H. Van der Auweraer

Multiple input orthogonal polynomial parameter estimation

Abstract The object of this paper is the development of a so-called global modal parameter estimation technique capable of analysing frequency response functions (FRFs) between several input and response stations simultaneously. The technique analyses the FRFs in their natural domain, the frequency domain. Highly consistent estimates of all modal parameters, including repeated modes, can be obtained. The effect of modes outside the analysis band can also be accounted for by explicitly locating these modes, by including residual terms, or by a combination of both. The use of orthogonal polynomials improves the numerical properties of the calculation procedure. It also reduces the order of the identification problem. All pertinent equations have a size which is proportional to the number of modes in the data, and are independent of the number of response and input stations for which data are analysed simultaneously. As a consequence, the technique lends itself readily to implementation with microcomputers.

Discriminating physical poles from mathematical poles in high order systems: use and automation of the stabilization diagram

System identification from measured MIMO data plays a crucial role in structural dynamics and vibro-acoustic system optimization. The most popular modeling approach is based on the i modal analysis concept, leading to an interpretation in terms of visualized eigenmodes. Typically, the number of nodes is very high (often over 100), including modes with high damping and high modal overlap. The paper discusses a key problem of the system identification process: the selection of the correct model order and related to this, the selection of valid system poles. A multi-order approach, followed by a heuristic selection process is outlined. A visual representation of the pole behavior is presented and the possible routes to automation are discussed. The process is illustrated with typical complex datasets, including full-scale industrial tests.

Excitation design for FRF measurements in the presence of non-linear distortions☆

Abstract In this paper, we discuss optimised strategies to measure the frequency response function in the presence of (non-linear) distortions. To do so we will compare three classes of excitation signals. These signals will be used in an optimised measurement strategy, reducing the leakage effects to acceptable (user-defined) levels, allowing to separate the disturbing noise influence from the impact of non-linear contributions, and resulting in the ‘best linear approximation’ of the system.

Continuous monitoring of the Øresund Bridge: system and data analysis

The Øresund Bridge opened in July 2000. It is the most striking part of the fixed link across the Øresund connecting Copenhagen (Denmark) and Malmø (Sweden), which further includes a tunnel and an artificial island. The bridge is equipped with a PC-based continuous monitoring system, capable of measuring both static and dynamic quantities such as temperatures, wind characteristics, air humidity, strains and accelerations. The challenges for the design of the monitoring system were the long distances between the monitoring points and the variety of sensors. This paper describes the bridge and the monitoring system components. Some typical measurement data are presented. Finally, the modal parameters of the bridge are extracted from the deck, cable and tower vibrations. This shows that the system does not only give information about sudden events that exceed a certain threshold, but can also be used as a health monitoring system by tracking the evolution of the modal parameters.

Application of stroboscopic and pulsed-laser electronic speckle pattern interferometry (ESPI) to modal analysis problems

Accurate structural models are key to the optimization of the vibro-acoustic behaviour of panel-like structures. However, at the frequencies of relevance to the acoustic problem, the structural modes are very complex, requiring high-spatial-resolution measurements. The present paper discusses a vibration testing system based on pulsed-laser holographic electronic speckle pattern interferometry (ESPI) measurements. It is a characteristic of the method that time-triggered (and not time-averaged) vibration images are obtained. Its integration into a practicable modal testing and analysis procedure is reviewed. The accumulation of results at multiple excitation frequencies allows one to build up frequency response functions. A novel parameter extraction approach using spline-based data reduction and maximum-likelihood parameter estimation was developed. Specific extensions have been added in view of the industrial application of the approach. These include the integration of geometry and response information, the integration of multiple views into one single model, the integration with finite-element model data and the prior identification of the critical panels and critical modes. A global procedure was hence established. The approach has been applied to several industrial case studies, including car panels, the firewall of a monovolume car, a full vehicle, panels of a light truck and a household product. The research was conducted in the context of the EUREKA project HOLOMODAL and the Brite-Euram project SALOME.

Trends in experimental modal analysis

Abstract The scope of this paper is to comment on current trends and new developments in the field of experimental modal analysis. The first section covers modal measurement and estimation procedures, with special emphasis on the use and limitations of recent techniques such as multiple input processing, total least square, global time- and frequency domain parameter estimation. In the second section reference is made to applications and use of modal parameters in techniques such as structural modification, fatigue and acoustic analysis. Emphasis is put on applications used at the Katholieke Universiteit, Leuven.

