Cristinel Mares

Optimisation of energy and exergy of turbofan engines using genetic algorithms

This paper presents an application of genetic algorithm (GA) metaheuristics to optimise the design of two-spool separated-flow turbofan engines based on energy and exergy laws. The GA was used to seek the optimum values of eight parameters that defined the turbofan engine. A computer program called the TurboJet-Engine Optimiser v1.0 (TJEO-1.0) has been developed to perform thermodynamic property calculations of the engine and implement the optimisations. The TJEO-1.0 was integrated with Pyevolve, an open source GA optimisation framework built for use with Python programming language. The optimum designs created by TJEO-1.0 were evaluated with the following criteria: 1) energy efficiency; 2) exergy efficiency; 3) combination of both of them. Compared with the designs optimised for maximum energy efficiency, the designs optimised with the combination of energy and exergy efficiencies were able to produce 3.3%–11.0% extra specific thrust at the expense of 1.5%–2.3% extra fuel consumption.

Wavelet approach for vibration analysis of fast moving load on a viscoelastic medium

This paper analyses theoretically the response of a solid for fast moving trains using models related to real situations: a load moving in a tunnel and a load moving on a surface. The mathematical model is described by Naviers elastodynamic equation of motion for the soil and Euler-Bernoulli equation for the beam with appropriate boundary conditions. Two modelling approaches are investigated: the model with half space under the beam and the model with finite thickness of supporting medium. The problem of singularities for displacements calculation is discussed in relation with boundary conditions and types of considered loads: harmonic and constant, point and distributed moving loads. The analysis in frequency-time and frequency-velocity domains is presented and discussed with regard to critical velocities.

Clustering of Parameter Sensitivities: Examples from a Helicopter Airframe Model Updating Exercise

The need for high fidelity models in the aerospace industry has become ever more important as increasingly stringent requirements on noise and vibration levels, reliability, maintenance costs etc. come into effect. In this paper, the results of a finite element model updating exercise on a Westland Lynx XZ649 helicopter are presented. For large and complex structures, such as a helicopter airframe, the finite element model represents the main tool for obtaining accurate models which could predict the sensitivities of responses to structural changes and optimisation of the vibration levels. In this study, the eigenvalue sensitivities with respect to Youngs modulus and mass density are used in a detailed parameterisation of the structure. A new methodology is developed using an unsupervised learning technique based on similarity clustering of the columns of the sensitivity matrix. An assessment of model updating strategies is given and comparative results for the correction of vibration modes are discussed in detail. The role of the clustering technique in updating large-scale models is emphasised.

Improving Spot Weld Models in Structural Dynamics

This paper gives an overview of the finite element modelling of spot welds for the analysis of the dynamic response of structures. In particular models for dynamic analysis that use coarse meshes and equivalent parameters are considered. A major requirement for these models is their accuracy in predicting the dynamic behaviour of spot welded structures despite the low number of degrees of freedom. Three different models of spot welds are investigated [1–3] and for each model physical parameters have to be assigned based on engineering insight. The aim of the present paper is to improve the accuracy of these three models by searching for the optimum values of the parameters characterising the spot weld models using experimental data. For this purpose a benchmark structure has been analysed, consisting of a thin walled hat section beam made of two plates welded together by twenty spot welds. The predicted natural frequencies and modes of the benchmark structure have been compared to the experimental modes. Updating of the finite element models has been performed and the accuracy of the three models has been significantly improved.© 2003 ASME

Model structure correction and updating of aeroengine casings using fictitious mass modifications

Abstract In this paper the results of a finite element model updating exercise, carried out on closely axisymmetric aeroengine casings, are presented. The correction of the combustion chamber outer casing (CCOC) model is considered and, after assembly with the turbine casing (TC), the quality of the resulting combined model is investigated. The dynamics of both casings is characterized by pairs of close modes, which may be separated by fictitious point mass modifications. The natural frequencies and mode shapes of the fictitiously modified CCOC are determined from receptances obtained from the CCOC in its original (unmodified) configuration. The modifications are shown to improve the understanding of both the CCOC and the system formed by connecting the CCOC to the TC. A particular problem is revealed when model updating is applied to the CCOC. An analysis of the mode shapes locates a modelling error on an inner shell of the structure but it is found that the finite element model is unable to be parameterized for the correction of two pairs of wrongly ordered predicted modes. This can only be achieved by firstly correcting the ‘structure’ of the model itself. The main error is found to be a geometrical inaccuracy, and, when this is put right, the sequence of the modes is corrected. Model updating is then applied to the thickness of certain shell elements and the CCOC is found to be in excellent agreement with measured data, as is the complete model formed from the two models of the CCOC and the TC together.

