Magnetic bearings levitate, suspend and guide rotors by magnetic forces without physical contact. The non-contact support offers many advantages and opportunities for a wide variety of applications. Since the First International Symposium on Magnetic Bearings was held in Zurich in 1988, this field of research has expanded, and competition in research and development has been very keen. This review presents the recent trends in applications, control methodologies, sensorless controls, bearingless motors, unbalance control, problems of flexible rotors, eddy current properties and power amplifiers.

View full abstract

This paper deals with the problem of gyroscopic effect and unbalance vibration of the Magnetic Bearing system. Using a gain scheduled H_∞ control with a free parameter φ, we design a control system which attenuates the unbalance vibration, and guarantees the stability against the gyroscopic effect in specified rotational speed. Further we implement the controller and evaluate the effectiveness of the proposed approach by experiments. First, our experimental setup is explained, a mathematical model of the magnetic bearing is derived. Then, we introduce the gain scheduled H_∞ control with free parameters to a magnetic bearing control, in order to reject the disturbances caused by unbalance of the rotor, and guarantee the stability against gyroscopic effect, even if rotational speed of the rotor changes. At last, several experimental results show the effectiveness of this proposed method.

View full abstract

NASA maintains the Geosynchronous Operational Environmental Satellite(GOES)fleet for observing severe weather and anomalous atmospheric behavior such as the greenhouse effect. A critical instrument flown on these satellites is the Fourier Transform Infrared(FTIR)interferometer used for atmospheric element detection. An integral part of the FTIR interferometer is the scanning mechanism, which must position a mirror with a high degree of accuracy along its optical axis. Traditional means of controlling the position of this mirror have used flexure-based designs, "porch swing" linkages, or other mechanical actuators. As an alternative mechanism, we have developed a magnetic bearing-based scanner. This scanner uses what we call a super-hybrid magnetic circuit, in which permanent magnets bias both the magnetic bearings and a voice coil actuator. Two designs incorporating this super-hybrid magnetic bearing configuration are presented herein, along with experimental results for the first design. In addition to their use in FTIR interferometers, these magnetic bearing designs may have applications in coordinate measuring machines, lithographic steppers, and other precision motion control devices.

View full abstract

In many applications of magnetic bearings, high system performance(disturbance rejection)is required. Often several eigenfrequencies of the rotor must be actively damped. Advanced controller design is then necessary, which in ture relies on accurate plant models. System identification is therefore an important prerequisite for fast and reliable commissioning. Magnetic bearing systems are unstable MIMO plants. In case of current control, they are characterized by a set of poles on the positive and negative real axis(rigid body modes of the rotor), and a set of poorly damped poles along the imaginary axis(flexible modes). Previously described identification algorithms fail with this kind of system. Therefore a novel algorithm has been developed. The system poles are identified from the determinant of the measured frequency response function matrix. The algorithm provides a state-space model of pre-defined order and structure, suited for controller design and verification. Experimental results obtained using measurement data from a magnetic bearing system with a flexible rotor are presented.

View full abstract

Recent developments in the field of electrical drives and magnetic bearings have led to so called "bearingless motors". The innovation is that the magnetic forces for the suspension of the electrically powered rotor are generated in the motor itself and not in separate magnetic bearings. Normally, two motor / bearing units are needed for the full stabilisation of five spatial degrees of freedom. In this paper a bearingless motor with a slice-shaped rotor is presented, where three spatial degrees of freedom are passively stable. Only one active radial bearing is needed. Possible applications for the "bearingless slice motor" lie typically in the field of small centrifugal pumps, high speed centrifuges, blowers for dangerous gases or flywheels. The possibilities and advantages of the bearingless slice motor are demonstrated by the example of a disposable blood pump for heart surgery.

View full abstract

Tools for the analysis and synthesis of robust adaptive open loop controllers are evaluated experimentally on a magnetic bearing supported flexible rotor. A detailed explanation of how these tools may be applied is given via this example problem. Five synthesized controllers were tested and each resulted in significantly greater robustness than the standard solution to the problem. The value of the theory is also demonstrated by the high correlation between experimental results and the theoretical predictions of stability margin, convergence rate, and worst case performance.

