The P92 steel has been recognized as a material for power fossil plant due to its high corrosion resistance, excellent creep rupture strength and improved oxidation resistance. These advantages allow the steam power plants to operate at higher steam temperature more than 600℃ and a steam pressure greater than 300bar. And many researchers have focused on creep property. But It is necessary to study the creep-fatigue property because power fossil plant materials are used at severe environment like quick cooling and heating.
Many researchers have reported that the decrease in fatigue life of ferritic steel like P92 is thought to be due to the surface oxidation and cavitation along prior austenite grain boundary. But most cracks after low cycle continuous fatigue and creep-fatigue test propagate through prior austenite grain boundary. Therefore, in this study to verify the cavitation mechanism at packet boundary by reason of fracture under tensile and compressive hold creep-fatigue interaction in P92 steel has been investigated.
All low cycle continuous fatigue and creep-fatigue tests are carried out in air atmosphere at 550 ~ 700℃ and strain wave form is symmetrical triangle. In order to investigate the creep damage on the high temperature, the hold times at tensile and compressive peak strain were imposed to be 10 and 30 min.
Prior austenite grain in P92 steel consists of packet, block and lath martensite. Many cavities are formed on various boundaries after creep-fatigue, especially most of cavities are located along packet boundaries. According to the results of SEM observation after creep-fatigue test, it is believed that cavitation is related with decohesion of $M_{23}C_6$ carbide from matrix.
To verify the boundary which cavitation occurred, precise polishing is performed before fracture by impact at liquid nitrogen temperature. The experimental analysis indicates that the specific boundary is packet boundary from comparison with Liquid Nitrogen Temperature (LNT) fractured surface and longitudinal metallographic surface.
Cavitation along packet boundary is occurred not only at tensile hold but also at compressive hold creep-fatigue test. From analysis of areal fraction of fracture along packet boundary, creep-fatigue life is decreased with increase of areal fraction of fracture along packet boundary and relaxed stress during hold time. It is considered that the cavity formation and growth at packet boundary is easy, because the increase of relaxed stress indicates the increase of plastic strain by creep damage.
The misorientation angles of all packet boundaries are above 25°. Especially low misorientation boundaries less than 10° are tangled as thread around cavities. Many dislocations are formed at specific region by non-uniform deformation under creep-fatigue test. The increase of dislocation density is confirmed by microstructure observation around packet boundary. It is suggested that the pile-up dislocations at packet boundary induce the large stress, therefore cavities can be formed at tip of dislocations piled up by stress concentration for creep-fatigue cycles.
