Calculation Method for Residual Strength of Tungsten-Copper Alloy After Hole-Drilling for Crack Arrest Based on a Thermomechanical Equivalent Phase-Field Model Combined with Multi-Fidelity Neural Network

This paper addresses tungsten-copper functionally graded materials, which are primary thermal sink materials in the power industry. a thermomechanical gradient phase-field fracture model was employed to calculate crack propagation paths and residual strengths after hole-drilling for crack arrest. The influence of different elastic energy decomposition methods and characteristic widths were compared. A method is proposed to reduce the sensitivity of characteristic width on residual strength calculations by integrating phase-field model and multi-fidelity neural network, while simultaneously optimizing the calculation accuracy.