View full abstract

For the assessment of machinery noise, we describe a method of estimating the normal particle velocity distribution from a sound pressure distribution measured at a single aperture close to the sound source based on near acoustic holography(NAH). In NAH, it is required that the distance between the surface of the sound source and the measurement aperture is one-tenth of a wavelength. In practice, the distance between the machinery and the measurement aperture is too large for NAH because of the 3-dimensional features of the machine, so this method is investigated under the condition that the sound pressure is measured at about one wavelength from the sound source. The size of the measurement aperture was determined to be 1m×1m. Through investigation by simulation, it has been shown that the effect of evanescent waves should be taken into consideration. Experimental verification shows that this method can be practically applied in echoic situations.

View full abstract

There are many instances in which measures for reducing machinery noise must simultaneously satisfy conflicting requirements for sound isolation and ventilation. These two requirements are contradictory because effective noise isolation requires a structure with high material density per unit area and to be extremely airtight. A new sound isolation panel has been developed that resolves this contradiction. It provides sound isolation in the required frequency band by controlling the acoustic anti-resonance frequency of the sound field inside the panel, while incorporating holes for ventilation. The working principle and characteristics of this sound isolation panel are described with the help of simulation and experimental data.

View full abstract

System identification and attitude control experiments have been performed as part of a series of bus experiments on the Engineering Test Satellite-VI, which was launched by an H-II rocket in August 1994. The aims of the experiments were to establish a pre-launch system dynamics modeling method, to evaluate typical on-orbit modal parameter identification methods and to develop robust and precise attitude controller design technology for future large flexible spacecraft and large space structures. In the on-orbit identification experiments, the central body of the satellite was excited by gas jet thrusters and the measured attitude and paddle accelerometer signals were downlinked by telemetry. In this paper, the identified parameters are compared with those of the pre-launch modeling and the accuracy of the pre-launch model is evaluated.

View full abstract

In this paper, we present a new deconvolution method for digital signals, which is distorted by the characteristics of the transfer system. First we show how to decompose the transfer system into subsystems using the discrete wavelet transform, and show that the unstable behavior of the inversion of the narrow bandpass system can be mitigated by removing subsystems which correspond to the stopband. Then we derive the regularized inverse system that estimates the wavelet coefficients of the input data series. Since the sampling interval of the wavelet coefficients are 2^j-times longer than that of the original signals, the proposed method can reduce the computational cost significantly.

View full abstract

The two-dimensional deformations of human left ventricles(Subjects were four normal volunteers and a patient with a hypertrophic cardiomyopathy)during the ejection period were investigated using the magnetic resonance tagging technique. The displacements were measured by tracking the intersections of tagged stripes formed within the left ventricular myocardial wall, and the strains were calculated from the changes of segment lengths of triangles defined by three adjoining intersections. It was recognized from the results for the normal volunteers that the magnitudes of minimum principal strains were almost uniform over all regions(anterior, septal, posterior and lateral)of the left ventricular wall, while the magnitudes of displacements were different from regions to regions. The magnitude of minimum principal strain in the anterior wall of the heart with the disease was smaller compared with those of the normal hearts. This study may suggest that the minimum principal strain could be an alternative effective index to evaluate the cardiac contractility.

View full abstract

The theory of design optimization based on stability maximization of a passive gyroscopic damper(PGD)is discussed and applied to the stabilization of a ropeway carrier. The design formulae maximize the degree of stability of the linearized model, which is defined as the negative of the maximum value of the real part of the characteristic roots. The new design is analyzed using the linearized model of the PGD and the nonlinear model in order to investigate the practical characteristics of the formulae. It is found that the transient responses are almost the same as those obtained from numerical optimization designs. The new design formulae are very simple, but the responses are improved compared to those obtained using previous linear designs. The formulae have been successfully used to design a V-gimbaled PGD for a full-scale ropeway carrier. The step responses of this test are presented.

