Spectral Finite Element Method Wave Propagation, Diagnostics and Con
In recent times, the use of composites and functionally graded materials (FGMs) in structural applications has increased. FGMs allow the user to design materials for a specified functionality and therefore have numerous uses in structural engineering. How
- PDF / 1,107,253 Bytes
- 57 Pages / 439.37 x 666.142 pts Page_size
- 64 Downloads / 277 Views
In this chapter the damage model derived in Chapter 9 will be used to develop numerical algorithms for damage detection studies. This chapter will present four such algorithms, namely, spectral power flow, the damage force indicator, genetic algorithms and the artificial neural network (ANN).
10.1 Strategies for Identification of Damage in Composites With the increasing use of composites as structural materials in aerospace and other industries, there is a growing need for identification of delamination and other modes of damage as part of the structural health monitoring and structural integrity evaluation. Several methods based on vibration characteristics for structural health monitoring have been reviewed in [10]. In more recent times in this direction, new sophisticated strategies for damage identification using modal parameters have been studied extensively [197], [198], [199], [200], [167]. Since modal parameters depend on the material property and geometry, the change in natural frequencies, mode shape curvature etc. can be used to locate damage in structures without knowing the excitation force when linear analysis is adequate. Lim and Kashangaki [201] located damage in space truss structures by computing Euclidian distances between the measured mode shapes and the best achievable eigenvectors. The best achievable eigenvectors are the projection of the measured mode shapes onto the subspace defined by the refined analytical model of structure and measured frequencies. Liu [202] used direct minimization of residue in the eigenequation for identification and damage detection in trusses using modal data. Manning [203] used active member transfer function data in conjunction with an ANN to detect damage in structures. It relies on training a neural network using active member transfer function pole/zero information to classify damaged structural measurements and predicts the degree of damage in a structure. The active members (transducers) that are already present in the
308
10 SFEM for SHM
controlled structure can be utilized for this purpose (input and output for interrogation). However, modal methods are not very sensitive to the small size delaminations which are of practical interest, and can be very cumbersome as well as computationally expensive when implementing in practice for on-line health monitoring. In most methods based on modal parameters, it is assumed that the modes under consideration are affected by damage. As pointed out in [200] the change in individual natural frequencies due to slight damage may become insignificant and may fall within measurement error. In practical situations, this can considerably reduce the effectiveness of the prediction. With a view to alleviating such difficulty, results of broadband analysis using the spectral element for delaminated beams are discussed in this chapter. Wave propagation analysis has been used extensively in non-destructive techniques (NDT). However, in the context of intelligent health monitoring tasks, there are several possibilities that such wave propag
Data Loading...
