In this research, a perturbation method using the element stiffness matrices is developed to detect and estimate damages in a truss model. A group parametric perturbation approach was used to study changes in the stiffness values of several structural members grouped as one parameter. These algorithms facilitate a parametric updating process. Simulations show that using the perturbation equations, damage was limited to a certain percentage. Therefore numerical methods were developed to tackle large percentage damage detection problems. Based on the eigenvectors of a truss model, some parametric functions can be derived by knowing the geometry and mechanical properties of the model. The functions describing the eigenparameters in terms of the structural parameters can be expressed by pol...[ Read more ]

In this research, a perturbation method using the element stiffness matrices is developed to detect and estimate damages in a truss model. A group parametric perturbation approach was used to study changes in the stiffness values of several structural members grouped as one parameter. These algorithms facilitate a parametric updating process. Simulations show that using the perturbation equations, damage was limited to a certain percentage. Therefore numerical methods were developed to tackle large percentage damage detection problems. Based on the eigenvectors of a truss model, some parametric functions can be derived by knowing the geometry and mechanical properties of the model. The functions describing the eigenparameters in terms of the structural parameters can be expressed by polynomials and other numerical methods. Algorithms were implemented to monitor localized damages and groups of structural members.

Simulations were performed on a finite element model. While, modal tests were carried out on the actual structure. The methods were evaluated under different damage situations. The technology was applied to a typical structural damage due to earthquake.