View full abstract

A nuclear reactor building structure consists of shear walls with complex geometry, beams and columns. The complexity of the structure is explained in the section 'Introduction'. Seismic analysis of the complex reactor building structure using the continuum mechanics approach may produce good results but this method is very difficult to apply. Hence, the finite element approach is found to be an useful technique for solving the dynamic equations of the reactor building structure. In this approach, the model which uses finite elements such as brick, plate and shell elements may produce accurate results. However, this model also poses some difficulties which are explained in the section 'Modeling Techniques'. Therefore, seismic analysis of complex structures is generally carried out using a lumped mass beam model. This model is preferred because of its simplicity and economy. Nevertheless, mathematical modeling of a shear wall structure as a beam requires specialized skill and a thorough understanding of the structure. For accurate seismic analysis, it is necessary to model more realistically the stiffness, mass and damping. In linear seismic analysis, modeling of the mass and damping may pose few problems compared to modeling the stiffness. When used to represent a complex structure, the stiffness of the beam is directly related to the shear wall section properties such as area, shear area and moment of inertia. Various beam models which are classified based on the method of stiffness evaluation are also explained under the section 'Modeling Techniques'. In the section 'Case Studies' the accuracy and simplicity of the beam models are explained. Among various beam models, the one which evaluates the stiffness using strain energy equivalence proves to be the simplest and most accurate method for modeling the complex shear wall structure.

View full abstract

The local velocity field in a self-excited oscillating flow through a collapsible tube has been precisely measured using a fiber optic laser Doppler velocimeter. Velocity distributions along the tube axis and across the tube cross section were measured. Time-varying flow rate at downstream end of the tube and pressure drop in the tube were also measured simultaneously. The following results were obtained:In the vicinity of the most strongly collapsed portion fluid is pushed to the downstream and upstream sides and then sucked back into the distending portion as the tube wall collapses and then recovers during the oscillation, and therefore superposition of this back-and-forth fluid motion to the gently varying flow gives rise to a sharp dent at the peak of the velocity waveform at upstream and a very sharp peak of the velocity waveform at downstream. Sufficiently far upstream from the collapsed portion, flow velocity is not influenced by the wall motion.

View full abstract

To reduce noise caused by current collectors on high-speed railways, a wing-Shaped low-noise current collector has been developed. This collector head is directly supported by a composite insulator. However, it becomes difficult to maintain the predetermined contact force of the contact strip against the trolley wire. Therefore, the vertical height of those apparatus should be controlled actively. The upper part of the composite insulator is under high voltage. When using sensors only under the insulator, it is difficult to distinguish wire disturbance from lift disturbance. In this study, the difficulty is solved by using H_∞ controller to estimate those disturbances in each frequency range. In traveling experiments using full-size prototype equipment on the running vehicle, it was confirmed that the contact force variation due to lift of 68 N at the speed of 112km / h was reduced by about 33% compared with active control without lift compensation, and the wire push-up variation was reduced by about 67%.

View full abstract

The self-sensing actuator is a new concept for intelligent materials, where a single piezoelectric element simultaneously functions as both a sensor and an actuator. This concept should be advantageous in many aspects of control, especially for vibration control of flexible structures. The key component of this device is an electric bridge circuit that includes a piezoelement. This circuit could provide significant information about strain in the element if it were wellbalanced. Our aim is to use μ-synthesis to design a controller which guarantees that the performance of a self-sensing actuator is robust against perturbations in the bridge balance and to confirm the advantages of this technique from the viewpoint of control. Experimental results show that the self-sensing actuator driven by the designed controller exhibits excellent performance in suppressing the vibration of a cantilever and is robust compared to the conventional control scheme.

View full abstract

This paper is concerned with robust control for actual milling AMB spindle. We apply a method of μ-Synthesis based on the descriptor form representation to magnetic bearing system. The unmodeled dynamics and the sensitivity performance are also taken into consideration. By using this method physical uncertainties are independently treated with and a controller is designed to achieve not only robust performance but also robust disturbance suppressions.

View full abstract

This paper deals with a multiple mobile working robot, which has a six-legged locomotion function and a vertically articulated manipulator. The walking mechanism consists of a parallel link mechanism connecting two frames with three linear actuators and six extendable legs fixed to the frames. The manipulation function has 4 degrees of freedom and locomotion function has 6 degrees of freedom, thus this working robot is a redundant mechanism with 10 degrees of freedom. We propose a trajectory following control method by switching the normalizing matrix according to the freedom. The simulation results and the experimental results leads to the availability of the method.

View full abstract

Fuzzy logic was employed in the vibration control of a single-link flexible arm. The control system has two feedback loops:one basic fuzzy control(BFC)loop to control the hub position and another loop to regulate the tip point vibration by either BFC or adaptive fuzzy control(AFC). The experimental results reveal that BFC is simpler and it requires less computing time, but the gains need to be found by a trial-and-error procedure. On the other hand, the AFC can generate and modify the control rules according to the performance of the system and it demands no tuning on the control gains for different operating ranges. Both the BFC and AFC are effective in suppressing the tip vibration and for adaptation to payload change.

