• 한국압력기기공학회 논문집
• /
• 제13권2호
• /
• pp.60-66
• /
• 2017

The standard master curve (MC) approach has the major limitation because it is only applicable to homogeneous datasets. In nature, materials are macroscopically inhomogeneous and involve scatter of fracture toughness data due to various deterministic material inhomogeneity and random inhomogeneity. RPV(reactor pressure vessel) steel has different fracture toughness with varying distance from the inner surface of the wall due to cooling rate in manufacturing process; deterministic inhomogeneity. On the other hand, reference temperature, $T_0$, used in the evaluation of fracture toughness is acting as a random parameter in the evaluation of welding region; random inhomogeneity. In the present paper, four regions, the surface, 1/8T, 1/4T and 1/2T, were considered for fracture toughness specimens of KSNP (Korean Standard Nuclear Plant) SA508 Gr. 3 steel to investigate deterministic material inhomogeneity and random inhomogeneity. Fracture toughness tests were carried out for four regions and three test temperatures in the transition region. Fracture toughness evaluation was performed using the bimodal master curve (BMC) approach which is applicable to the inhomogeneous material. The results of the bimodal master curve analyses were compared with that of conventional master curve analyses. As a result, the bimodal master approach considering inhomogeneous materials provides better description of scatter in fracture toughness data than conventional master curve analysis. However, the difference in the $T_0$ determined by two master curve approaches was insignificant.