摘要
高钢级管线钢的断面凹凸不平并且生成的裂纹没有尖锐整齐的裂纹尖端存在。因此,通过标准的测试方法无法获得传统的断裂力学参数。文章基于离散内聚力模型和有限元方法,对高钢级管线钢拉伸试样的测试数据进行了分析,并提出了一种通过人工神经网络拟合获得三角形内聚力法则中的聚合强度和断裂能两个参数的方法。结果表明,在离散内聚力模型中代入该参数后,通过仿真得到的载荷-位移曲线与试验获得的结果吻合很好。因此,该论文的研究工作可为基于高钢级管线钢构建的船舶与海洋结构物断裂分析打下良好的基础。
The fractured surface of toughen high strength steel is not fiat and crack front is not straight, ods. In and thus, conventional fracture parameters cannot be obtained through standardized test meth- this paper, discrete cohesive zone model （DCZM） was employed to analyze the fracture test results on toughen high strength steel. The simple triangular cohesive law in conjunction with finite element analysis was adopted. An artificial neural network （ANN） approach is proposed to determine the cohesive strength and fracture energy, which are the fundamental parameters to establish the tri- angle type cohesive law. In engineering sense, the load vs. displacement curves predicted by the DCZM using the determined parameters agree with the test curve quite well. Therefore, this research lays a good foundation for analyzing the fracture of the ship and offshore structure constructed from toughened high strength steels.
出处
《船舶力学》
CSCD
北大核心
2013年第6期672-679,共8页
Journal of Ship Mechanics
关键词
裂纹扩展
聚合强度
断裂能
高钢级管线钢
离散内聚力模型
人工神经网络
crack growth and propagation
cohesive strength
fracture energy
toughened high strength steel
discrete cohesive zone model （DCZM）
artificial neural network （ANN）
