The integrity of a reactor pressure vessel (RPV) has to be ensured throughout its entire life in accordance with the applicable regulations. Typically an assessment of the RPV against brittle failure needs to be conducted by taking into account all possible loading cases. One of the most severe loading cases, which can potentially occur during the operating time, is the loss-of-coolant accident, where cold water is injected into the RPV at operating conditions. High pressure in combination with a thermal shock of the ferritic pressure vessel wall caused by the injection of cold water leads to a considerable load at the belt-line area known as Pressurized Thermal Shock (PTS).

Usually the assessment against brittle failure is based on a deterministic fracture-mechanics analysis, in which common parameters like J-integral or stress intensity factor are employed to calculate the load path for an assumed (postulated) flaw during the PTS event. As an alternative to this standard approach a fracture mechanics assessments based on eXtended Finite Element Method (XFEM) approach can be performed.

The most important input data for the fracture-mechanics analysis is the transient thermal-hydraulics (TH) load of the RPV during the emergency cooling. Such data can be calculated by analytical fluid-mixing codes verified on experiments, such as KWU-MIX, or by numerical Computational Fluid Dynamics (CFD) tools after suitable validation. In KWU-MIX, which is the standard used for TH calculations within PTS analyses, rather conservative analytical models for the quantification of mixing and, depending on the water level, condensation processes in the downcomer (including simplified stripe and plume formations) are utilized. On the contrary, the numerical CFD tools can provide best-estimate results due to the possibility to consider more realistically the stripe and plume formations as well as the geometry of the RPV in detail.

In a previous paper [1] results of standard and XFEM analyses of the RPV Gösgen 1 based on thermal-hydraulics input data from KWU-MIX were presented. This paper presents new results based on thermal-hydraulics input data from CFD. The new results are compared with those from [1] in order to show additional safety margins obtained by using thermal-hydraulics input data from CFD.

This content is only available via PDF.
You do not currently have access to this content.