Comparative study of model match techniques used in damage detection on reticulated shell

Mismatching between the degrees of freedoms (DOFs) of the measurement model and those of the finite element model is the principal drawback when damage detection technique is used to monitor actual civil constructions. Model reduction algorithms and eigenvector expansion algorithms are the special methods to resolve the above problem. The two-step damage detection algorithm based on the residual modal force and the minimum rank perturbation theory is proposed. Some different types of model reduction algorithms and eigenvector expansion algorithms are compared when they have application in damage detection on the reticulated shell. The feasibility, strength and scope of application of each algorithm are introduced. Under the incomplete measurement information, a numerical simulation is deeply conducted for damage detection on the single-layer reticulated shell model. Meantime, the measurement error is taken into account and identification capacity of the proposed algorithm corresponding to damage extent is analyzed. The results show that with the proposed damage detection method the single location damage can be determined. It is feasible for health diagnoses of reticulated shells. Furthermore, the results using Guyan reduction and Guyan expansion algorithms to match the DOFs are of high accuracy. They are the best selection of matching the DOFs of models.