• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2024, Vol. 60 ›› Issue (18): 32-42.doi: 10.3901/JME.2024.18.032

• 仪器科学与技术 • 上一篇    下一篇

扫码分享

复合材料基体裂纹损伤演化的自适应粒子滤波预测

刘小峰, 张天瑀, 韦代平, 柏林, 陈兵奎   

  1. 重庆大学机械与运载工程学院 重庆 400044
  • 收稿日期:2023-10-10 修回日期:2024-05-18 出版日期:2024-09-20 发布日期:2024-11-15
  • 作者简介:刘小峰(通信作者),女,1980年出生,博士,教授,博士研究生导师,主要研究方向为超声导波无损检测、设备监测与故障诊断,智能测试与仪器。E-mail:liuxfeng0080@126.com
    张天瑀,男,1998年出生,硕士,主要研究方向为超声导波无损检测。E-mail:1530622497@qq.com
    韦代平,男,1991年出生,博士,主要研究方向为超声导波无损检测。E-mail:daipingw@126.com
    柏林,男,1972年出生,博士,教授,主要研究方向为机械设备状态监测与故障诊断。E-mail:bolin0001@aliyun.com
    陈兵奎,男,1966年出生,博士,教授,博士研生导师。主要研究方向为精密传动,不确定性分析及优化设计。E-mail:bkchen@cqu.edu.cn
  • 基金资助:
    国家科技重大专项(J2019-IV-0001-0068)和国家自然科学基金(52175077)项目资助。

Adaptive Particle Filtering Prediction of Crack Damage Evolution in Composite Materials

LIU Xiaofeng, ZHANG Tianyu, WEI Daiping, BO Lin, CHEN Bingkui   

  1. College of mechanical and Transportation Engineering, Chongqing University, Chongqing, 400044
  • Received:2023-10-10 Revised:2024-05-18 Online:2024-09-20 Published:2024-11-15

摘要: 针对复合材料层合板基体裂纹的监测数据小样本特性、损伤参数量测的随机性、退化模型的不确定性等问题,结合基于能量释放率的基体裂纹演化模型,提出复合材料剩余使用寿命的自适应粒子滤波预测方法。该方法首先采用复合材料裂纹损伤的剪滞模型法,从微观力学角度分析了层合板基体裂纹密度与其有效刚度间的解析关系,在此基础上构建基于应变能释放率的层合板裂纹损伤正演模型。然后,综合考虑层合板损伤演化的个体差异性、模型误差、测量噪声等不确定性因素,以复合材料层合板的有效刚度及基体裂纹密度为观测量,建立层合板疲劳损伤的多参量动态状态空间。最后,采用自适应粒子滤波算法对层合板裂纹损伤的演化趋势进行预测评估,同时对每个载荷循环次数下的层合板剩余寿命进行迭代估计。有限元仿真与试验分析表明,模型预测方法揭示了层合板裂纹损伤的渐进演化规律,不仅能够对损伤状态进行有效追踪而且能准确地预测出层合板的剩余使用寿命。

关键词: 复合材料层合板, 应变能释放率模型, 自适应粒子滤波, 寿命预测

Abstract: In view of the small samples of composite laminate fatigue crack data, the randomness of damage parameters measurement, and the uncertainty of degradation model, the method of predicting the remaining service life of composite materials is proposed by combining the matrix crack evolution model based strain energy release rate and the particle filtering with adaptive parameter optimization. Firstly, the shear-lag model of composite crack damage is used to analyze the analytical relationship between the matrix crack density and the laminate effective stiffness from the micromechanical point of view, based on which the strain energy release rate based laminate cracking evolution model is constructed. Secondly, taking into account the individual variability of laminate damage evolution, modeling errors, measurement noises and other uncertainties, by taking the effective stiffness and matrix crack density in the laminate as the observed quantities, a multi-parameter dynamic state space of laminate fatigue damage is established. Finally, the adaptive particle-filtering algorithm is utilized to evaluate the continuous evolution of laminate crack damage, and then the remaining useful life under different load cycles is predicted iteratively. The finite element simulation and experimental analysis showed that the proposed model and prediction method not only reveal the progressive evolution law of laminate crack damage, but also can effectively track the damage state and accurately predict the remaining useful life of the laminate.

Key words: composite laminates, strain energy release rate model, adaptive particle filtering, remaining useful life prediction

中图分类号: