• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2022, Vol. 58 ›› Issue (6): 110-118.doi: 10.3901/JME.2022.06.110

• 材料科学与工程 • 上一篇    下一篇

扫码分享

基于贝叶斯推断的复合材料层间内聚力模型参数反演

赵文滔1, 杨颜志2,3, 王长焕2,3, 鞠雪梅2,3, 严刚1   

  1. 1. 南京航空航天大学机械结构力学与控制国家重点实验室 南京 210016;
    2. 上海宇航系统工程研究所 上海 201109;
    3. 上海市空间飞行器机构重点实验室 上海 201109
  • 收稿日期:2021-05-20 修回日期:2021-11-12 出版日期:2022-03-20 发布日期:2022-05-19
  • 通讯作者: 严刚,男,1981年出生,博士,副教授,硕士研究生导师。主要研究方向为多功能复合材料结构、结构健康监测和结构疲劳断裂。E-mail:yangang@nuaa.edu.cn
  • 作者简介:赵文滔,男,1996年出生。主要研究方向为复合材料结构强度。E-mail:zwt@nuaa.edu.cn
  • 基金资助:
    上海市空间飞行器机构重点实验室开放课题(YYHT-F805201808012)、南京航空航天大学基本科研业务费(NS2021002)和江苏高校优势学科建设工程资助项目。

Parameter Inversion for Composite Interlayer Cohesive Zone Model Based on Bayesian Inference

ZHAO Wentao1, YANG Yanzhi2,3, WANG Changhuan2,3, JU Xuemei2,3, YAN Gang1   

  1. 1. State Key Laboratory of Mechanics and Control of Mechanical Structures, Nanjing University of Aeronautics and Astronautics, Nanjing 210016;
    2. Shanghai Institute of Aerospace Systems Engineering, Shanghai 201109;
    3. Shanghai Key Laboratory of Spacecraft Mechanism, Shanghai 201109
  • Received:2021-05-20 Revised:2021-11-12 Online:2022-03-20 Published:2022-05-19

摘要: 复合材料因为比强度大、比刚度高、可设计性强等优点,在航空航天等领域被广泛应用,其主要的损伤模式为分层损伤。内聚力模型可以很好地模拟复合材料的分层损伤,被广泛应用于复合材料结构设计和分析,因此准确确定内聚力模型参数对计算结果的可靠性具有重要作用。通过复合材料双悬臂梁试验获取载荷-位移数据信息,结合对应的有限元模型,分别采用双线性和指数型内聚力模型,基于贝叶斯推断方法反演确定层间内聚力参数的概率分布。数值仿真和试验结果表明,两种内聚力模型对断裂韧性的识别结果影响较小,而对界面强度的识别结果影响较大;贝叶斯推断能够考虑模型误差和测量噪声等不确定性因素,反演得到的参数信息比确定性方法更丰富,可以对结果进行不确定性分析。

关键词: 贝叶斯推断, 内聚力模型, 参数反演, 双悬臂梁试验, 不确定性

Abstract: Due to the advantages of high specific strength, high specific stiffness and high designability, composites have been widely used in aerospace and other fields, and their main damage mode is delamination damage. Cohesive zone model(CZM) is widely used in design and analysis of composite structures due to its capacity of simulating delamination damage, thus it is very important to accurately determine the parameters of CZM to increase the reliability of the numerical results. The load-displacement data is obtained through the composite double cantilever beam(DCB) experiments. Combined with the corresponding finite element model, Bayesian inference is employed to inversely determine the probability distributions of the parameters for bilinear and exponential CZMs, respectively. Numerical simulation and experimental results have demonstrated that, the two CZMs have neglected effect on the identification of fracture toughness but important effect on the identification of interface strength; since Bayesian inference considers uncertainties from modeling error and measurement noise, compared to deterministic methods, it can obtain more information about the CZM parameters, providing a new method to identify parameters of CZM and perform uncertainty analysis on the results.

Key words: Bayesian inference, cohesive zone model, parameter inversion, double cantilever beam tests, uncertainty

中图分类号: