چكيده به لاتين
Abstract:
Major engineering constructions which are made by humans are damaged by factors such as
earthquakes. Therefore, it is necessary and vital to pay attention to the damage localization and
quantification occurring in these structures. For this reason, many researchers have introduced
various structural damage detection methods in recent decades. In this regard, a set of practical
and important methods that can be used in engineering structures due to its large size,
complexity, and restricted access to local experiments tend to identify the location and extent
of damage are damage detection methods based on the obtained dynamical data of structures.
The basic idea behind these methods is that any change in the physical properties of the
structure (mass, attenuation, and stiffness) will lead to changes in structural modal parameters
(frequency, modal shape, and modal damping). The methods which are based on finite element
model updating are group of methods in which the difference between the analytical and
measured results depend on modifying analytical model properties such as mass, stiffness and
damping by solving as an optimization problem. In other words, defining the objective
function, determine parameters which are updated, and selecting the type of optimization
algorithm to solve damage identification problem as an inverse problem using optimization
approach are the most effective factors in these methods.
In this thesis, two new objective functions based on differences between damaged structure
and analytical model depending on finite element model updating have been presented. In order
to find the location and severity of damage in the numerical studies, the problem of identifying
damage is proposed as an optimization problem. Moreover, to solve the problem, the MothFlame Optimization (MFO) algorithm is used. In the first method, an objective function based
on Static Strain Energy (SSE) and Generalized Flexibility Matrix (GFM) is introduced in order
to determining the location and extent of damage in structures. In the second method, an
objective function based on the combination of modal strain energy and natural frequency is
proposed. To evaluate the proposed methods, for each method four different test examples
including 13 and 25 bars planar truss, 23 and 28 elements plane frame, 16 elements spatial
frame, 10 and 15 story shear frame are investigated. Furthermore, numerous challenges
including the effect of number of vibrating modes used for identifying damage and random
noise effect in results of damage detection in order that they indicate capability of abovementioned proposed methods are taken into consideration.
Keywords: Damage detection, Finite element model updating, Moth-falame optimization
algorithm, Modal strain energy, Generalized flexibility matrix, Natural frequency.
