Static and dynamic characteristics are experimentally investigated for annular plain seals with porous materials applied to the seal surface by insertion into the middle of the seal. Experimental results show that annular plain seals with porous materials have a higher leakage flow rate, larger main stiffness coefficients, and smaller cross-coupled stiffness coefficients and main damping coefficients than conventional annular plain seals with solid surfaces. In the porous seals, an increase of approximately 30% in the leakage flow rate and reduction of the same amount in the main damping coefficients are obtained, whereas the main stiffness coefficients for the porous seals are four to six times those for the solid seals due to the increase in the hydrostatic force induced by a function of the hydrostatic porous bearing. This suggests that the quantitative effects of the porous materials on the main stiffness coefficients are much more significant than the effects on the leakage flow rate and the other dynamic coefficients. The larger main stiffness coefficients for the porous seals yield larger radial reaction force for a small concentric whirling motion, which would contribute to rotor stability from the viewpoint of increasing speed limit by a stiffer rotor support.

View full abstract

Harmonic resonances and subharmonic resonances of order 1/3 in a continuous rotating shaft with distributed mass are discussed. The restoring force of the shaft has geometric stiffening nonlinearity due to the extension of the shaft centerline. The effect of a distributed lateral force, such as the gravity, is assumed. The possibility of the occurrences of harmonic resonances and subharmonic resonances of order 1/3, the shapes of resonance curves, and internal resonance phenomena are investigated.

View full abstract

A method for calculating the response of nonlinear spring-mass systems subjected to transient force, using the convolution integral is proposed. The response of systems with a localized nonlinear spring is represented in the form of the convolution integral of the nonlinear restoring force which is treated as an external force and can be easily calculated by the forward substitution method after the discretization of time with a small interval. The response of systems connected by a nonlinear spring can also be calculated using impulse responses of subsystems before the connection. Examples of a 4-degrees-of-freedom system with a localized nonlinear spring and of a system in which two 2-degrees-of-freedom systems are connected with a nonlinear spring and a damper are shown. The result of using the convolution integral becomes valid when using a smaller time interval than in the Runge-Kutta-Gill method.

View full abstract

It was shown in the second report that a collocation method using analytic solutions can be applied to the analyses of bending vibrations of plates with various shapes and that it yields accurate results. This means that the bending vibrations of plates, which have been treated by various methods of approximation so far, can be analyzed by only the simple collocation method. However, some problems remain. In particular, a marked change of determinant in a frequency equation makes calculation unstable. It remains unknown how to treat reasonably the boundary conditions at corner points and how to determine degenerated natural frequencies. In this paper, improvements of the calculating procedure concerning the aforementioned problems are shown.

View full abstract

The transient state of parametric resonance can be analyzed using the amplitude equations, which are obtained by the method of multiple scales and the averaging method. Using the analysis, the growth of the amplitude and the time variation of the phase angle have been theoretically clarified. On the other hand, to our knowledge, no experiment has been conducted in order to examine the above state of the parametric resonance. The reason seems to be that FFT, which is widely used in the spectral analysis, is not an appropriate method for transient phenomenon analysis. In this research, we propose a method for analyzing the transient state of the parametric resonance of Duffing type, by applying adaptive spectral analysis. The growth of the amplitude and the time variation of the phase angle are experimentally clarified. By comparing with the theoretical results obtained from the amplitude equations, we show that the adaptive spectral analysis is effective for experimental analysis of the transient state of the parametric resonance of Duffing type.

View full abstract

In this paper, we discuss a self-excited driving system for a vibratory machine which is excited by electromagnets. The open-loop transfer function of an electromagnetic-type vibratory machine has a 180-degree phase lag at the natural frequency (phase crossover frequency). However, in the absence of coil current, no self-excited vibration is generated because the gain of the open-loop transfer function is zero at the phase crossover frequency. This zero-gain phenomenon is derived from the second-power nonlinearity of an electromagnet. To avoid the zero gain in the electromagnet, a half-wave rectification circuit with a diode is added to the vibratory machine. Self-excited vibration is generated by displacement feedback in the vibratory machine which has a half-wave rectification circuit and this driving system follows the resonance frequency change automatically.

View full abstract

We investigate the shock vibrations of a sandwich beam having a viscoelastic core and elastic layers subjected to inpulsive excitation. In the analysis, Young's modulus of the viscoelastic core is assumed to be small compared with that of the elastic layer, and the response to the excitation and the frequency equation are derived. In numerical calculations, effects of the thickness and stiffness of a viscoelastic core on the response and loss factor of the system are examined. A series of experiments were carried out on the sandwich beam subjected to impulsive excitation. Measured results are in good agreement with the theoretical results obtained by the method in this paper.

View full abstract

This paper is concerned with the sway control system for rotary cranes during swing operation. The sway control system in lever operation is successfully performed by developing a sway angle sensor and a sway controller. The proposed sensor is composed of strain gauges and a FIR filter which can measure the sway angle with time-lag using boom strain. The load sway is regulated by applying the regulator theory to a linearized crane model. It is confirmed by comparison of analytical and experimental results that the proposed sway system shows excellent performance and allows easy operation from a practical viewpoint.

View full abstract

This paper is concerned with the sway control system for rotary cranes during boom hoisting operations. The sway control system in lever operations has been performed by developing a sway angle sensor and a sway controller, and the load sway is regulated by a feedback control, as reported previously. A sway control system during boom hoisting operations is developed by applying the regulator theory to a linearized crane model, which consists of the feedback of boom hoisting and sway angles, and their derivatives. It is confirmed by comparison of analytical and experimental results that the proposed sway system during boom hoisting operations shows excellent control performance.

