| With the development of intelligent sensing technology,structural health monitoring technology has been widely developed in the fields of engineering structure,aerospace,civil and mechanical structure.The damage diagnosis and identification of engineering structure under dynamic loads have attracted more and more attention from scholars in terms of safety assessment,post-disaster reconstruction and risk estimation.Complicated and large-scale structural system will inevitably suffer from various types of damage during long-term service.When the damage is not found or repaired,it will lead to further development of the damage and even damage to the structural system.The theory and method of structural damage identification based on vibration test has been a hot topic in the field of civil engineering and infrastructure.At the same time,compared with the linear structure damage diagnosis,the damage development process of engineering structure system under dynamic loading is often nonlinear.Previous damage identification theories and methods based on structural dynamic response measurement and eigenvalue extraction are not applicable in many nonlinear systems.Moreover,for many structural systems,the dynamic characteristics of the structure cannot be determined before identification,and there are obvious differences between the identification of linear and nonlinear structural systems.Based on the above background,this paper takes the restoring force coefficient as an index to describe the linear or nonlinear characteristics of the structure,and proposes a multi-degre of freedom structure non-parametric damage identification method based on power series polynomials.The specific research contents are as follows:In the part of theoretical research,a non-parametric damage identification method based on the power series expansion model is proposed,which can effectively identify the damage state of the structure system when the linear or nonlinear type and mass of the structure are unknown.Based on the dynamic response information of acceleration,velocity and displacement,this method which starts from the structural dynamic balance equation can identify the restoring force parameters by using power series polynomial,least square method and Newton’s third law.In the experimental part,the linear and nonlinear shaking table tests of a five-layer shear aluminum alloy frame model are carried out.The vibration table is used to apply random load excitation to the structure base and the wireless acceleration sensor is used to collect the time history response information of the structure acceleration.The response information of velocity and displacement structure is obtained by the time domain integration.Based on the response information above,the model parameters and damage degree of the structure are identified effectively.The linear structure experiment simulates the damage by changing the stiffness of each layer of the structure and compares the experimental results with the theoretical values.Meanwhile,the influence of non-contact members and looseness of bolts on the nonlinear characteristics of the structure is studied on the basis of linear structure,and the parameters of the restoring force of the structure are identified effectively.In the numerical simulation,the finite element model is established to study the damage identification problem of different degrees in each layer of the structure under unidirectional and bi-directional excitation loads,and the influence of 5% and 10% degree of random noise is considered.The difference of recognition results between circular column and rectangular column is studied in the two-direction loading simulation.The damage states of different floors with different degrees are designed and the parameters of restoring force are effectively identified.In addition,the numerical simulation results are compared with the theoretical and experimental results to verify the effectiveness and robustness of the method in structural damage identification. |
PDF Full Text Request