View full abstract

This paper presents a study of asymptotic stability for the control of flexible systems. It shows that when the controller structure is a collocated direct output feedback design with positive definite feedback gain, the number and placement of sensors and actuators are the only factors required to determine necessary and sufficient conditions to ensure closed-loop asymptotic stability. The minimum number of control devices required is dependent on the maximum number of multiple eigenvalues. The positioning of the control devices is restricted by checking several simple subsidiary rank conditions.

View full abstract

In this paper, a novel sliding mode controller is designed to track periodic reference signal and reject periodic disturbance for the periodically rotating mechanisms. The novel sliding mode controller includes an integral compensator and a low pass filter to reduce the chattering phenomena and the accumulated periodic errors of the periodic motion. The sliding condition for the control system is derived in detail. Then the proposed method is extended its application to the slider-crank mechanism and its tracking performance is obtained and compared with that of the repetitive control.

View full abstract

This paper is concerned with the problem of robust stabilization by a continuous sliding mode controller(CSMC)for systems subjected to nonlinear multiple time-varying delayed perturbations. A simple robust stability condition to test the stability of the sliding mode and a CSMC have been recently developed for multiple time-varying delayed systems. The effect of delayed perturbations in the sliding mode are discussed by seperating them into matching and unmatching conditions. This stability condition is less conservative than those derived in the literature, and chattering is removed due to the CSMC. Examples are included to illustrate our results.

View full abstract

For Clarifying the mechanism of tool and machine handle grasping, sensory tests to determine the optimum grasping diameter were performed on 43 people aged seven to fifty-three years. It was confirmed that the sensory evaluation for determining the optimum grasping diameter was difficult for children under age ten due to insufficient growth of the skeletal structure. The optimum grasping diameter was thus determined for persons above age ten. The optimum grasping diameter was found to depend on the hand size, strongly on the length of the hand but only weakly on the hand breadth or fist size. The optimum grasping diameters for the males were 30-40 mm, about 10% larger than those for the females. It was also found that a cylinder could be grasped well by a human hand if all the finger tips and the entire palm of the hand were in contact with it.

View full abstract

Our aim is to develop a microactuator for use in endoscopic-surgery. The purpose of this study was to construct cylinder-piston-type microactuators actuated by thermal expansion of paraffin. We use paraffin as the material for expansion because paraffin has a large coefficient of expansion in the temperature range of 35 to 45 degrees centigrade. For smooth actuation of the piston, we must prevent the paraffin from sticking to the inside surface of the cylinder. Thus, the paraffin is enclosed in a silastic tube. The cavity between the cylinder and the tube is filled with silicone oil for lubrication and prevention of expansion loss. The outer diameter of the actuator is 2.5 mm and its length is 130 mm. The actuator can generate a stroke of 8mm against a load of 1 kg.

View full abstract

A numerical method based on a finite volume approach has been developed to predict the rotordynamic forces on an eccentric labyrinth seal. The SIMPLER algorithm and a low-Reynolds-number κ-ε turbulence model are used to compute the complex turbulent flow field within the seal. To verify the validity of the developed method, the turbulent flow in the labyrinth seal was calculated and the resulting fluid forces was estimated. The predicted fluid forces are in good agreement with the reported measured values. It is also shown that the flow field near each labyrinth tooth must be accurately predicted to evaluate the rotordynamic forces. Labyrinth tooth throttling has more effect on the tangential force component than on the radial one. The developed code was proven to be an effective tool for predicting rotordynamic forces.

View full abstract

Two different methos have been proposed to determine the reference surfaces for the tooth contact analysis of bevel and hypoid gears. One is to use the theoretical tooth flank coordinate, for both pinion and gear, calculated from the machine settings. The other is to use the same theoretical tooth flank coordinate for the gear but the conjugate pinion. Based on these two methods, two different computer based simulation programs were developed. The transmission errors, path of contact and contact pattern were calculated and compared between the two methods, under both no load and loaded conditions. The results for the two methods were also compared with the experimental results and showed good agreement. The results for four gear sets with different composite error surfaces showed the power of the conjugate approach. However, each approach has features and problems, and thus the characteristics and scope of the two methods are compared and discussed.

