Improvement in Weldment of Dissimilar 9% CR Heat-Resistant Steels by Post-Weld Heat Treatment

: The effect of the post-weld heat treatment on the microstructures and mechanical properties of the dissimilar joint of G115, a novel developed martensite heat resistant steel, and CB2 steel, currently used in an ultra-super-critical power unit, was investigated. The results indicate that the quenched martensite underwent decomposition and transformation, and the amount of dislocations were sharply decreased in the weld metal after post-weld heat treatment (PWHT). Many nano-scale M 23 C 6 precipitates present in the weldment were distributed on the grain and grain boundary in a dispersed manner with PWHT. The average microhardness of the weldment decreased from about 400 HV to 265–290 HV after PWHT and only a slight decrease in the microhardness of CB2 steel was detected after PWHT at 760 °C. In contrast to the case of the as-received joint, the tensile strength of the joint was improved from 630 MPa to 694 MPa and the fracture location moved from the weld metal to the base metal after PWHT. The fracture surface consisted of a cleavage fracture mode without PWHT, whereas many dimples were observed on the fracture surface with PWHT.


Introduction
Heat resistant steels are widely used as structural materials for boilers, main steam pipes, and turbines of power plants [1,2]. In order to improve steam parameters and heat efficiency, researchers have turned their targets to higher-parameter thermal power units above 630 °C for developing highcapacity ultra-super-critical power plants [3]. At present, the upper limit of the operating temperature of the commercial heat-resistant steel is about 600 °C. Beyond the temperatures, problems in the still, such as thermally unstable microstructure, insufficient durability and insufficient environmental corrosion resistance, need to be considered [4]. Thus, the key factor restricting the construction of a 630 ℃ power station is the ability to produce heat-resistant steel in the higher temperature range of 630 °C [5,6].
The welding research on heat-resistant steel was mostly concentrated on P91 [17,18] and P92 [19,20] steels. The joints of other steels with a 9Cr-3W-3Co composition system, such as MARBN steel, have also been reported [21,22]. As G115 steel is a newly developed heat-resistant steel, which will be used in the high temperature components; so far researches on the welding of G115 steel were limited. In this study, G115 steel was selected for welding with a heat-resistant steel (CB2), still currently used in the unit. After welding, microstructure similar to that found by casting formed in the weld bead, the heat-affected zone (HAZ) that appears adjacent to the weld line, and residual stress can be produced. To reduce these influences on the joint, post-weld heat treatment (PWHT) was employed. The improvements in the microstructure and mechanical properties of the dissimilar joints of G115 and CB2 steels with PWHT are discussed in detail in this paper.

Materials and Methods
G115 and CB2 steel blanks, measuring 80 × 70 × 2 mm, were machined from as-received plates and utilized as the base metals. The G115 steel plate with dimensions of 200 × 150 × 40 mm, which was produced by China BaoWu Steel Group Corporation Limited (Shanghai, China), was in the condition of quenching and high temperature tempering. A filler wire with a diameter of 2.4 mm was used and its composition was similar to that of G115 steel. The filler wire was made by Atlantic China Welding Consumables INC (Zigong, China). The chemical composition of the substrates and filler are given in Table 1. A common welding method, tungsten inert gas (TIG) welding, was used to weld G115 and CB2 steels by a TIG welding machine (YC-300WP5HGN, Panasonnic industrial machinery Co., Ltd., Tangshan, China). Based on the results of previous research, the welding current, the wire feed speed and the welding speed were 130 A, 0.5 m/min and 0.15 m/min, respectively. The post-welding heat treatment was conducted at 700, 730 and 760 °C for 2 h in a resistance heating furnace. The temperature of the solid solution of CB2 steel was about 730 °C.
Specimens for metallographic examination were sectioned from the welded joints. Ground and polished cross sections were etched in a solution (5 g FeCl3 + 15 mL HCl + 80 mL H2O) to observe the joint microstructure. Microstructural observations were conducted using an optical microscope (OM, DM2000X, Chongqing, China) and a scanning electron microscope (SEM, TESCAN VEGA 3 LMH, Brno, Czech Republic). Morphology and precipitates were analyzed by TEM. Microhardness testing was used to determine the hardness profile of the joint region. The testing was carried out using a load of 4.9 N on a hardness tester (MH3N). The hardness testing was performed at a position 1 mm below the surface, and the distance between the hardness points was about 0.25 mm. Tensile tests were conducted by an electronic tensile test machine (AG-X 50KN, Shimadzu, Kyoto, Japan), where the loading rate was 1 mm/min. The tensile testing is shown schematically in Figure 1. Three specimens were tested for each condition. After tensile testing, the fracture surfaces of the samples were observed with SEM.     Figure 4a,d was relatively larger, due to the higher temperature effect of these areas closer to the weld line. There are a few precipitates in the substrate of G115 steel in the joint without PWHT; that is because the as-received condition of the G115 steel is different from the condition of the CB2 steel. As compared with the as-welded joint, many precipitates were observed in the HAZ for the base metal involving CB2 and G115 steel after PWHT at 730 °C for 2 h.    Figure 5, the matrix phase of the weldment was martensite, and the width of the martensite lath was approximately several hundred nanometers. The distribution of the lath martensite with a certain length was regular and directional. The quenched martensite was formed by shearing and resulted in a high defect density. A large number of dislocations and dislocation tangles were detected on and between the martensite laths. These dislocations were unevenly distributed, and the dislocation cells of the cellular substructure might be formed in some high-density dislocation regions.   Figure 3, the directionality and length of the martensite laths were clearly reduced and shortened after PWHT, which can be explained by the following three reasons: (1) Firstly, the long laths of quenched martensite are decomposed, owing to the effect of the high temperature. (2) Secondly, some martensite phases undergo recovery and recrystallization processes, and then the equiaxed crystals with low dislocation density are transformed from the lath crystals. (3) Thirdly, sub-grains are formed from remaining dislocations, through multilateralization and then gradually grow up when undergoing higher temperature for an extended long time. The dislocation density was decreased significantly, which can be attributed to the reduction and disappearance of dislocations and dislocation cells in the recovery process during PWHT. The carbides were identified as M23C6 phase by the diffraction patterns. Many carbide particle phases were formed on the grains and grain boundaries. The precipitate size ranged from approximately several nanometers to tens of nanometers, and the size of very few carbides was approximately hundreds of nanometers.  Figure 7 displays the variation of microhardness in the joint region without and with PWHT. For the as-welded joint, the microhardness of the weld metal, whose average value is about 400 HV, is significantly higher than that of the G115 steel and CB2 steel. This might be interpreted in the following four aspects: (1) Firstly, the quenched martensite transformed by rapid cooling has higher microhardness, which results in transformation hardening in the weld zone. (2) Secondly, there are no carbide precipitates in the quenched martensite matrix, and almost all carbon atoms are dissolved into the matrix in a supersaturated form. This can lead to solid solution hardening in the weld zone.

Microhardness of the Joint
(3) Thirdly, the residual stress produced by welding and stored in the weldment makes the measured microhardness values higher. (4) Finally, a large number of dislocations (Figure 5b) are produced in the rapid cooling process and during the quenched martensite formation. This might cause dislocation hardening in the weld seam. The microhardness values of the base metal areas were higher than those of the center area, because the above-mentioned four hardening effects are greater at the faster cooling rate. The microhardness of the G115 steel was similar to that of the CB2 steel, and its value was approximately 250 HV. By comparing the two cases of without and with PWHT, an obvious change in the microhardness profile is that the microhardness of the weld seam was sharply reduced to 265-290 HV after PWHT. The transformation of tempered martensite from quenched martensite, the reduction of the solid solution carbon content in martensite, the elimination of the remaining stress and the decrease and disappearance of dislocation density during PWHT can make the microhardness of the weldment notably low. As the PWHT temperature increased, the microhardness of the G115 steel side did not change significantly, which indicates that the G115 steel has good high temperature stability. However, the microhardness of the weldment and CB2 steel side was slightly reduced especially at 760 °C. The further coarsening of grains at the higher temperature may be responsible for the decrease in microhardness.  Figure 8 presents the tensile strengths of the joints without and with PWHT. The tensile strength of the joint without PWHT was about 630 MPa, and the fracture during tensile testing was located in the weld seam. Due to the large amount of fine carbides with dispersed distribution (Figure 6), the tensile strength values of the joints that underwent PWHT were notedly larger than those without PWHT, and the failure during tensile testing took place in the base metal of CB2, which indicates that the thermal strength of CB2 is lower than that of G115. As the PWHT temperature increased from 700 to 730 °C, the joint strength improved markedly, whereas the joint strength clearly decreased by further increasing the PWHT temperature from 730 to 760 °C. The reduction in the properties including microhardness and tensile strength of CB2 steel at 760 °C might be because the temperature was higher than the solid solution temperature of CB2 steel. The maximum strength of the joint that was reached was 694 MPa at the PWHT temperature of 730 °C, which was approximately 94 % of the tensile strength of the base metal (Table 2). Table 3 shows the yield strength and elongation to failure of the welded joints without and with PWHT. The yield strength values of the joints with PWHT are much larger than those without PWHT. The yield stress exhibits a change that firstly increases and then decreases as the temperature of PWHT enhances, which was similar to the change in the tensile strength. The maximum yield strength obtained with PWHT at 730 °C was higher than that of the joint without PWHT by about 159 MPa. The ductility of the joint increased significantly after PWHT, and the elongation values of the joints with PWHT were an order of magnitude larger than those without PWHT.    Figure 9 represents the fracture surface images of the joints without and with PWHT of 730 °C. The fracture of the as-welded joint occurred in the weldment; the failed surface contained cleavage fracture characteristics and it is considered to be a mainly brittle fracture. In all cases investigated for PWHT, the weld seam was strengthened and CB2 steel became a weak part, causing fractures that occurred in the CB2 side. The post fracture microstructures of the joints were similar and there were many dimples on the fracture surface. The results indicate that the PWHT process improved the microstructure and mechanical properties of the weldment, thereby avoiding brittle fracture at the weld.

Conclusions
In this paper, the microstructures and mechanical properties of the weldment between G115 and CB2 steels without and with PWHT were studied. The related conclusions can be drawn as follows: 1) Without PWHT, the weld seam was mainly composed of the quenched martensite with a lath thickness of approximately several hundred nanometers Many dislocations and dislocation tangles were found on and between the martensite laths. With PWHT, the tempered martensite was transformed from the quenched martensite and the number of dislocations was drastically reduced. In addition, M23C6 secondary-phases with the size of approximately several nanometers to tens of nanometers were dispersedly distributed on the grains and grain boundaries. 2) Without PWHT, the average microhardness of the weldment (~400 HV) was evidently larger than that of the G115 steel and CB2 steel (~250 HV). With PWHT, the microhardness of the weldment decreased to 265-290 HV and the microhardness of CB2 steel was slightly reduced at 760 °C for PWHT. 3) Without PWHT, the tensile strength of the joint was about 630 MPa, and the fracture with mainly cleavage fracture characteristics occurred in the weld seam. With PWHT, the joint strength increased significantly and reached up to a maximum value (694 MPa) at 730 °C, which might result from many dispersed precipitates, and the fracture with many dimples was located in the base metal of CB2 steel.

Conflicts of Interest:
The authors declare no conflict of interest.