The vibration induced fatigue failure of small-bore piping is one of the common causes of failure trouble at nuclear power plants. This failure used to be prevented by calculating and screening vibration induced stresses using the accelerations measured by portable vibrometers, which are easy to handle in the working areas. Though the conventional evaluation method for calculating the vibration induced stress of small-bore piping often adopts the single-mass model, the stresses calculated by the model may be different from the actual ones because of being too simplified. So the purpose of this study is to develop the calculation methods of vibration induced stress for the screening preventing from fatigue failure troubles of small-bore piping using portable vibrometers. Firstly, for comparatively simple small-bore piping using the mock-up model simulating actual simple small-bore piping, shaking table experiments are conducted using sine wave and the field response wave measured on-site. By comparing the vibration induced stresses measured by the strain gauges and calculated using the accelerations, at first the validity of a single-mass model was conducted, and then the evaluation of a two-mass model developed as an improvement calculation model was conducted. As results of comparison, the single-mass model was found to be useful only for screening although the calculated stresses had the deviations and the tendency of an underestimate, and the two-mass model was found to be utilized as better screening because the calculated stresses had better agreement with the measured ones. Next, for small-bore piping with typical pattern configurations consisted of several masses and supports, the model considering the supports and the center of gravity being out of pipe centerline was developed and proposed. Finally, for the more complex small-bore piping with general piping configurations consisted of many bends, branches or joints, the method based on the finite element method analysis and the values measured by a portable vibrometer was developed. In this method, the analytical model was optimized, and the stresses were obtained considering vibration modes as dynamically. Judging from the results checked by numerical analysis, this method was found to be accuracy enough to use for screening, because the analytical model was optimized smoothly and the estimated stresses became to be from 1.1 to 1.4 times to the original true ones that corresponds to the actual ones measured in site.
A 1-D analytical model for the thermally induced stresses in the mold surface during die casting
