OECD benchmark on thermal fatigue problem. Synthesis of the benchmark
T. Payen, S.Chapuliot, O. Ancelet, E. Mathet,
Third International Conference on Fatigue of Reactor Components, Seville, Spain, October 3-6, 2004,
Rapport DSR 33
Thermal fatigue in a fluid mixing area is a well-known phenomenon that has already been studied in the past. Generally, this phenomenon is linked to turbulent mixing of two fluids at two different temperatures and creates “elephant skin” type damage at the inner surface of the component and some cracks, which remain relatively small, compared to the thickness of the structure.
However, as was the case for a Tee junction of the French Super Phenix fast breeder reactor (chosen configuration for an international benchmark study ) and more recently for a pressurized water reactor at Civaux Nuclear Power Plant , this kind of fatigue damage can create cracks that propagate through the entire wall thickness.
CEA considers that 3D thermo mechanical loading is a major factor influencing crack propagation through the thickness. This factor is linked to the complex thermal hydraulic loading and has an impact on the stress distribution in the structure and the damage or crack propagation estimates. For this reason an R&D program, based on a test and numerical interpretations, was launched to quantify experimentally the influence of the 3D aspects on crack initiation and propagation. The main objective is to work on a configuration with a 3D thermal load easy to reproduce using numerical simulations, so that accurate mechanical studies can be carried out and the assessment methodologies validated or modified. However, due to the complexity of the problem, it appeared necessary for the CEA to hold technical discussions with international partners to compare different approaches used for mechanical assessment of this 3D configuration. Under the auspices of the OECD/NEA Committee for the Safety of Nuclear Installations (CSNI) and its Working Group on Integrity and Aging (IAGE), a benchmark on the integrity assessment under thermal fatigue loading was launched in early. Six organisations from 4 countries are contributing to this effort.
By now, the benchmark enters in its synthesis phase. Blind estimations of cracking (crack initiation on the inner surface of the mock up and crack propagation through the thickness) proposed by the participants are compared to experimental results. This article describes the different analyses made by the participants, a description of the experimental observations and the main conclusions of the benchmark.