Search

This paper reports the development of a vibration monitoring system for wind turbine blades. This system is used to estimate the deflection at the tip blade on a wind turbine tower. Technical accidents of wind turbine blades have become increasingly common. Thus, regular monitoring of the blades is very important to prevent breakdowns, especially in cases when the blades begin to vibrate excessively. The monitoring system developed in this study satisfies two main objectives for practicality. First, our system is easy to install on existing wind turbines. Second, blade deflection is measured in real time. Our system can be operated using a few strain gages attached at the blade root, and the deflection is calculated based on the monitored stress. Thus, direct measurement of deflection at the blade tip is unnecessary. An estimation algorithm for this purpose is adopted based on the experimental modal analysis. This paper focuses on the evaluation of the estimation algorithm to investigate the feasibility of our system. Basic experiments were conducted using a simple blade model of a 300 W scaled wind turbine under rotation. Signals from the strain gages were acquired by a sensor network and sent to a computer through a wireless connection. The results show that the estimation accuracy is acceptably high. Therefore, we conclude that our proposed system is practical.

The bending stiffness of the wind turbine blades has decreased due to the trend of wind turbine upsizing. Consequently, the risk of blades breakage by hitting the tower has increased. In order to prevent such incidents, this study proposes a deflection monitoring system that can be installed to already operating wind turbine's blades. The monitoring system is composed of an estimation algorithm to detect blade deflection and a wireless sensor network as a hardware equipment. As for the estimation method for blade deflection, a strain-based estimation algorithm and an objective function for optimal sensor arrangement are proposed. Strain-based estimation algorithm is using a linear correlation between strain and deflections, which can be expressed in a form of a transformation matrix. The objective function includes the terms of strain sensitivity and condition number of the transformation matrix between strain and deflection. In order to calculate the objective function, a simplified experimental model of the blade is constructed by interpolating the mode shape of a blade from modal testing. The interpolation method is effective considering a practical use to operating wind turbines' blades since it is not necessary to establish a finite element model of a blade. On the other hand, a sensor network with wireless connection with an open source hardware is developed. It is installed to a 300 W scale wind turbine and vibration of the blade on operation is investigated.

This paper presents a new type of damping material composed of flexible porous material such as polyurethane foam and liquid such as water. This new type of damping material costs less than viscoelastic materials such as Sorbothane, and is anticipated to have a better damping effect. At the same time, installation of the new type of damping material is as easy as that of usual viscoelastic materials. The damping effect of this material on the vibration of a cantilever iron plate is experimentally compared with the one of a viscoelastic material. The damping effect on a rail piece with a length of 1.2m and a weight of 72kg is also experimentally investigated. The experimental results show a good damping effect. The design variables of this new damping material such as the liquid viscosity and the hardness of the porous material are thought to influence the damping effect, consequently are investigated by carrying out vibration testing of a cantilever iron plate (material:SS400, size:300×440×5mm, weight:5.2kg).

Friction between two relatively sliding surfaces in mechanical structures possibly excites unstable vibration, which emits squeal noise. In this paper, the authors propose an idea to predict the possibility of squeal noise occurrence in actual structures having relative sliding surfaces with friction by means of experimental modal analysis and experimental spatial matrix identification. Using spatial matrices that derived from experimental modal data, unstable mode coupling that causes squeal noise can be predicted. By this method it is possible to calculate the possibility of squeal noise incident and its frequency within the practical realistic range of friction coefficient and contact stiffness. The method is presented with showing an experimental verification study using a simple L shape frame.

This paper presents a study of experimental identification of vibration force generated in a single cylinder outboard engine under a steady-state operation by means of linear structural modeling and equivalent force transformation. The identification is to determine a set of equivalent force vectors that can express actual vibration force generated by firing and inertia force due to dynamic motion of moving components in the engine. The vibration force generated in engines under steady-state operation will be independent of the boundary conditions excluding load condition. Therefore, once such vibration force can be modeled with practical accuracy, the result can be used as a database for CAE simulations of trouble shooting and design improvement analysis for noise and vibration, etc.

Many researchers have made efforts to develop structural health monitoring methods to detect damages, cracks, corrosion and bolt loosening using modal characteristics of structures. However, modal characteristics are affected by not only damages but also boundary conditions and loading conditions etc. In fact, local damages induce very slight change of modal parameter. Therefore, it will be difficult to apply most of these conventional methods to actual structures. In this paper, we propose a new structural health monitoring method to detect bolt loosening using the characteristics of local vibration modes of a thin plate sensor sandwiched between a bolt, a nut and washers in main structures. This sensor consists of a thin plate and some piezoelectric elements. In local mode vibration of the thin plate sensor, the bolt’s axial tension determines its boundary condition. Bolt loosening changes the boundary condition and natural frequencies of the sensor. However, main structure’s characteristics don’t affect sensor’s modal characteristics. Since our method attempts using local vibration modes it is important to distinguish the local vibration modes of interest from the global ones related to structures. We propose a technique to distinguish local modes using finite element method. Finally, we confirm the validity of the thin plate sensor experimentally.

