A Nonlocal Meso-Scale Damage Model for Nuclear Graphite
Yan Ma,Yue Qian,Yuchen Hao,Musen Lin,Yue Li,Jinhua Wang,Haitao Wang
DOI: https://doi.org/10.1115/icone31-134775
2024-01-01
Abstract:Abstract Due to its excellent nuclear physical properties and high-temperature resistance, graphite is widely used in many areas of the nuclear industry. In high-temperature gas-cooled reactors (HTGRs), graphite is used not only in the core structural components as the neutron moderator and reflector, but also as the matrix material for the fuel elements. During the operation of HTGRs, the core structure and fuel element are subject to various static and dynamic loading. Therefore, it is particularly important to study the reliability evaluation of nuclear graphite for the safe operation of reactors. As a typical brittle material, the quantity, size and shape of internal defects of nuclear graphite have strong randomness, which leads to a large discrete strength of nuclear graphite. It is more reasonable to use a probabilistic method to evaluate the strength of nuclear graphite. At present, the dominant approach to the probabilistic evaluation of nuclear graphite is based on finite element methods, such as the Burchell model. However, compared with the meshless method, the finite element method make it difficult to handle the stress singularity brought by the displacement discontinuity, the computation complexity brought by the crack bifurcation and the non-locality of failure. As an emerging meshless method, Peridynamics has been adopted by most scholars. In this paper, we develop a uniaxial tensile fracture model of nuclear graphite based on the ordinary state-based Peridynamic theory. After that, a random number and size of initial defects are added to the nuclear graphite according to the pore-size distribution in the graphite microstructurally parameters. In order to verify the reliability of the model, this paper compares the prediction results of the model with the experimental data on two typical graphites using the two-parameter Weibull function. Besides, this paper compares the model in this paper with the Burchell model, and the results show that the spherical defect shape used in this paper is more suitable for predicting isostatic graphite.