基于SVM的A508-3钢样品断裂韧性的尺寸影响研究

打开文本图片集

中图分类号：TB9；TG142.1-5 文献标志码：A文章编号：1674-5124（2025）07-0072-08

Abstract： The standard CT specimens typically employed for fracture toughness testing of A508-3 RPV steel present challenges due to their large volume and increased induced radioactivity after irradiation. Therefore， there is a trend toward testing with smaller specimens. Fracture toughness values （KJc） are influenced not only by specimen size but also by several parameters such as crack ratio and testing temperature.The conventional methods of size transition for KJc require that experimental values satisfy high constraint conditions，and they are intricate in addressing the coupled effects of other factors and size.Utilizing the SVM to model fracture toughness-related values （elastic/plastic J-integrals Je and Jp， fracture toughness KJcΔ⋅ ）based on experimental data offers the capability to comprehensively consider the influence of multiple factors with minimal data.

Compared to theoretical models，low constraint effects are considered. The results reveal a significant reduction in Jp with an increase in sample size， a gradual increase in Je， and a decrease in Kr .Each value decreases with the increase of crack proportion. Coupled efect analysis indicates that with decreasing sample size， the sensitivity of Jp to temperature intensifies， while the sensitivity of Je to temperature diminishes. Keywords：RPV steel; machine learning; fracture toughness; size effect

0 引言

A508-3钢作为反应堆压力容器（RPV），其在辐照后低温脆性增大，为预防断裂事故发生，需要评估材料韧脆特性。（剩余9083字）