In order to develop a method capable of determining the time variant spectrum of time series, various existing approaches have been investigated. Although the Fourier-based methods are superior in their computational efficiency, their inherent characteristics may sometimes limit applications. The AR method gives the best results even for small data sets. However, insufficient information is available for determining its applicability. In this report, a brief review, as well as the performance, of various AR methods applied to a certain class of stationary time series is systematically documented. The covariance method is found to be the best solution for the determination of AR coefficients, and many trials using sinusoidal data sets indicate the usefullness and applicability of AR-based spectrum analysis.

View full abstract

This paper proposes active vibration control of flexible structures using wave-absorbing method. The method is based on the active restraint of the formation of standing waves which are created in continuous vibration. The Euler-Bernoulli beam is used as the control object. Considering the general solution of the wave equation, we obtain an active control method which eliminates the reflection waves atthe boundary. The effectiveness of the restraint of the resonance is shown in frequency domains. We then apply the method to impulse responses which have wide frequency domains. Consequently, we show that the vibration of the beam is effectively and quickly controlled before the standing waves are formed.

View full abstract

In this paper, a method of simultaneous analysis of waves propagating through multilayered media consisting of gas and solid fields using the boundary element method is proposed. Usually the waves propagating in gas fields are analyzed using the Helmholz wave equation, where the unknown functions of sound pressures are taken to be scalars. Then, the sound pressures obtained as the result of the analysis of the gas field are imposed on to the solid fields as external forces and the waves propagating in the solid are been analyzed, taking the unknown functions of particle velocities to be vectors. That is, the waves propagating in gas and solid fields are analyzed separately. Naturally, it is considered to be more effective to analyze the waves propagating in gas and solid fields simultaneously than to analyze them separately. In this work, a method of simultaneous analysis of waves propagating through multilayered media using the boundary element method has been developed, using the same equation of motion for gas and solid and considering the same unknown functions of particle velocities as vectors. Also several conditions under which the solutions obtained by the proposed method converge numerically, have been clarified through numerical evaluations.

View full abstract

Here we present detailed experimental results concerning flow-induced vibrations of long-span, shell-type gates in which the upstream gate face consists of vertical and inclined skin plates (also referred to as weir plates). Such shell-type gates possess two degrees of freedom, one each in the streamwise (horizontal) and vertical directions, due to bending flexibility in those two directions. The streamwise and vertical vibrations can become closely coupled with each other through the hydrodynamic forces acting on the weir plates, resulting in severe self-excited vibrations. A two-dimensional laboratory model of a long-span, shell-type gate was operated with underflow only (i.e., no overflow) at small gate openings with several different inclined weir plate geometries, ranging from a 17.5°to a 65°inclination angle (relative to the horizontal) for the inclined weir plate. By measuring the gate's vertical and horizontal displacements, it was possible to determine the vibration frequency, the excitation ratio (negative damping ratio) and the trajectories of gate motion. These results show that the inclination angle of the inclined weir plate angle plays a significant role in determining the gate's susceptibility to this type of dynamic instability. Long-span, shell-type gates with an inclined weir plate angle of about 60°relative to the horizontal were found to be the most unstable.

View full abstract

A method of homology design is presented for linear and undamped vibration of elastic structures in eigenvalue problems based on the finite element method. It is aimed at to realize the homologous mode shape, which maintain a prescribed geometrical property in a structure both in standstill and vibrant state. The original eigenvalue problem is separated into two modified eigenvalue problems with square and rectangular coefficient matrices by introducing the homologous mode shape constraint. The governing equation of design change for homologous mode shape is derived by means of the finite element sensitivity analysis of the modified eigenvalue problems. The Moore-Penrose generalized inverse is used for the solution since the coefficient matrix of the governing equation is rectangular. The validity and versatility of the proposed method are demonstrated by the numerical examples of a planar lattice frame.

View full abstract

The analysis of problems concerning transient sounds radiated by impacted bodies using the boundary element method (BEM) is investigated, and a general-purpose analysis program which is valid for any type of sound source is developed. The Laplace transformation of the Helmholtz equation is employed in the formulation of the BEM. The analysis is first carried out in the Laplace domain by using the boundary input values obtained by a numerical Laplace transform. Then, results in the time domain are obtained by a numerical inverse Laplace transform. As an example of transient sound radiated by impacted bodies, the transient acoustic radiation emanating from a column and a cube impacted by a sphere are considered. The validity of these analytical results for time-dependent sound radiation is examined through comparison with the experimental results.

