Search results

This paper aims to describe a method for efficient frequency domain model order reduction. The method attempts to combine the desirable attributes of Krylov reduction and proper orthogonal decomposition (POD) and is entitled Krylov enhanced POD (KPOD).

The KPOD method couples Krylov’s moment-matching property with POD’s data generalization ability to construct reduced models capable of maintaining accuracy over wide frequency ranges. The method is based on generating a sequence of state- and frequency-dependent Krylov subspaces and then applying POD to extract a single basis that generalizes the sequence of Krylov bases.

The frequency response of a pre-stressed microelectromechanical system resonator is used as an example to demonstrate KPOD’s ability in frequency domain model reduction, with KPOD exhibiting a 44 per cent efficiency improvement over POD.

The results indicate that KPOD greatly outperforms POD in accuracy and efficiency, making the proposed method a potential asset in the design of frequency-selective applications.

The purpose of this paper is to highlight that the need for improved system identification methods within the domain of modal analysis increases under the impulse of the broadening field of applications, e.g., damage detection and vibro-acoustics, and the increased complexity of today’s structures. Although significant research efforts during the last two decades have resulted in an extensive number of parametric identification algorithms, most of them are certainly not directly applicable for modal parameter extraction. So, based on this, the aim of the present work is to develop a technique for modal parameter extraction from the measured signal.

A survey and classification of the different modal analysis methods are made; however, the focus of this thesis is placed on modal parameter extraction from measured time signal. Some of the methods are examined in detail, including both single-degree-of-freedom and multi-degree-of-freedom approaches using single and global frequency-response analysis concepts. The theory behind each of these various analysis methods is presented in depth, together with the development of computer programs, theoretical and experimental examples and discussion, in order to evaluate the capabilities of those methods. The problem of identifying properties of structures that possess close modes is treated in particular detail, as this is a difficult situation to handle and yet a very common one in many structures. It is essential to obtain a good model for the behavior of the structure in order to pursue various applications of experimental modal analysis (EMA), namely: updating of finite element models, structural modification, subsystem-coupling and calculation of real modes from complex modes, to name a few. This last topic is particularly important for the validation of finite element models, and for this reason, a number of different methods to calculate real modes from complex modes are presented and discussed in this paper.

In this paper, Modal parameters like mode shapes and natural frequencies are extracted using an FFT analyzer and with the help of ARTeMiS, and subsequently, an algorithm has been developed based on frequency domain decomposition (FDD) technique to check the accuracy of the results as obtained from ARTeMiS. It is observed that the frequency domain-based algorithm shows good agreement with the extracted results. Hence the following conclusion may be drawn: among several frequency domain-based algorithms for modal parameter extraction, the FDD technique is more reliable and it shows a very good agreement with the experimental results.

In the case of extraction techniques using measured data in the frequency domain, it is reported that the model using derivatives of modal parameters performed better in many situations. Lack of accurate and repeatable dynamic response measurements on complex structures in a real-life situation is a challenging problem to analyze exact modal parameters.

During the last two decades, there has been a growing interest in the domain of modal analysis. Evolved from a simple technique for troubleshooting, modal analysis has become an established technique to analyze the dynamical behavior of complex mechanical structures. Important examples are found in the automotive (cars, trucks, motorcycles), railway, maritime, aerospace (aircrafts, satellites, space shuttle), civil (bridges, buildings, offshore platforms) and heavy equipment industry.

Presently structural health monitoring has become a significantly important issue in the area of structural engineering particularly in the context of safety and future usefulness of a structure. A lot of research is being carried out in this area incorporating the modern sophisticated instrumentations and efficient numerical techniques. The dynamic approach is mostly employed to detect structural damage, due to its inherent advantage of having global and location-independent responses. EMA has been attempted by many researchers in a controlled laboratory environment. However, measuring input excitation force(s) seems to be very expensive and difficult for the health assessment of an existing real-life structure. So Ambient Vibration Analysis is a good alternative to overcome those difficulties associated with the measurement of input excitation force.