View full abstract

To clarify static and dynamic properties in the circumferential direction of a tapered coupling joint, the relationship between loading torque and relative angle of torsion between a taper shank and a taper socket is determined for Morse tapers #3, #4 and #5. Static properties such as torsional stiffness of a tapered coupling joint and maximum sustainable torque are strongly influenced by the contact state between the taper shank and the taper socket. Dynamic properties such as natural frequency and damping capacity are determined using an impulsive response test of the tapered coupling joint. These properties are also influenced by the contact state between the taper shank and the taper socket.

View full abstract

In this paper, a duplex spread-blade method for cutting hypoid gears with modified tooth surfaces is proposed. The duplex spread-blade method is a rapid and economical manufacturing method because both the ring gear and the pinion are cut by a spread-blade method. In the proposed method, the non-generated ring gear is cut by a cutter in which each cutting edge is altered from the conventional straight line to a circular arc with a large radius of curvature and the pinion is generated by a cutter with conventional straight line cutting edges. Use of the circular arc cutting edges enables the application of the duplex spread-blade method in theory and produces modified tooth surfaces. The main step in this method is the determination of the cutter specifications and machine settings so as to satisfy the duplex spread-blade conditions while taking into account the positions of the tooth bearings. The proposed method was validated by using it in gear manufacture.

View full abstract

Frictional behavior in sliding pairs of pig articular cartilage and glass plate has been studied to elucidate the tribological role of constituents in a synovial fluid and a surface layer of the articular cartilage. Pig synovial fluid and water solution of hyaluronic acid were used as lubricants. The synovial fluid had a significantly superior lubricating ability to a sodium hyaluronate solution of equivalent viscosity under physiologically high load condition. The superiority of the synovial fluid seems to be responsible for the boundary lubricating ability of constituents other than hyaluronic acid. Langmuir-Blodgett films of phospholipid(Lα-dipalmitoyl phosphatidylcholine, Lα-DPPC)on the glass plate maintained low and stable friction, depending on the number of film layers. In conditions of mixed films containing γ-globulin and Lα-DPPC, the frictional behavior was improved by increasing the quantity of γ-globulin. A model of the boundary lubricating mechanism is proposed in which the effective adsorbed film is composed of proteins, phospholipids and other conjugated constituents on the articular surfaces, and controlled by hydrophobic groups in those amphiphiles.

View full abstract

In this study, to investigate mechanical remodeling of cancellous bone from the viewpoint of trabecular surface remodeling due to trabecular level mechanical stimuli, a rate equation for trabecular surface remodeling is proposed based on the uniform stress hypothesis, which suggests that nonuniformity in the local stress distribution on the trabecular surface is the driving force for remodeling. A proposed local rate equation for trabecular surface remodeling is applied to a computational simulation of morphological changes, in which the trabeculae are modeled as an assemblage of two-dimensional boxel elements and their morphological changes are simulated by removal / addition of the elements from / to the trabecular surface. The basic features of the proposed rate equation and simulation method are investigated through remodeling simulations applied to the basic trabecular-level structure, cancellous-bone-level structure and the proximal femur. In all cases, the trabecular orientation, thickness and connectivity are adapted to the mechanical environment, which demonstrates the capability of the proposed rate equation for computational prediction of the adaptation phenomenon and supports the hypothesis regarding the optimality of trabecular architecture as a mechanical load bearing structure.

View full abstract

In this paper we describe an operation planning system for process planning. An operation planning system called the computer-aided operation planning system(COPS)which generates cutting parameters(tool size, depth and width of cut, cutting speed, and feed rate)has been developed. The functional features assigned to machining operations are categorized into 8 machining features such as face milling, pocket milling, solt milling, hole making, tapping, counter boring, and counter sinking. Algorithms to carry out operation planning for those categories according to the shape of each feature, the capacity of the selected machine tool, the workpiece material and the cutting tool, were established and implemented. For supporting this software, five data files have been designed - material file, machine tool file, cutting tool file, sample cutting file, and cutting technology file - and a software program file has been made for the maintenance of those data files. The output of the established system has been proved to be practical by means of the metal cutting test.

View full abstract