View full abstract

Internal resonances of cracked rotors are studied using rotor systems with two degrees of freedom. Phenomena at the major critical speed and other subcritical speeds in the precritical speed range are discussed theoretically and experimentally.Phenomena in a deflection (lateral) oscillation model with internal resonances and an inclination oscillation model with no internal resonance are compared. In the former model, critical speeds of a forward and a backward whirling modes coincide and, as the results, the following phenomena occur: (a) The vibration of a synchronous backward whirling mode changes from a steady-state oscillation to an unstable oscillation due to an internal resonance. (b) Superharmonic resonances of 2nd and 3rd orders of backward whirling modes and a super-subharmonic resonance of 3/2-order of a backward whirling mode occurs due to an internal resonance.

View full abstract

The design of near-optimal observer-based controllers for singularly perturbed discrete systems is investigated in this work. The near-optimal control is achieved by combining optimal observer-based controls of the relative fast and slow subsystems.The minimum value of the performance index when the near-optimal control is used is O(ε²) close to that of the full-order optimal control. The near-optimal trajectory is O(ε) close to the full-order optimal trajectory. The effect of the singular perturbation parameter ε on the stability problem is also examined. Finally, a twin-engine aircraft model is used to illustrate the theoretic analysis.

View full abstract

In this paper, the author proposes a method for improving control performance in feedback control systems by introducing a multiple eigenvalue filter which has been deduced from parallel learning models. First, the control system model is copied to i (i=1, 2, ..., k) systems corresponding to learning times k. The actuating signal of the first model is added to the actuating signal of the second model, and then the actuating signal of the second model is added to the actuating signal of the third model. Likewise, the actuating signal of the k-1-th model is added to the actuating signal of the k-th model. The thus obtained k-th model is equivalent to a system which has a filter as a series compensator that is composed of the sum of i (i=1, 2, ..., k-1) multiples of the left side of the characteristic equation. In this paper, the sum is called "multiple eigenvalue filter" and it is concluded that the filter can eliminate control variable deviation without losing stability when disturbance is imposed.

View full abstract

This paper deals with the eigenvalue problem of a multiple eigenvalue filter which was proposed in the previous paper. The system with the proposed filter derived by means of parallel learning not only reduces control variable deviation without losing stability, but also has PID (Proportional, Integral and Derivative) control action although the original control system before learning only has just proportional control action. In this paper, the reason for what is shown from the viewpoint of an eigenvalue problem and Nyquist stability criterion.

View full abstract

This paper deals with recursive formulations of flexible bodies connected by rotary joints based on the method of weighted residual. As the state variables, the translational velocity of the base body, the angular velocities of the bodies, and the modal amplitudes of elastic deformation of the bodies are employed. Using two sets of weighted functions, two types of mathematical models of the system are obtained in recursive forms; The first one is an extension of the model of a rigid multibody system based on the Newton-Euler's method. The second one is the same as the model based on the Lagrangian formulation with quasi coordinates. The recursive algorithms of the inverse dynamics of a flexible manipulator are derived using the models obtained.

View full abstract

This paper reports the kinematic characteristics of the mass center of the entire human body while ascending and descending stairs, using large force plates which can measure both three-dimensional force components between the feet and the stair and the point of application of the force. Kinematic characteristics in terms of the velocity and the displacement of the body center in the three-dimensional space can be obtained by numerical integration of the force components with respect to time after proper calibration of measured data. For the calibrations, two methods were use: the first one compensates the measured force according to the sensitivity of the force plates due to their size were large; the second one tunes integration constants to avoid drifting features in integrated data. The following results are obtained: our new method, i.e., obtaining kinematic characteristics for stair gait using force plates, gives reliable results in the three orthogonal directions in space; characteristics in the sagittal and the vertical directions change depending on gait cadence; characteristics in the lateral direction are similar to those observed in level walking.

View full abstract

To protect precision equipment from microvibrations, it is important to be able to measure the microvibration sensitivities of these machines. As yet, however, no measuring method has been developed. Conventional vibration testing devices cannot generate microvibrational waves under 10 Gal. The authors have developed a microvibration testing device that employs active control technologies. The trial device, which comprises pneumatic-type actuators and a new DSP (Digital Signal Processor) controller, successfully generates sinusoidal microvibrational waves at a magnitude of 0.1 Gal on a vibrating floor. These waves were realized with a noise rate of less than 10%. This paper outlines the developed system and describes some of the experimental results.