Three single bay two storey frame structure has been chosen for the experiment. The frame has been divided into six small elements. An algorithm has been developed to determine the natural frequency of those frame structures of which one is undamaged and the rest two damages in single element and double element, respectively. The experimental results from ARTeMIS and from developed algorithm have been compared to verify the effectiveness of the developed algorithm. Modal parameters like mode shapes and natural frequencies are extracted using an FFT analyzer and with the help of ARTeMiS, and subsequently, an algorithm has been programmed in MATLAB based on the FDD technique to check the accuracy of the results as obtained from ARTeMiS. Using singular value decomposition, the power Spectral density function matrix is decomposed using the MATLAB program. It is observed that the frequency domain-based algorithm shows good consistency with the extracted results.

The purpose of this paper is to develop an easy-to-implement and accurate fast boundary element method (BEM) for solving large-scale elastodynamic problems in frequency and time domains.

A newly developed kernel-independent fast multipole method (KIFMM) is applied to accelerating the evaluation of displacements, strains and stresses in frequency domain elastodynamic BEM analysis, in which the far-field interactions are evaluated efficiently utilizing equivalent densities and check potentials. Although there are six boundary integrals with unique kernel functions, by using the elastic theory, the authors managed to accelerate these six boundary integrals by KIFMM with the same kind of equivalent densities and check potentials. The boundary integral equations are discretized by Nyström method with curved quadratic elements. The method is further used to conduct the time-domain analysis by using the frequency-domain approach.

Numerical results show that by the fast BEM, high accuracy can be achieved and the computational complexity is brought down to linear. The performance of the present method is further demonstrated by large-scale simulations with more than two millions of unknowns in the frequency domain and one million of unknowns in the time domain. Besides, the method is applied to the topological derivatives for solving elastodynamic inverse problems.

An efficient KIFMM is implemented in the acceleration of the elastodynamic BEM. Combining with the Nyström discretization based on quadratic elements and the frequency-domain approach, an accurate and highly efficient fast BEM is achieved for large-scale elastodynamic frequency domain analysis and time-domain analysis.

This paper aims to present an approach for calibrating the numerical models of dynamical systems that have spatially localized nonlinear components. The approach implements the extended constitutive relation error (ECRE) method using multi-harmonic coefficients and is conceived to separate the errors in the representation of the global, linear and local, nonlinear components of the dynamical system through a two-step process.

The first step focuses on the system’s predominantly linear dynamic response under a low magnitude periodic excitation. In this step, the discrepancy between measured and predicted multi-harmonic coefficients is calculated in terms of residual energy. This residual energy is in turn used to spatially locate errors in the model, through which one can identify the erroneous model inputs which govern the linear behavior that need to be calibrated. The second step involves measuring the system’s nonlinear dynamic response under a high magnitude periodic excitation. In this step, the response measurements under both low and high magnitude excitation are used to iteratively calibrate the identified linear and nonlinear input parameters.

When model error is present in both linear and nonlinear components, the proposed iterative combined multi-harmonic balance method (MHB)-ECRE calibration approach has shown superiority to the conventional MHB-ECRE method, while providing more reliable calibration results of the nonlinear parameter with less dependency on a priori knowledge of the associated linear system.

This two-step process is advantageous as it reduces the confounding effects of the uncertain model parameters associated with the linear and locally nonlinear components of the system.

The purpose of this paper is to detail the design of a fractional controller which was developed for the suppression of the flexural vibrations of the first mode of a smart beam.

During the design of the fractional controller, in addition to the classical control parameters such as the controller gain and the bandwidth; the order of the derivative effect was also included as another design parameter. The controller was then designed by considering the closed loop frequency responses of different fractional orders of Continued Fraction Expansion (CFE) method.

The first, second, third and fourth order approximations of CFE method were studied for the performance analysis of the controller. It was determined that the increase in the order resulted in better vibration level suppression at the resonance. The robustness analysis of the developed controllers was also conducted.

The experimentally obtained free and forced vibration results indicated that the increase in the order of the approximations yielded better performance around the first flexural resonance region of the smart beam and proved to yield better performance than the classical integer order controllers.

Evaluation of the performance of a developed fractional controller was realized by using different approach orders of the CFE method for the suppression of the flexural vibrations of a smart beam.

