Search results

To develop a new method for estimation of damage configuration in composite laminate structure using acoustic wave propagation signal and a reduction‐prediction neural network to deal with high dimensional spectral data.

A reduction‐prediction network, which is a combination of an independent component analysis (ICA) and a multi‐layer perceptron (MLP) neural network, is proposed to quantify the damage state related to transverse matrix cracking in composite laminates using acoustic wave propagation model. Given the Fourier spectral response of the damaged structure under frequency band‐selective excitation, the problem is posed as a parameter estimation problem. The parameters are the stiffness degradation factors, location and approximate size of the stiffness‐degraded zone. A micro‐mechanics model based on damage evolution criteria is incorporated in a spectral finite element model (SFEM) for beam type structure to study the effect of transverse matrix crack density on the acoustic wave response. Spectral data generated by using this model is used in training and testing the network. The ICA network called as the reduction network, reduces the dimensionality of the broad‐band spectral data for training and testing and sends its output as input to the MLP network. The MLP network, in turn, predicts the damage parameters.

Numerical demonstration shows that the developed network can efficiently handle high dimensional spectral data and estimate the damage state, damage location and size accurately.

Only numerical validation based on a damage model is reported in absence of experimental data. Uncertainties during actual online health monitoring may produce errors in the network output. Fault‐tolerance issues are not attempted. The method needs to be tested using measured spectral data using multiple sensors and wide variety of damages.

The developed network and estimation methodology can be employed in practical structural monitoring system, such as for monitoring critical composite structure components in aircrafts, spacecrafts and marine vehicles.

A new method is reported in the paper, which employs the previous works of the authors on SFEM and neural network. The paper addresses the important problem of high data dimensionality, which is of significant importance from practical engineering application viewpoint.

Deals with the formulation and application of temporal and spatial spectral element approximations for the solution of convection‐diffusion problems. Proposes a new high‐order splitting space‐time spectral element method which exploits space‐time discontinuous Galerkin for the first hyperbolic substep and space continuous‐time discontinuous Galerkin for the second parabolic substep. Analyses this method and presents its characteristics in terms of accuracy and stability. Also investigates a subcycling technique, in which several hyperbolic substeps are taken for each parabolic substep; a technique which enables fast, cost‐effective time integration with little loss of accuracy. Demonstrates, by a numerical comparison with other coupled and splitting space‐time spectral element methods, that the proposed method exhibits significant improvements in accuracy, stability and computational efficiency, which suggests that this method is a potential alternative to existing schemes. Describes several areas for future research.

The transient simulation of integrated circuit has become very expensive in terms of computer time due to increase in the number of transistors in typical simulation. Spectral technique and Chebyshev polynomials offers an efficient alternative algorithm for simulation of integrated circuits. In this paper an automatic formulation of circuit elements and transistor models, built in MOS technology, for analysis using spectral technique is presented. The algorithm is implemented and the simulation is proven to require less computer time than in the case of SPICE or ASTAP

The purpose of this paper is to develop a new formulation using spectral approach, which can predict the wave behavior to uncertain parameters in mid and high frequencies.

The work presented is based on a hybridization of a spectral method called the “wave finite element (WFE)” method and a non-intrusive probabilistic approach called the “polynomial chaos expansion (PCE).” The WFE formulation for coupled structures is detailed in this paper. The direct connection with the conventional finite element method allows to identify the diffusion relation for a straight waveguide containing a mechanical or geometric discontinuity. Knowing that the uncertainties play a fundamental role in mid and high frequencies, the PCE is applied to identify uncertainty propagation in periodic structures with periodic uncertain parameters. The approach proposed allows the evaluation of the dispersion of kinematic and energetic parameters.

The authors have found that even though this approach was originally designed to deal with uncertainty propagation in structures it can be competitive with its low time consumption. The Latin Hypercube Sampling (LHS) is also employed to minimize CPU time.

The approach proposed is quite new and very simple to apply to any periodic structures containing variabilities in its mechanical parameters. The Stochastic Wave Finite Element can predict the dynamic behavior from wave sensitivity of any uncertain media. The approach presented is validated for two different cases: coupled waveguides with and without section modes. The presented results are verified vs Monte Carlo simulations.

A splitting technique for solutions of the Navier—Stokes and the energy equations, in Boussinesq approximately, is presented. The equations are first integrated in time using a splitting procedure and then discretized spatially by means of a high‐order spectral element method. The whole technique is validated on the flow in a differentially‐heated cavity at intermediate and transitional Rayliegh numbers. The results are in a very good agreement with other available numerical solutions.

The Rayleigh‐Benard‐Marangoni problem for natural convection in a rectangular cavity with thermocapillary forces on a free surface is investigated using a stream function‐vorticity formulation. The nonlinear system is iteratively decoupled and high‐degree p finite elements are used for the discretization of the physical domain. The linear systems arising from the discretization at each iteration are solved using a spectral multilevel scheme, which is a natural preconditioner for high‐p (spectral) elements. The spectral multilevel solver lends itself to parallelization in an element‐by‐element (EBE) framework. Simulation results are presented and compared to previously published results. The multilevel efficiency is compared to previous results for the driven cavity problem. Parallel performance studies are presented for the Cray T3E distributed memory architecture.

The purpose of this paper is to develop a spectral element modeling to predict electromechanical admittance in the surface-bonded piezoelectric wafer and beam structure considering temperature effects.

For modeling the beam, the axial and transverse vibrations of the beam have been considered, and temperature-dependent mechanical and electromechanical properties of piezoelectric wafer active sensor and aluminum have been analyzed. The influences of temperature effects on electromechanical admittance are investigated.

The results show that a frequency left shift and a decrease in amplitude of admittance in any natural frequencies with increasing temperature have been observed. The mechanism of these changes is discussed.

The numerical results may be considered helpful for structural health monitoring using electromechanical impedance technique.

The purpose of this paper is to offer a fast and reliable discretisation scheme for computing the electromagnetic fields inside a ferromagnetic cylinder, accounting for motional eddy currents under high velocities and accounting for the severe ferromagnetic saturation of the rotor surface.

A nonlinear spectral‐element (SE) formulation is developed and compared to existing analytical and finite‐element approaches.

The proposed SE method results in a higher accuracy, allows for smaller models, avoids upwinding and needs less computation time. Disadvantages are the dense system matrix and the bad condition number.

The SE approach is only developed and tested for 2D models with a single cylindrical domain.

The results of the paper may improve the design and optimisation of solid‐rotor induction machines and magnetic bearings.

The paper offers an appropriate solution for a computational problem, which already has been encountered by a large community of researchers and engineers dealing with high‐speed rotating devices.

The aim is to apply probabilistic approaches to electromagnetic numerical dosimetry problems in order to take into account the variability of the input parameters.

A classic finite element method is coupled with probabilistic methods. These probabilistic methods are based on the expansion of the random parameters in two different ways: a spectral expansion and a nodal expansion.

The computation of the mean and the variance on a simple scattering problem shows that only a few hundreds calculations are required when applying these methods while the Monte Carlo method uses several thousands of samples in order to obtain a comparable accuracy.

The number of calculations is reduced using several techniques: a regression technique, sparse grids computed from Smolyak algorithm or a suited coordinate system.

A bibliographical review of the finite element methods (FEMs) applied for the linear and nonlinear, static and dynamic analyses of basic structural elements from the theoretical as well as practical points of view is given. The bibliography at the end of the paper contains 1,726 references to papers, conference proceedings and theses/dissertations dealing with the analysis of beams, columns, rods, bars, cables, discs, blades, shafts, membranes, plates and shells that were published in 1996‐1999.