View full abstract

Multiple elevator doors are driven with one servomotor through a multi-degree-of-freedom linkage system. This paper deals with the derivation of the equation of motion of the door mechanical system. The movement of the doors is described by giving various constraint conditions to multiple single-chain and multi-degree-of-freedom rigid-body systems. The door motion was simulated based on this formulation. The conclusions are as follows. [1] It is clarified by computer simulation that the change in the constraint condition caused by integration error is extremely small. [2] The method of obtaining the motor driving torque, in which a given movement of the door or the motor is regarded as a constraint condition, is proposed. [3] A mutual conversional relations among the joint torques, door driving force, and motor torque is shown.

View full abstract

Cooperative manipulation by autonomous intelligent robots promises fertile potential for improving the ability of robotic manipulation and expanding the domain where robots may be used. This paper focuses on the method for calculating the optimal force of each robot based on local sensor feedback and high-level communication. First, we propose adaptive force direction control, which is useful for monitoring the state of force equilibrium, and then describe the theoretical background of this control method. Second, we describe the strategy to find the proper force using our proposed control method. We also describe a control system architecture which is suitable for these operations. In this control system, each manipulator motion function, each sensor function, and each message-passing operation are programmed as primitive tasks which run in parallel. Third, we illustrate the usefulness of our proposed method with some experiments such as tumbling operation by two autonomous robots.

View full abstract

We developed an electrostatic linear actuator for use as an artificial muscle that consists of parallel-plate capacitors and a slider made of dielectric material. We must integrate many unit actuators to generate a large output force. To integrate the unit actuator, we attempted to determine the optimal aspect ratio, i.e., distance between two electrodes of capacitor/length of electrode, which represents the integration rate of the actuator. It became clear through the computer simulation and the experimental measurement that the optimal range of the aspect ratio is from 1.6 to 1.7. We made a two-dimensional integrated actuator with twenty pairs of electrodes whose aspect ratio was 0.8. It was confirmed by the measurement of force that the experimental values agreed well with the corrected theoretical ones. The maximum output force of 42.3 [mN] was generated at the supply voltage of 200 [V].

View full abstract

An enveloping grasp is stronger than other grasps. This grasp, however, has a disadvantage that it is difficult to find contact points and contact forces. In this paper we discuss the identification of unknown parameters, which are contact points and contact forces, in the enveloping grasp.

View full abstract

End milling of a sculptured surface is attempted employing an articulated robot with six degrees of freedom. In this application, the machining error due to the positioning accuracy and elastic deformation of robots should be considered. In order to improve positioning accuracy, a new robot calibration method has been developed and applied. After the calibration, the robot is operated by off-line teaching programs effectively, and experimental end milling is finished with improved machining accuracy. Furthermore, a postprocess error compensation has been realized to minimize the machining error caused by all of factors such as the positioning error, elastic deformation and joint backlash of the robot. Experimental end milling shows that the machining accuracy is improved to near the positioning repeatability of the robot after repeating compensation twice. Additionally, a sensor feedback teaching method which is another way to achieve end milling of a sculptured surface has been developed.

View full abstract

Kinematically redundant manipulators possess a number of advantages over conventional manipulators, including superior performance with respect to workspace obstacles and mechanism singularities. These manipulators also have the potential for providing better dynamic performance than the current manipulators. This paper presents a very efficient means of including dynamic effects in the obstacle avoidance control of redundant manipulators. The approaches are based on the decomposition technique, in which a manipulator with high redundancy is first decomposed into nonredundant systems (and/or subredundant ones with only one or two degrees of redundancy) and each of these is analyzed using a well-defined Jacobian matrix taking into account obstacle avoidance and manipulator dynamics, thus constructing a set of possible solutions for obstacle avoidance. Numerical simulations were performed to show the effectiveness of the proposed approaches.

View full abstract

In order to develop "Active Human Interface, (AHI)" that realizes emotive communication between intelligent machine and human being, we have been investigating machine recognition of human emotions from facial expressions. Using a neural network (NN), we have already obtained a high correct recognition ratio of about 90[%] for 6 basic facial expressions, but we find that, depending on the NN learning information, there exist certain characteristics in the NN misrecognition and correct recognition. This paper analyzes these characteristics using action units (AUs) employed in the psychological field for describing the basic muscle movements of human face and analyzing the relationship between the NN input and output information. We clarify, in this paper, the reason why misrecognition occurs by using the Jacobian of the NN input-output function and how AUs influence the recognition results through sensitivity analysis done by assigning a zero value to each AU used in NN input information.

