13th International Conference on Fracture June 16–21, 2013, Beijing, China -9- This paper proposes a new damage detection strategy which can successfully identify the damage of the plate structure. Inheriting the results of Step 1, the non-dimensional stiffness parameters, only from the suspicious damage region, are updated using the Bayesian model updating method in Step 2. The significances of the method are that: when the identification of the detailed damage parameters focuses on the suspicious damage area, the number of the parameters needs to be identified has been greatly reduced, which can improve the convergence performance of the stochastic simulation method. The uncertainty of identification parameters will be reduced. More important, the nonuniqueness issue of the results of damage identification could be effectively relieved. The finite element model which introduces local stiffness reductions representing damage may be the simplest method among these damage modelling methods for SHM. But different from the traditional usage way as [1, 9, 10] in which a damage threshold must be pre-assumed, this paper uses the stochastic simulation method to obtain the samples of the stiffness parameters. The sample means are the identified values of the parameters, which represent actual severity of damage. The value of the θi is closer to 1, the more health structure; closer to 0, the more serious structure damage. The sample variance gives the credibility of the recognition results. The size and orientation of the damage can be observed by the combination of the different damage elements. The accuracy of damage identification depends on the sizes of the finite elements, which can be controlled artificially but usually on the cost of the computation effort. Although the experiment result has obtained encouraging results, how to improve the accuracy of the damage identification is the focus of the future work. Acknowledgements The first author would like to sincerely thank Professor Jianye Ching and Dr. Yi-Chu Chen in Taiwan for their selfless help of the TMCMC algorithm. This work was supported by the National Natural Science Foundation of China (No. 50935005) and the Major State Basic Research Development Program of China (973 Program) (No. 2009CB724306). Authors express the most sincere thanks to these organizations. References [1] Sohn H, Law KH, A Bayesian probabilistic approach for structure damage detection, Earthquake Engineering & Structural Dynamics, 26 (1997) 1259-1281. [2] Cheung SH, Beck JL, Calculation of Posterior Probabilities for Bayesian Model Class Assessment and Averaging from Posterior Samples Based on Dynamic System Data, Computer-Aided Civil and Infrastructure Engineering, 25 (2010) 304-321. [3] Panopoulou A, Loutas T, Roulias D, Fransen S, Kostopoulos V, Dynamic fiber Bragg gratings based health monitoring system of composite aerospace structures, Acta Astronautica, In Press, Corrected Proof (2011). [4] Doebling SW, Farrar CR, Prime MB, A summary review of vibration-based damage identification methods, Shock and Vibration Digest, 30 (1998) 91-105. [5] Wu ZS, Li SZ, Two-level damage detection strategy based on modal parameters from distributed dynamic macro-strain measurements, Journal of Intelligent Material Systems and Structures, 18 (2007) 667-676. [6] Beck JL, Au SK, Bayesian updating of structural models and reliability using Markov chain Monte Carlo simulation, Journal of Engineering Mechanics-Asce, 128 (2002) 380-391. [7] Ching JY, Chen YC, Transitional markov chain monte carlo method for Bayesian model updating, model class selection, and model averaging, Journal of Engineering Mechanics-Asce, 133 (2007) 816-832. [8] Cheung SH, Beck JL, Bayesian Model Updating Using Hybrid Monte Carlo Simulation with Application to Structural Dynamic Models with Many Uncertain Parameters, Journal of
RkJQdWJsaXNoZXIy MjM0NDE=