhere as confirmed by hardness distribution, and the fatigue crack grows preferentially through the temper softened region in the multi-pass welds. In Type 308 stainless steel weld metals, the ductility reduction causes reductions in low-cycle fatigue life. This study leads to the conclusion that fairly accurate estimates of the low-cycle fatigue life of weld metals can be obtained using Manson’s universal slope method. However, life estimates of the Type 304 stainless steel is difficult due to a lack of ductility caused by a deformation-induced martensitic transformation.

A 1 GPa class weld metal was prepared by multi-pass welding using a prototype flux-cored wire, and the correlation between microstructure and low-temperature impact toughness of the weld metal was investigated. The microstructure of the as-deposited columnar region and the equiaxed region reheated by subsequent welding passes was observed, and it was found that low-temperature transformation products such as bainite and martensite mainly developed in both regions. Since the prior austenite grains were coarse in the columnar region, the effective grain size of the final microstructure could be large; in this case, it was confirmed that the low-temperature impact toughness of the weld metal was very low. However, in the weld metal used in this study, acicular ferrite was formed inside the grains of the columnar region; the effective grain size and grain boundary misorientation distribution of the final microstructure were similar to those of the equiaxed region, and a good low-temperature impact toughness of the weld metal was obtained. Key words: High strength steel, Multi-pass weld metal, Acicular ferrite, Impact toughness

Completely reversed strain controlled low cycle fatigue properties at room temperature are investigated for structural carbon steels and weld metals. The weld metal specimens deposited by gas metal arc welding (GMAW) were taken from multi-pass weld metal such that gage length consisted entirely of the weld metal. The results indicate that there is a trend toward reductions in the low cycle fatigue life of weld metals as compared with the base metals. They are explained in terms of local plastic strain concentration by lack of uniformity of the multi-pass weld metals. The weld metals have not the same mechanical properties anywhere as confirmed by hardness distribution, and the fatigue crack grows preferentially through temper softening region of multi-pass welds. This study leads to the conclusion that fairly accurate estimates of the low cycle fatigue life of structural carbon steel weld metals are obtained by the results of base metals from the ratio of temper softening region. It is also confirmed by experiment that low cycle fatigue life of multi-pass welds can be improved significantly using half bead grinding technique and/or welding conditions.

In multi-pass welding of high manganese austenitic steel, weld metal is solidified in austenite single phase and hot cracking may occur depending on residual stress. In order to identify the cracking morphology, a fracture surface of the cracking of weld metal in multi-pass welding test was observed. The fracture surface indicated a dendritic morphology and a bunching pattern; therefore, it is considered that the crack in the multi-pass weld metal is caused by residual liquid film or remelting liquid film. Varestraint test was carried out in order to clarify the characterization of the crack, and then the fracture surface of cracks obtained by the test was observed. Transverse-type and longitudinal Varestraint test were used for getting typical fracture surfaces of solidification cracking in weld metal, and liquation cracking in reheat weld metal and heat affected zone of base metal, respectively. Comparing the fracture surfaces of cracks of Varestraint test and crack in practice, the fracture surface of the solidification cracking was most similar to the crack in multi-pass weld metal in this study. Thus, it is considered from comparing the fracture surfaces that solidification cracking might mainly occur in the multi-pass welding of the high manganese steel.

The effects of the heat input and preheat/interpass temperatures on the tensile strength and impact toughness of multipass welded weld metal were investigated and interpreted in terms of the recovery of the alloying elements and microstructure. Increases in both the heat input and preheat/interpass temperatures decreased the tensile strength of the weld metal. A lower recovery of alloying elements, especially Mn and Si, and smaller area fraction of acicular ferrite in the weld metal were observed in higher heat input welding, resulting in a lower tensile strength. In contrast, only a microstructure difference was observed at a higher preheat/interpass temperature. The impact toughness of the weld metal gradually increased with an increase in the heat input because of the lower tensile strength. However, it decreased again when the heat input was larger than 45 kJ/cm because of the much smaller area fraction of acicular ferrite. No effect of the preheat/interpass temperature on the impact toughness was observed. The formation of a weld metal heat-affect zone showed little effect on the impact toughness of the weld metal in this experiment.