View full abstract

Hard coatings consisting of TiN films were deposited onto high tensile steel by ion mixing. Three laminated layers were prepared on each specimen, with variations of TiN and Ti₂N phases. The effect of the laminated layers on adhesion, friction and wear behavior was studied. The investigation methods included of scratch testing, pin-on-disc wear testing, and dynamic micro-hardness measurement. The laminated layer does not significantly affect the adhesion for the hardness of the film. The unlaminated films have stronger adhesion, showing that interfacial bonding between the layers weakens the overall adhesion of the coatings. A Ti₂N phase laminated layer improves the adhesion slightly when it is either the top layer or at the substrate-coating interface. The TiN and Ti₂N phases have different deformation mechanisms during the scratch test. The steady-state friction coefficient does not vary significantly with sliding up to a distance of 400 meters. High sliding speeds cause more severe pin-on-disc wear.

View full abstract

This paper presents a study on the friction and wear characteristics of gears in high vacuum. A power-circulating-type gear testing machine with a vacuum chamber was developed, in which the coefficient of friction between the contacting tooth surfaces, and the temperatures of the gear tooth and the shaft can be measured simultaneously under various running conditions. The running tests for S 45 C, SUS 440 C and MoS₂-coated SUS 440 C gears were carried out in high vacuum and air environments by using the gear testing machine, and the coefficient of friction between the contacting tooth surfaces, the wear of the gear teeth, and temperature changes of the gear teeth, the shaft, etc. were measured. On the basis of these results, the friction and wear characteristics of gears in high vacuum were determined.

View full abstract

Oil film seals of large diameter sometimes show long-period cyclic deformation of the floating seal ring whose rotation is constrained by a stopper pin. Cyclic change of deformation and oil film temperature may cause fatigue damage of the seal ring after long-term operation. We present a theory for the mechanisms of thermal deformation and its periodic transformation, the results of which agree well with those of measurements, and which provides a practical design criterion to prevent this phenomenon.

View full abstract

To solve the tribological problems such as wear and loosening in existing knee prostheses composed of ultrahigh molecular weight polyethylene (UHMWPE) and metal or ceramics, a new design was attempted to realize adaptive multimode lubrication in the artificial joint. For this purpose, artificial articular cartilage is required, which has a similar structure and function to natural articular cartilage. In this study, PVA (polyvinylalcohol) hydrogel was used as artificial cartilage in the total knee prostheses. A PVA hydrogel layer was placed on an articulating surface of the tibial component, and their lubrication property during walking was evaluated by the knee joint simulator test. Silicone rubber (another compliant material) and UHMWPE were also tested and their frictional behaviors were compared with that of PVA hydrogel. As a result, PVA hydrogel showed suitable frictional behavior as artificial cartilage.

View full abstract

This paper proposes a method of application of micro amounts of adhesive resin by using a needle. The process of micro-application of the resin is experimentally investigated. The amount of adhesive resin on the needle is influenced by the shape and size of the needle. The micro-application of adhesive resin is dependent on the handling process of the needle to transfer adhesive resin on the adherent surface. The design parameters of mechanical systems for micro-application of adhesive resin is discussed.

View full abstract

Sliding accuracy characteristics such as the variation of floating height and the pitching-mode angular fluctuation of an externally pressurized gas-lubricated pad bearing traveling on a linear guide way are investigated theoretically with special attention paid to the relationship between the sliding accuracy and the machining error of the guide-way rail, where sinusoidal undulation of the guide-way surface is assumed. The effects of pitch length of the undulation of the rail surface on the sliding accuracy characteristics are discussed. The experimental results verify the validity of the theoretical analysis of the sliding accuracy of the bearing pad.

View full abstract

An optimal space partitioning method is proposed based on rectangular duals of planar graphs and a simulated annealing algorithm. The layout problem, in which a region should be partitioned into plural subregions of layout components so as to satisfy relationships between neighbors and size conditions for a whole region and respective subregions, occurs in several layout designs. Such layout problems are characterized by the combinatorial property on topological structure among subregions. In our method, rectangular duals of planar graphs are used for representing such a structure, and it is optimized through a simulated annealing algorithm. In the annealing process, topological layouts represented by rectangular duals of planar graphs are manipulated with rules, each of them is embodied into an actual layout using the generalized reduced gradient method which is one of the numerical optimization techniques for constrained nonlinear optimization problems. Finally, we show an example of an access control room layout in a power plant design in order to check the effectiveness and validity of the proposed method.