Some fundamental aspects of self-levitating sliding contact bearings and their practical implementations:

In this study, fundamental aspects and mechanisms of acoustic levitation together with governing equations are presented first. Then, the acoustic levitation phenomenon is considered as a new way to design air suspension systems capable of self-levitation. A particular emphasis is laid on journal bearings and their specific geometrical configuration. A practical feasibility of using acoustic levitation to separate contacting surfaces is supported and illustrated by results of experimental testing of a number of prototype devices.

A pattern recognition approach to acoustic emission data originating from fatigue of wind turbine blades

The identification of particular types of damage in wind turbine blades using acoustic emission (AE) techniques is a significant emerging field. In this work, a 45.7-m turbine blade was subjected to flap-wise fatigue loading for 21 days, during which AE was measured by internally mounted piezoelectric sensors. This paper focuses on using unsupervised pattern recognition methods to characterize different AE activities corresponding to different fracture mechanisms. A sequential feature selection method based on a k-means clustering algorithm is used to achieve a fine classification accuracy. The visualization of clusters in peak frequency−frequency centroid features is used to correlate the clustering results with failure modes. The positions of these clusters in time domain features, average frequency−MARSE, and average frequency−peak amplitude are also presented in this paper (where MARSE represents the Measured Area under Rectified Signal Envelope). The results show that these parameters are representative for the classification of the failure modes.

Image enhancement for minutiae-based fingerprint identification

This paper introduces a novel method for enhancement of fingerprint image by means of eliminating the artificial induced boundaries, precise background omission, avoid oversaturation of the image in homogeneous areas, with no intermediate filtering and computational complexity. In order to eliminate the induced boundaries, CLAHE (contrast limited adaptive histogram equalization) technique is employed. dasiaClip limitpsila is then applied in order to avoid oversaturation of the image in homogeneous areas. Subsequently, the image is disintegrated into an array of distinct blocks and the discrimination of the blocks is obtained by computing the standard deviation of the matrix elements to remove the image background. Once the boundaries for the region of interest is obtained, a new modified thinning method using slide neighborhood processing is applied to clarify the endpoints and the bifurcations in each specific pixel. The purpose of this paper is to develop a simple and accurate method to improve and enhance the fingerprint image through a simulated investigation.

Wavelet analysis of beam-soil structure response for fast moving train

This paper presents a wavelet based approach for the vibratory analysis of beam-soil structure related to a point load moving along a beam resting on the surface. The model is represented by the Euler-Bernoulli equation for the beam, elastodynamic equation of motion for the soil and appropriate boundary conditions. Two cases are analysed: the model with a half space under the beam and the model where the supporting medium has a finite thickness. Analytical solutions for the displacements are obtained and discussed in relation to the used boundary conditions and the type of considered loads: harmonic and constant. The analysis in time-frequency and velocity-frequency domains is carried out for realistic systems of parameters describing physical properties of the model. The approximate displacement values are determined by applying a wavelet method for a derivation of the inverse Fourier transform. A special form of the coiflet filter used in numerical calculations allows to carry out analysis without loss of accuracy related to singularities appearing in wavelet approximation formulas, when dealing with standard filters and complex dynamic systems.

Shape descriptors for mode-shape recognition and model updating

The most widely used method for comparing mode shapes from finite elements and experimental measurements is the Modal Assurance Criterion (MAC), which returns a single numerical value and carries no explicit information on shape features. New techniques, based on image processing (IP) and pattern recognition (PR) are described in this paper. The Zernike moment descriptor (ZMD), Fourier descriptor (FD), and wavelet descriptor (WD), presented in this article, are the most popular shape descriptors having properties that include efficiency of expression, robustness to noise, invariance to geometric transformation and rotation, separation of local and global shape features and computational efficiency. The comparison of mode shapes is readily achieved by assembling the shape features of each mode shape into multi-dimensional shape feature vectors (SFVs) and determining the distances separating them.