In this paper, for the application to the Japan Sodium-cooled Fast Reactor, JSFR, the creep-fatigue damage evaluation method is improved to consider the intermediate holding condition. The improved method is validated through both of the uni-axial and the structure model creep-fatigue tests. In these validations, the target material is 316FR steel, which is planned to use for the reactor vessel. The reactor vessel portion near the liquid sodium surface is one of the most probable points where the creep-fatigue damage is considerable. Because of the relaxation of the temperature gradient, the steady operation stress on the portion near the liquid sodium surface is less than the maximum stress in the transient. In the conventional method, in order to evaluate the creep damage conservatively, the maximum tensile value in the thermal stress transient cycle is used as the initial stress. The improved method evaluates the creep damage using the lower initial stress than the conventional method, while it has the rational margin. For the validation of the improved method, uni-axial creep-fatigue tests and structure model tests are carried out. A series of uni-axial creep-fatigue tests was carried out in the following conditions: 600 degree C testing temperature, 1% total strain range, 1 hour holding time, vacuum or air environments, and the various holding position. While the test environment affects the fatigue damage, it didn’t have significant effect on the creep damage. In the cases with high holding position, the creep damages were evaluated based on the given initial stress with high precision. In the other cases, by the assumption of the steady-stress existence, the rational margin is given for the evaluation. Furthermore, in the design stage, the evaluated creep-fatigue damage has enough margins derived from the conservative evaluation of the initial stress. The structural tests modeled the movement of the liquid sodium surface in the start-up and the shut-down stages, and the relaxation of the temperature gradient in the operation stage. In these tests, the temperature distribution was given by coolant water and an external high-frequency heating coil for the cylindrical specimen, and moved in the axial direction. In addition, the primary stress, which was caused by the weight of the reactor vessel, was given by the screw jack. As a result, using the strain range evaluated by the elastic analysis, the improved method evaluated the crack initiation life due to the creep-fatigue damage with the sufficient safety margin. In the case when the strain range was evaluated by the elastic-plastic analysis, the method predicted the crack initiation life with the good precision. While the evaluation of the crack penetration life was possible, further examination was desired for the precision improvement.