View full abstract

This research work deals with the development of a simple and practical method for manufacturing high aspect ratio micro mechanical parts using three methods. The first method is mask-based method in which an image is transferred to a liquid photopolymer by irradiating a UV laser through a patterned mask. The irradiated portion of the photopolymer is then solidified and it becomes a high aspect ratio polymer structure. In the second and third methods, high aspect ratio polymer structures are produced using a shaped UV laser beam writing and a focused UV laser beam writing, respectively. The accuracy of the solidified polymer when using these methods is examined. Various shapes of micro polymer structures are produced using these methods.

View full abstract

This paper deals with fundamental research on the development of human-oriented interface for CNC machine tools which allow workers to feel pride and pleasure in making full use of their ability in manufacturing. As a first step toward human-oriented CNC machine tools, three information feedback channels designed to provide workers with much useful information on machining phenomena have been proposed. They are the machining phenomenon visualizing system (MPVS), the machining vibration display dial (MVDD) and the machining conditions display system (MCDS). These systems have been evaluated by inexperienced and skilled operators of CNC machine tools using the information integration method. The results show that the combination system of MPVS, MVDD and MCDS greatly improves workers' satisfaction in CNC lathe work.

View full abstract

This paper deals with an aluminum-clad stainless steel sheet produced by hot rolling. Basic press-forming tests, which include punch stretching and deep-drawing, were conducted. The characteristics of the joining process within the range of uniform elongation shows that the aluminum portion has higher ductility than that of stainless steel. However, the latter has higher local deformation and a larger Lankford, r, owing to the inherent characteristic of the material. The formability in both stretching and deep-drawing, are higher when the aluminum is set on the outer side of the cup. Factors related to the deformation of the clad sheet, such as wrinkling and cup height are controlled by the material with the higher strength, stainless steel. The fracture and deformation mechanisms are clarified by the detailed measurements of thickness strain.

View full abstract

In scientific computing, developing effective methods to display complex phenomena represented by numerical data is a key issue. In this study, virtual reality technology is applied to develop a multisensory data sensualization environment in which the user can use visual, acoustic, and touch sensations to understand scientific data. In particular, a prototype wind sensation display system was developed to generate touch sensation. In the multisensory data sensualization method, accurate and efficient transmission of data from the computer to the user is expected through the effective integration of several sensations. We conducted several experiments on the perception of scalar data and vector data in three dimensional space, using multisensory data sensualization methods, and verified their effectiveness in providing the user correct perceptions of data.

View full abstract

Opposed dies shearing process proposed by one of the authors is based on the idea of smooth sheared surface formation utilizing a cutting mechanism and has excellent features not found in other shearing processes. This process involves two stages; after sufficient shearing between die and protruding die, a product part is separated successively by a knock-out punch, where knock-out timing is important to prevent fracture of the surface and burr formation. Therefore, this process requires use of a precise hydraulic press which facilitates this complicated action and exact knock-out timing. This requirement seems to prevent wide practical application of this process. Therefore, we propose a simplification of Opposed dies shearing process utilizing differential pressure. First, this idea was tested using a coil spring substituted for hydraulic pressure and its feasibility was confirmed. Secondly, a differential pressure mechanism was realized with a hydraulic press and straight shearing of mild steel was achieved. Basic phenomena of this method were studied and it was found that knock-out timing was controllable by varying the pressure difference. Furthermore, a die set unit utilizing this mechanism was developed and disk blanking of aluminium with a mechanical press was demonstrated.

View full abstract

The purpose of the present research is to establish a simple method of forming holes in brittle materials by a conventional press-working. By utilizing a specially designed blunted spherical punch, a small hole is formed at the center of the specimen at its first stage, and subsequently the hole is enlarged by a continuous punching operation. The present improved technique realizes a very low load for punching operation as well as a considerably low clamping stress for preventing serious radial cracks around the hole. Some punching works in engineering ceramic plates were performed by the method for various aspect ratios AR=3∼32(AR=D/t, D; the diameter of the hole and t; the plate thickness). As long as the AR value is larger than about 3, the present method proved to be very successful in forming holes in tough engineering ceramics such as alumina, silicon-nitride and toughend zirconia, as well as in making small holes less than 1 mm in diameter. The technique is also successful in forming many holes by a single operation.

View full abstract