摘要
疲劳裂纹是金属结构常见的损伤之一,提出了一种构建数字孪生模型对裂纹扩展结构进行健康管理的方法。在数字孪生模型的建立过程中,首先建立裂纹结构的有限元模型,并结合Paris公式构建裂纹扩展模型;其次基于动态贝叶斯网络表示裂纹状态随时间演化的规律;最后采用粒子滤波算法作为模型的推理算法,从而在数字空间中完成对裂纹扩展行为的描述。通过应变传感器实时在线监测裂纹状态,驱动孪生模型动态更新,实现裂纹扩展状态和剩余使用寿命更准确的预测。通过试验验证了该方法在裂纹扩展结构健康管理上的有效性与实用性。
Fatigue crack is one of common damages of metal structures.Here,a method to construct digital twin model for health management of crack propagation structures was proposed.In establishing process of digital twin model,the finite element model of crack structure was constructed firstly,and the crack propagation model was built by combining Paris formula.Secondly,the law of crack state evolution versus time was expressed based on dynamic Bayesian network.Finally,the particle filter algorithm was used as the reasoning algorithm of the model to complete the description of crack propagation behavior in digital space.Strain sensors were used to monitor crack states online in real time,drive the twin model to update dynamically,and realize more accurate prediction of crack propagation state and residue service life.The effectiveness and practicability of the proposed method in health management of crack propagation structures were verified with experiments.
作者
常琦
陈乐乐
赵恒
谢方勤
高鹤明
CHANG Qi;CHEN Lele;ZHAO Heng;XIE Fangqin;GAO Heming(School of Mechanical and Precision Instrument Engineering,Xi’an University of Technology,Xi’an 710048,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2023年第9期253-260,303,共9页
Journal of Vibration and Shock
基金
国家自然科学基金(51775429)
西安电子科技大学电子装备结构设计教育部重点实验室开放基金(EESD1901)
西安市科技计划(2020KJRC0080)。
关键词
数字孪生
疲劳裂纹扩展
健康管理
动态贝叶斯网络
粒子滤波算法
digital twin
fatigue crack propagation
health management
dynamic Bayesian network
particle filter algorithm