Effect of Twist Extrusion Deformation on Microstructure and Creep Property of Low Activation Martensitic Steel

Abstract

In order to improve the high temperature creep properties of low activation martensitic steel, large plastic deformation was carried out by twist extrusion process, the longterm creep properties of the samples before and after deformation were tested, and the microstructure of the samples before and after deformation and creep were studied and analyzed by OM, SEM and TEM. The results show that twist extrusion deformation can refine the martensitic laths and precipitates of low activation martensitic steel. After deformation, the martensitic laths with an average width of 340 nm are divided into 250 nm subgrain/dislocation cells. The average sizes of M23C6 phase and MX phase reduce from 137 nm and 35 nm to 112 nm and 18 nm respectively. The volume fraction and number density of M23C6 phase in undeformed samples are 4.93% and 9.91×106 mm-2 respectively, increased to 6.89% and 2.19×107 mm-2 after deformation. The long rod like M23C6 phase distributes along the original austenite grain boundary, martensitic lath boundary becomes short rod like after deformation, and is dispersed in the matrix. The dislocation density in the undeformed sample is about 2.37×1014 m-2, and it increases to 4.9×1014 m-2 after deformation. In the creep test, when the high temperature stress ratio is 0.7, the steadystate creep rate of the deformed sample is 1.56×10-4 h-1, the fracture strain is 25.76%, and the creep time is 187 h. The steadystate creep rate of the undeformed specimen is 6.24×10-4 h-1, fracture strain is 24.8%, creep time is 98.5 h. That is, under the same hightemperature stress ratio, the creep time of the deformed sample is longer than that of the undeformed sample, the steadystate creep rate is lower than that of the undeformed sample, and the fracture strain of the deformed sample is equivalent to that of the undeformed sample, indicating that the deformed sample can maintain good creep ductility while enhancing the hightemperature creep resistance. After shorttime creep with a stress ratio of 0.7, the average size of M23C6 phase in the undeformed sample is about 146 nm, the coarsening rate is 6.6%, and the volume fraction is 5.32%. The average size of M23C6 phase in the deformed sample is about 126 nm, the coarsening rate is 12.5%, and the volume fraction is 7.12%. After creep at the stress ratio of 0.65, the average size of M23C6 phase of the deformed sample increases to about 139 nm, the coarsening rate is 24.1%, and the volume fraction is only 7.79%, indicating that there is no obvious coarsening in the creep process of M23C6 phase. Through theoretical calculation, pinning force PB of M23C6 phase relative interface and the resistance PZ of MX phase relative to dislocation movement in undeformed sample are 0.38 MPa and 0.041 MPa respectively. After deformation, PB and PZ increase to 0.51 MPa and 0.065 MPa respectively. Dislocation strengthening before and after deformation σd is 165 MPa and 237 MPa respectively, which shows that dislocation strengthening plays a major role. Compared with the creep stress index of undeformed low activated martensitic steel, the stress index of deformed sample increases significantly, which is 28.3. The creep fracture surfaces of the two samples show typical ductile fracture characteristics. The difference is that the diameter of the creep fracture dimple of the undeformed sample is large and deep, and carbides can be found at the bottom of the larger dimple, while the creep fracture dimple of the deformed sample is small, uniform and shallow. Comprehensive analyses suggest that the refinement of precipitates and the increase of dislocation density in the deformed sample are conducive to reduce the creep rate and strengthening.

Keywords

• twist extrusion process
• low activation martensitic steel
• high temperature creep
• microstructure evolution