چكيده به لاتين
Civil engineering has made it possible to build a variety of structures and infrastructure for the benefit of the community. These structures and infrastructure may be damaged for a variety of reasons, including aging, earthquakes, flood and other external loads. Damage to a structure leads to reduced stiffness and stability and adversely affects the optimal performance of the structure. The need to monitor the process of damage and the resulting damages has led to the development of various methods of damage detection.The issue of structural health monitoring is addressed in order to identify the damage in a timely manner for restoration and prevention of severe damage. Since most structures are asymmetric due to functional and aesthetic requirements, the process of identifying damage to structures with respect to their symmetry and asymmetry effect is important.
In this study, the model updating method was used to identify damage. Based on two modal parameters namely natural frequency and mode shape, three objective functions are defined and then optimization of the objective functions is performed by the multi-verse meta-heuristic algorithm.
It should be noted that the formation of symmetry and asymmetry states in structures is very diverse, and the study of all states is not limited to one study or research. In this study, three symmetrical planar trusses, five asymmetric planar trusses, two symmetric two-dimensional bending frames, ten asymmetric two-dimensional bending frames, one symmetric shear frame, three asymmetric shear frames and a total of twenty-four different structures were evaluated in this study. Each of these evaluations was performed with full and incomplete modal information with noise and without noise.
In general, damage detection results in asymmetric trusses are better than symmetric trusses. Creating asymmetrically in material and geometrical properties to the truss symmetry line yields the most accurate results, with the maximum error being 3.4% (equal to the minimum error among the eight modeled trusses).
In bending frames, in all symmetry and asymmetry modes, changes in the vertical spans of the frame causes an increase in error. The most accurate results for the bending frame are obtained when asymmetrically applying by separating the beam and column material properties, with changing in the horizontal spans and the maximum error in this case being 3.2% (equal to the minimum error among the twelve modeled bending frames).
The results of damage detection in the asymmetric shear frame are more accurate than the symmetric shear frame, and the best results are obtained when the masses of stories are different and their stiffness are same. The maximum error in this case is 3.5% (equal to the minimum error between the four modeled shear frames).