Accurate modal analysis measurements with programmed sine wave excitation

Abstract Due to the development of several new dynamic analysis methods which use experimental modal analysis results as input data, these results need to be subjected to stringent accuracy demands. One of the techniques to achieve this goal is the use of stepped since excitation. Although the principle of this technique has been known for a long time, a feasible implementation, suited for testing large structures or non-linear systems, has only been possible by making full use of todays techniques in digital signal processing. In this paper, the characteristics of the method are briefly reviewed, for both single input and multiple input testing. An implementation with response adaptive definition of excitation frequency and amplitude is presented and illustrated with some examples. The impact of this method on the estimation of the modal parameters is discussed.

The complex stiffnes method to detect and identify non-linear dynamic behaviour of SDOF systems

Abstract A new method to detect and identify non-linear dynamic behaviour of single-degree-of-freedom (SDOF) systems is presented. The method is based on the characteristic evolution of the least squares equivalent damping and stiffness as a function of the velocity and the displacement respectively. With these two functions, it is possible to detect the non-linearity, to identify its basic type and to obtain quantitative information on the system parameters. The method is evaluated by means of several examples and compared to other detection and identification methods. Extensions to multi-degree-of-freedom (MDOF) systems and to modeling non-linear systems are outlined.

Automated misfire diagnosis in engines using torsional vibration and block rotation

Even though a lot of research has gone into diagnosing misfire in IC engines, most approaches use torsional vibration of the crankshaft, and only a few use the rocking motion (roll) of the engine block. Additionally, misfire diagnosis normally requires an expert to interpret the analysis results from measured vibration signals. Artificial Neural Networks (ANNs) are potential tools for the automated misfire diagnosis of IC engines, as they can learn the patterns corresponding to various faults. This paper proposes an ANN-based automated diagnostic system which combines torsional vibration and rotation of the block for more robust misfire diagnosis. A critical issue with ANN applications is the network training, and it is improbable and/or uneconomical to expect to experience a sufficient number of different faults, or generate them in seeded tests, to obtain sufficient experimental results for the network training. Therefore, new simulation models, which can simulate combustion faults in engines, were developed. The simulation models are based on the thermodynamic and mechanical principles of IC engines and therefore the proposed misfire diagnostic system can in principle be adapted for any engine. During the building process of the models, based on a particular engine, some mechanical and physical parameters, for example the inertial properties of the engine parts and parameters of engine mounts, were first measured and calculated. A series of experiments were then carried out to capture the vibration signals for both normal condition and with a range of faults. The simulation models were updated and evaluated by the experimental results. Following the signal processing of the experimental and simulation signals, the best features were selected as the inputs to ANN networks. The automated diagnostic system comprises three stages: misfire detection, misfire localization and severity identification. Multi-layer Perceptron (MLP) and Probabilistic Neural Networks were applied in the different stages. The final results have shown that the diagnostic system can efficiently diagnose different misfire conditions, including location and severity.

Neural network based diagnosis of mechanical faults in IC engines

Oversize clearance induced piston slap and big end bearing knock are two common mechanical faults in the operation of internal combustion (IC) engines. A previous study has shown that the vibration signals measured on the engine block can be used to diagnose such mechanical faults in engines. However this requires some advanced signal processing techniques to be applied. Envelope analysis converts the signals from piston slap and bearing knock (second order cyclostationary signals) into deterministic signals, to which synchronous averaging can be applied. Before generating the envelope, the “kurtogram” was used to filter the signal and find the frequency bands with high impulsiveness. The amplitudes and phases of the Fourier series of the averaged envelope signals were extracted as diagnostic features. In order to realize automated and intelligent fault diagnosis of the engine, Artificial Neural Networks (ANN) were trained using the features characteristic of different faults. The networks comprise three stages: fault detection, fault localization and severity identification. The critical issue for successful application of ANNs in fault diagnosis is the selection of optimal features from the candidate ones. Two feature selection methods – “filter” and “wrapper”- were used to select the optimal features for the fault detection. The Relief approach is a typical “filter” feature selection method and Genetic Algorithm (GA) is a typical “wrapper” method. The features selected from the two approaches were separately used as the inputs to ANNs and the results are compared. Because the “wrapper” method takes into account the relevance of the individual features, the comparison showed that the GA approach has advantages in feature selection for fault diagnosis.