This article describes a new method and device for measuring the rigid-body properties of mechanical structures (the inertia tensor, the center of gravity coordinates, and optionally the mass). The setup consists of a platform constrained from underneath by soft coil springs and supported at its center by a 5-DOF air bearing mechanism. Following a random push by hand, laser distance sensors measure the platform’s free vibrations. The rigid-body properties are then obtained by fitting a numerical model to the free vibration signals. The model’s key components (a mass matrix, a stiffness matrix, and a kinematic transformation matrix) are obtained by calibration. Together with the unknown rigid-body properties, the three matrices define the natural frequencies and unscaled mode shapes, which are fitted to the sensor signals directly in time domain. A prototype of the device is presented and its accuracy is evaluated in an extensive series of tests. Advantages over previous methods include a faster and simpler procedure, higher accuracy, lower frequencies, and an unlimited work space.

The purpose of the study described in this paper is to analyze and predict the vibration responses of an internal combustion engine (ICE). Firstly, all substructures of the ICE are modelled by the finite-element method, and their dynamic characteristics obtained by a component mode synthesis method (CMS) or by a multiple component mode synthesis method (MCMS) are checked by comparison with experimental results. Secondly, the dynamic characteristics of the whole system synthesized by MCMS are compared with experimental results to verify the accuracy of analytical results and the proprieties of the adopted connecting conditions. Thirdly, a computational prediction of the vibration characteristics, when the supplementary parts for reducing the vibration response level are added to the bottom of the cylinder block of the original engine, is carried out and also compared with experimental results. In addition, the overall level of the force at intake/outtake valve setting points is experimentally estimated in the frequency domain for the purpose of application as a factor of excitation forces in the engine during operation. Then, by using the FE model and the estimated valve-driving forces, the forced vibration analysis is carried out with MCMS. It is shown that the result corresponds well with the experimentally obtained results.

In this paper, the authors propose a drastic improvement with regard to the force-response window function, which is commonly used on impact vibration testing. The customary force-response window function pairs a rectangular window function for impulsive force signals and an exponential function for vibration response signals. The three dynamic characteristics : inertia, damping and stiffness, are measured with distortion since neither is an actual force signal an exact impulse, nor is data acquisition triggered at t=0 for the exponential window function. Consequently, modal parameters are necessarily identified with bias errors that cannot be ignored in the measured data. Bias errors cannot be removed by any post-treatment techniques because an actual force is never an exact impulse. An improved force-response window function frees the experimental results obtained by impact vibration testing from the problem of the bias error.

This paper is to present a promising method of experiment-based substructure modeling and synthesis and a case study for low frequency noise and vibration analysis using a basic specimen of small boat consisting of a boat-hull, a single-cylinder-outboard-engine and its suspension-mechanism. Any machine generally consists of substructures and sub-machines. Cruiser is an example. Various combinations can be possible in choice of hulls, engines, engine-mounts or suspension-mechanisms. Noise and vibration characteristics are much dependent on the combinations. Therefore, it is desired to develop simple and speedy methods that are available to use for engineering design development and techno-business discussion with customers.

The accurate prediction of the sound power radiated from complicated structures, whose precise models are practically difficult to be made by the finite element method, is performed by an experimental-based approach combining impulse vibration testing and the boundary element method. This approach is applied to a small boat hull comprising of some panels with some longitudinal bending stiffeners and ribs as a basic research work in the series of a research for establishing an integrated analysis and optimization of mechanical systems to be assembled some components such as boat hulls and outer engines by means of using both experimental and theoretical modeling. It is verified that the predicted sound power level radiated from the hull is in a good agreement with an experiment. The parts of the hull radiating the noise dominantly can be identified.Copyright © 2005 by ASME

This paper presents a new method for identifying the stiffness of mounts under the condition of being built in a structural system as in practical use. A conventionally standard method is to identify an individual mount using a dynamic stiffness measurement equipment. In the new method, the stiffness values of mounts built in a structural system are identified using the modal parameters of the system which are obtained by experimental modal analysis. Vibration testing of the structural system with a posture in an operational condition can make it possible to consider preload effect that is caused by the gravitational and thrust force applying to the structural system in practical conditions. Additionally, vibration testing using impact hammer is often made available so that the new method can make it much easier to identify the stiffness of mounts than the conventional measurement of dynamic stiffness. In this paper, the validity of the method is demonstrated using both a simulation study using a finite element model and an experimental identification of an actual outboard engine. The experimental result shows a dependence of the mount stiffness on the gravitational force and the thrust force at the propeller of the outboard engine.

This paper presents the result of using an experimental spatial matrix identification method to predict the dynamics of a frame structure under a different boundary condition. The single-input-multiple-output frequency response functions of the test structure under the free-free boundary condition are measured by hammer testing. Using the FRFs, a set of spatial matrices is determined to represent its structural dynamic characteristics by the method. Then, using the identified spatial matrices, the dynamic characteristics of the test structure under the boundary condition of clamping 4 points is predicted. The prediction is practically accurate. The result of the prediction demonstrates that the spatial matrices identified by the method can be used for structural modification and substructure synthesis in the field of computer aided mechanical engineering.