Recommendations for analysis of stress corrosion in pipe systems exposed to thermohydraulic transients

Transient thermohydraulic events often control the design of piping systems in nuclear power plants. Water hammers due to valve closure, pressure transients caused by steam collapse and pipe break all result in structural loads that are characterised by a high frequency content. What also characterises these pressures/forces is the specific spatial and time dependence that is acting on the piping system and found in the wave propagation in the contained fluid. The aim with this project has been to develop recommendations for analysis of the stress response in piping systems subjected to thermohydraulic transients. Basis for this work is that the so called two-step-method is applied and that the structural response is calculated with modal superposition. Derived analysis criteria are based on the assumption that the associated volume strain energy in the wave propagation for the contained fluid may be well defined by a parameter, here called εPN. The stress response in the piping system is assumed to be completely determined with certain accuracy for that part of the volume strain energy in the wave propagation associated with this parameter. A comprehensive work has been done to determine the accuracy in loadings calculated with RELAP5. Properties such as period elongation and associated spurious oscillations in the pressure wave transient have been investigated. Furthermore, has the characteristics of the artificial numerical damping in RELAP5 been identified. Based on desired accuracy of the thermohydraulic analysis together with knowledge about the duration of the thermohydraulic perturbation, the lowest upper frequency limit fPipe, in the modal base that is required for the structure model is calculated. With perturbation is meant such as a valve closure. According to suggested criteria and with the upper frequency limit set, the essential parameters i) largest size of the elements in the structure model and ii) the largest applicable time step in the structural analysis are determined. The criteria given for these two essential parameters and criteria for data reduction of loading signals are shown to be reasonable. The present study shows that the expected error in calculated stress response is an order of magnitude larger than that of that of the error in the generated loading signals given a certain εPN. Investigations with the aim to determine appropriate requirements on the thermohydraulic model in generating the loading, i.e. the largest applicable value of εPN, has not been part of this project. While waiting for this investigation to be done, the recommendation in determining fPipe is that a value of εPN < 0.005 is used. Recommendations for analysis of the stress response in piping systems subjected to thermohydraulic loads are given in the report. Two new projects are also suggested. The first one includes a limited study with the aim to verify suggested requirements on the thermohydraulic model in generating the thermohydraulic loading. The second project concerns elastic-plastic analysis of piping systems subjected to dynamic loading and how to perform these analyses and evaluate the results according to ASME