The purpose of this paper is to present an original design procedure for a flight control system.

An optimization process, based on a genetic algorithm (GA), is used to meet the frequency domain handling qualities requirements in the longitudinal plane for an unconventional platform characterized by nonlinear aerodynamics. The parameters are implemented in the search process as fitness functions related to the expected magnitude of bandwidth and delay for an existing micro aerial vehicle. The bandwidth and the delay of the longitudinal short‐term attitude response are estimated before and after the inclusion of the flight control system in the simulation model, and the parameters are compared with the expected handling qualities levels. A qualitative analysis of handling qualities levels is also performed by implementing the augmented aircraft in a simulator with a realistic visual environment.

The results show that an optimal search process based on a GA can implement the handling qualities requirements with a computational procedure that is straightforward.

Even if the requisites for bandwidth and delay implemented in the search process are general in use as no specific aircraft response type is taken as a reference for the estimation of handling qualities requirements, only future experimental work will provide insight for the definition of specific Level 1 boundaries for micro aerial vehicles in remotely piloted flight.

The virtual environment is useful to test remote piloting with unconventional onboard visual cues. This is important in applications in which technical limitations may preclude complete real time data link during flight tests in the first development phase of the vehicle.

The purpose of this paper is to propose an identification method of acquiring aircraft mode characteristics based on fast Fourier transform and half-power bandwidth method, aiming at the common oscillation met in flight test.

The feasibility of this method is demonstrated through derivation; the robustness analysis is conducted through three examples, and finally the method was applied on a set of sideslip angle record from flight test.

The derivation and numerical analysis both show that the presented method can have high accuracy and good robustness under coupled mode and noise condition.

The method proposed is of robustness, and it is concise and easy to apply on flight data record.

This paper demonstrates the feasibility of half power bandwidth to be applied on oscillation mode characteristics identification from flight data record, which is different from other method applied.

A fast algorithm is proposed to calculate the lightning horizontal electric field over a lossy ground.

The lightning horizontal electric field in frequency domain is approximated by a number of piecewise cubic polynomial functions by using the proposed adaptive Hermite strategy. To utilize the Hermite strategy, the frequency domain spectrum and its derivative with respect to frequency are required. The integral kernel of the derivative appears singular along the real axis. To overcome this singularity and accelerate the calculation, a new integration path is proposed. With the help of the Hermite interpolation model and the new path, the lightning horizontal electric field in time domain can be obtained rapidly.

The singularity problem has been overcome with the new integration path and the adaptive Hermite strategy proposed in this paper is at least 50 times faster than the one using the equally spaced sampling approach.

The adaptive Hermite approach can be a good candidate for fitting a wideband frequency domain response and the revised new integration path can be utilized when the calculation of the generalized Sommerfeld integral (GSI) or its derivative with respect to frequency is involved.

The paper aims to present an overview of techniques for the identification in the frequency domain of reduced order models for distributed passive electromagnetic structures.

Most known approaches proposed in different application contexts are described within a unified framework.

A passive reduced order model of an unshielded twisted pair is fully developed with the combination of vector fitting algorithm and the passivity enforcement via Hamiltonian perturbation.

A state‐of‐the‐art picture of the frequency domain identification and passivity enforcement techniques is given, and a test case of actual interest fully analysed.

The purpose of this paper is to investigate limit cycles in digitally Proportional, Integral and Derivative (PID) controlled buck regulators. Filtering is examined as a means of removing the limit cycles in digitally controlled buck regulators.

The paper explains why limit cycles occur in a digitally PID controlled buck converter. It then proceeds to propose two filters for their elimination. Results indicate the effectiveness of each of the filters.

The paper gives a mathematical analysis of the occurrence of limit cycles in digitally controlled PID buck regulators. It finds that notch and comb filters are effective for the purpose of eliminating limit cycles in buck regulators.

The paper employs a model of the buck regulator inclusive of the inductor loss – this was not done to date for this type of work. The paper analyses PID control. This was not done in the manner given. The paper addresses filtering as a means of removing limit cycles. It examines the effect of changing the digital controller parameters on the requirements of the filters.