TY - JOUR
T1 - Sensitivity influence of initial crack characteristics on structural damage propagation based on the VB-PCE model and POD reduced order algorithm
AU - ZHU, Lin
AU - WANG, Junhao
AU - QIU, Jianchun
AU - CHEN, Min
AU - JIA, Minping
N1 - Funding Information:
This work was supported by National Natural Science Foundation of China ( 51805447 ), Natural Science Foundation of the Jiangsu Higher Education ( 22KJB460010 ), Yangzhou Science and Technology Plan (City School Cooperation Project) (YZ2022193).
Publisher Copyright:
© 2023 Elsevier Ltd
PY - 2023/12
Y1 - 2023/12
N2 - Fatigue crack growth is normally influenced by multiple factors, including crack length, crack depth, crack angle and section radius. In this paper, an improved sensitivity surrogate model is proposed based on the VB-PCE model and POD reduction algorithm. Through the dynamic and static structural analysis and multi-order sensitivity calculations of parameters, a quantitative analysis of the multiple uncertain variables’ influence on the structural stress intensity factor is achieved. Finally, the accuracy and effectiveness of the model are verified through a typical port crane. The results demonstrate that the angle of the crack has the greatest effect on the crack parameters, while the radius of the crack interface has the least effect. The average sensitivity prediction accuracy of the influencing parameters is 93.34%. The modified sensitivity model can rapidly solve the problem of high-dimensional parameters with high efficiency and good convergence.
AB - Fatigue crack growth is normally influenced by multiple factors, including crack length, crack depth, crack angle and section radius. In this paper, an improved sensitivity surrogate model is proposed based on the VB-PCE model and POD reduction algorithm. Through the dynamic and static structural analysis and multi-order sensitivity calculations of parameters, a quantitative analysis of the multiple uncertain variables’ influence on the structural stress intensity factor is achieved. Finally, the accuracy and effectiveness of the model are verified through a typical port crane. The results demonstrate that the angle of the crack has the greatest effect on the crack parameters, while the radius of the crack interface has the least effect. The average sensitivity prediction accuracy of the influencing parameters is 93.34%. The modified sensitivity model can rapidly solve the problem of high-dimensional parameters with high efficiency and good convergence.
KW - Initial crack characteristics
KW - Polynomial chaos expansions
KW - Proper orthogonal decomposition
KW - Sensitivity influence
KW - Variational Bayesian
UR - http://www.scopus.com/inward/record.url?scp=85168554202&partnerID=8YFLogxK
U2 - 10.1016/j.ress.2023.109551
DO - 10.1016/j.ress.2023.109551
M3 - Article
AN - SCOPUS:85168554202
SN - 0951-8320
VL - 240
JO - Reliability Engineering and System Safety
JF - Reliability Engineering and System Safety
M1 - 109551
ER -