Joint performances such as tensile strength and hardness in multi-pass welds are induced from both metallurgical and mechanical heterogeneity due to the difference of welding conditions. Hardness distribution in multi-pass weld metal is evaluated with a numerical simulation considering multiple heat cycles and phase transformation. Hardness of multipass weld metal is calculated with the rule of mixture by using fraction and hardness of each microstructure. In order to calculate fraction of each microstructure, CCT diagram was used. Conventional CCT diagrams of weld metal is not available even for single pass weld metal, thus new diagrams for multi-pass weld metals are created in this study. Modified diagrams for multi-pass weld metals with reheating effect were more dependent on the maximum temperature in reheating than the welding conditions. Hardness distribution is precisely predicted when the created CCT diagram for the multipass weld metal was used and the detailed calculation of weld thermal cycle is done.

The occurrence of microcracks, especially ductility-dip crack in multipass weld metal during GTAW and laser overlay welding processes of Ni-base alloy 690 was predicted by the mechanical approach. The stress/strain analysis in multipass welds was conducted using the thermo elasto-plastic finite element method. The brittle temperature range for ductility-dip cracking (DTR) of the reheated weld metal was determined by the Varestraint test. Plastic strain in the weld metal accumulated with applying the weld thermal cycle in multipass welding. The plastic strain-temperature curve in the La free weld metal did not cross the DTR in the cooling stage of GTAW process, however, it crossed the DTR in the cooling stage of reheating process by subsequent welding. On the other hand, the plastic strain-temperature curves of any weld passes in the La added weld metal did not cross the DTR. Ductility-dip cracks occurred in the La free weld metal except for the final layer, however, any ductility-dip cracks did not occur in the La added weld metal during multipass welding. It could be understood that ductility-dip crack would occur during not only single-pass welding but also multipass welding when plastic strain intersected the DTR.

Completely reversed strain controlled low cycle fatigue properties at room temperature are investigated for SUS304 austenitic stainless steel and weld metals. The metal specimens deposited by shield metal arc welding (SMAW) and gas metal arc welding (GMAW) were taken from multi-pass weld metal such that gage length consisted entirely of the weld metal. The results indicate that there is a trend toward reductions in the low cycle fatigue life of weld metals as compared with the base metals in the range of 3×102<Nf<6×104. They are explained in terms of lack of ductility by the deformation-induced martensitic transformation in SUS304. The martensite transformation is hardly induced in weld metals with 311% delta ferrite and the fatigue properties of weld metals are clearly different from the base metal (SUS304). This study leads to the conclusion that fairly accurate estimates of the low cycle fatigue life of weld metal can be obtained and the life estimation of SUS304 is impossible using Manson's universal slope method.

ABSTRACTIn this study, the relationship between impact toughness and microstructure in Cr–Mo–V multi-pass weld metals has been systematically investigated. The Charpy impact energy of two weld metals with various alloy elements increased remarkably. The primary cause of the change of impact toughness was attributed to the difference of acicular ferrite (AF) content and prior-austenite grain size, and the size and distribution of necklace martensite–austenite (M–A) constituents. With increasing Ti content, Ti-containing inclusions were increased, which resulted in an increased number of nucleation sites for AF, a change in the microstructure from allotriomorphic ferrite to AF, and refined ferrite grain size. In addition, smaller and more dispersive M–A constituents were observed in the weld metal with higher impact toughness.

The study was conducted to evaluate the impact toughness of flux-cored arc welded of SM570-TMC steel joint under different heat inputs, 0.9 kJ/mm (low heat input) and 1.6 kJ/mm (high heat input). Welding wire containing 0.4%Ni was selected on this experiment. Multi-pass welds were performed on SM570-TMC steel plate of 16 mm in thickness with a single V-groove butt joint on flat position (1G). The evaluation consists of observations on microstructure using an optical microscope and SEM-EDS, and mechanical properties including tensile, microhardness Vickers and Charpy V-notch (CVN) impact test at temperatures of 25, 0 and-20 °C. Results showed that the impact toughness of the base metal (BM) was higher than the weld metal (WM) at all test temperatures. Hardness and impact toughness of WM at low heat input was observed higher than when applied a high heat input. The welded samples at low and high heat inputs had high of tensile strength, and the fracture seemly occurs on the BM. Microstructure observation showed that at a high heat input, larger grains and microsegregation were observed. It might affect on decreasing their impact property.

The effects of both microstructure and oxygen content on the toughness of weld metals of a 950 MPa high strength steel have been investigated. The weld metals were prepared by SMAW (shielded metal arc welding), SAW (submerged arc welding), GMAW (gas metal arc welding) and GTAW (gas tungsten arc welding) processes. The microstructures of the weld metals were characterized with optical microscopy, SEM (scanning electron microscopy) and TEM (transmission electron microscopy) as well as continuous cooling transformation (CCT) diagram. The oxygen content was measured with a non-dispersive infrared absorption analyzer.The microstructures of all weld metals in as-welded state prepared by SMAW, SAW, GMAW and GTAW processes were martensitic structures, although slight differences in their morphologies were observed. The microstructures of the reheated weld metals in multi-pass weldment depended on both reheating temperatures TP and welding process. When TP≥Ac3, the reheated microstructures of SAW and GMAW weld metals consisted of fine martensitic structures, while that of SMAW was martensite with coarsened laths. The microstructure of the reheated weld metal of GTAW was lath martensite with a large amount of carbide precipitates. When TP≥Ac3, especially when Tp-Ac1, a large amount of precipitation were observed in the reheated weld metals of SAW, GMAW and GTAW, while the precipitation in that of SMAW was only slight. The toughness of the weld metals was closely related to their microstructures and oxygen contents. In particular, superior toughness could be associated with the formation of fine martensitic structures. Furthermore, the toughness of the weld metal was increased with the decrease in oxygen content, when the microstructure consisted primarily of martensite.

A large fraction of reheated weld metal is formed during multi-pass welding, which significantly affects the mechanical properties (especially toughness) of welded structures. In this study, the low-temperature toughness of the simulated reheated zone in multi-pass weld metal was evaluated and compared to that of the as-deposited zone using microstructural analyses. Two kinds of high-strength steel welds with different hardenabilities were produced by single-pass, bead-in-groove welding, and both welds were thermally cycled to peak temperatures above Ac3 using a Gleeble simulator. When the weld metals were reheated, their toughness deteriorated in response to the increase in the fraction of detrimental microstructural components, i.e., grain boundary ferrite and coalesced bainite in the weld metals with low and high hardenabilities, respectively. In addition, toughness deterioration occurred in conjunction with an increase in the effective grain size, which was attributed to the decrease in nucleation probability of acicular ferrite; the main cause for this decrease changed depending on the hardenability of the weld metal.

The effect of the addition of rare earth metals (REM) to the weld metal on the microcracking susceptibility in the multipass welds of alloy 690 was examined by using the La or Ce containing filler metals. The amounts of the La and Ce in the filler metal were varied in several levels. The microcracking susceptibility of the reheated weld metal was evaluated by the spot and transverse-Varestraint tests using pre-welded specimens made by GTAW. The augmented strain levels were varied from 0.50–2.44%. Cracks that occurred in the reheated weld metal evaluated by the Varestraint test could be classified into three types; ductility-dip, liquation and solidification cracking. The ductility-dip cracking susceptibility could be significantly improved by adding 0.01–0.025mass%REM to the weld metal. Adversely, the excessive REM addition led to the liquation and/or solidification cracking in the weld metal. Microstructural analyses revealed that phosphide and sulphide of La or Ce were formed in the REM containing weld metals, and Ni-La or Ni-Ce intermetallic compound was additionally identified in the excessively REM containing weld metals. High temperature tensile test indicated that hot ductility of the weld metal was ameliorated by adding 0.01–0.03mass%REM, implying that the ductility-dip cracking susceptibility was decreased as a result of lowering the grain boundary segregation of impurity elements such as P and S due to the scavenging effect of REM. The liquation and solidification cracking in the excessively REM containing weld metals were considered to be due to the formation of liquefiable intermetallic compounds of Ni and REM. The multipass welding test confirmed that microcracks in the multipass welds of alloy 690 were completely prevented by using the filler metals containing approx. 0.03mass%REM.

Microstructure evolution and impact toughness variation for high strength steel multi-pass weld metals with various cooling rates

The fatigue crack propagation thresholds of SAW weld metal of 25Cr2Ni2MoV simulating product of fossil and nuclear power low pressure turbine rotor at different stress ratios are tested. There is a big dispersity of the test results, even at the same stress ratio. The double logarithm curves of the fatigue crack growth rate and stress intensity factor range are researched. The difference of critical points between stable propagation region and near-threshold region in different specimens is found to be an important cause to the dispersity. Their locations in the specimens can be determined by the method of backward inference. After the observation of the microstructures around the critical points, a good correspondence between the size of prior austenite grain and the maximum size of monotonic plastic zone on the crack tip is confirmed. The difference of the critical points at the same stress ratio is caused by the inhomogeneous microstructures. So the inhomogeneous microstructures in the multi-pass and multi-layer weld metal contribute to the dispersity of the experimental threshold values.