View full abstract

Active suspensions have been considered for improving the ride quality of railway vehicles. In this study a combined control by preview and H_∞ control is considered for active suspension of railway vehicles. The control effect on yawing motion and the effects on other motions are confirmed by simulation using a full vehicle model with 17 DOF. In real railway vehicles, creep coefficient between wheels and rails varies according to the condition. Robustness of the control especially of creep coefficient is also confirmed by simulation. Furthermore, the number of equations which estimate the effect of combined control system is reduced and used for checking the simulation results.

View full abstract

In this paper we describe the estimation theory of sound power radiated from a circular plate with a solid shaft and a solid cylinder, which is a simple model of an aircraft gear blank. The forced vibration of the circular plate with a solid shaft and a solid cylinder was solved by assuming that the displacement of forced vibration is equal to the linear sum of the eigenfunctions of the displacements of the circular plate part and the cylinder part. Using the obtained result, the equations for calculating the frequency spectrum of the three important parameters for estimation-radiation loss factor, driving point mobility and total loss factor-were derived.

View full abstract

In this paper we describe the estimation result of sound power radiated from a circular plate with a solid shaft and a solid cylinder, which is a simple model of an aircraft gear blank and is excited axially, radially or three dimensionally. The three parameters for estimation, the radiation loss factor, the driving point mobility and the total loss factor, were calculated according to the formulae derived in the first paper. Using these calculated parameters and measured excitation force spectrums, the sound powers radiated due to the excitations were estimated and compared to the measured values obtained by means of the sound intensity method. The estimated frequency spectrums of sound radiation power coincided with the measured values. Consequently, the estimation theory and the associated formulae are valid.

View full abstract

A load-sensing hydraulic system is one of the most effective means to save energy. Such a system has been successfully used in construction machinery and injection molding machines. This kind of system, however, has demerit that it often falls into instability, which may be caused by the interaction between the dynamics of the load-sensing mechanism and that of the load. It is important, therefore, to investigate the dynamic characteristics of the system. In this study, both experiment and simulation are performed in the case where the friction characteristics of an actuator strongly affect the dynamic characteristics of the system. This system is simulated using bond graphs, and computation is carried out using a program named BGSP which was developed in Japan. By comparing experimental and computed results, a suitable model of the system is obtained. Using this model the overall efficiency of this system is calculated, taking into consideration the dynamic characteristics of the system.

View full abstract

The present paper shows the dynamic responses of a rarefied gas lubricated slider bearing on a magnetic recording disk. We have already presented the scheme of the computation and the results without rolling movement. If there is any irregularity on the disk, such as a bump or a grain of dust, rolling will occur. The object of the present study is to simulate the behavior of the head slider in this case. Results of the analysis are shown by dynamic response diagrams of a head slider passing over a bump. The adequacy of the computation and the scheme has been confirmed by comparison with results shown in [Hayashi and Ohkubo, Trans. Jpn. Soc. Mech. Eng., 55-511, C(1989), 743].

View full abstract

The purpose of this study is to develop a system that can regulate the parameters of Tchebycheff iteration method in order to accelerate the convergence of the approximate solution toward the exact solution. A self-tuning regulator was adopted in the present study and the optimum parameters in the Tchebycheff iteration method are determined according to the results of identification. The diffusion problem with the convection term was used as a test problem and the present method was compared with the CGS method, which is regarded as one of the most efficient methods. The memory capacity required for the present method was much smaller than that of the CGS method, but the calculation results showed that the present method required almost the same CPU time as the CGS method.

View full abstract

In this paper, we present a new deconvolution method for a time series, 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 effects of ill-posedness due to the narrow bandwidth of the 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 rate of the wavelet coefficients is coarser than that of the original signal, the proposed method can reduce the calculation cost extremely.

View full abstract

Fluidized bed incinerators can burn waste in a short time by stirring waste and fluidized sand. However, it is difficult to stabilize an incinetor plant without operators in abnormal situations and to obtain exact mathematical models. Moreover they tend to change their characteristics. Thus, it is necessary for the plant control to realize an intelligent control method which stabilizes plants when abnormal situations occur, which are not only expected but also unexpected and when their characteristics change. As a solution to such a control problem, we present a method of intelligent control using "cubic neural network" (CNN) which possesses multi-levels of different degree of abstracted signals and is capable of parallel distributed signal processing. As an unexpected abnormal situation, we consider sensor failure. The control ability of this method was verified by experiments on an actual plant.

View full abstract

It is assumed in an estimation problem with an observer that all inputs to a system can be measured and are available. However, there exists a system such as a boiler heating surface in which the observer causes estimated errors because the input called heat absorption is unmeasurable. This paper presents a solution to such a problem. The solution is based on the suggestion of the observer which acts as a feedback control that tracks its output to the system output and is found in the method of adding a feedback loop from the tracked error to the observer input corresponding to the unmeasurable input. The estimated input coincides with the unmeasurable input with infinite feedback gain which is realized by pI operation in an actual field. Favorable results are obtained in the application to boiler heating surface as shown by stability analysis and simulation.

View full abstract

It is assumed in an estimation problem with an observer that all inputs to a system can be measured and are available. However, there exists a system such as a boiler heating surface in which the observer causes estimated errors because the input called heat absorption is unmeasurable. The authors proposed, in a previous paper, a solution to such a problem for linear systems. This paper presents an extension of the proposed solution to nonlinear systems such as the actual boiler heating surface obtained by applying fuzzy controls to the feedback loops in the observer that track its output to the system output. The authors describe an example for the final superheater in a supercritical pressure once-through boiler and favorable estimated results obtained by simulations for both state variables and the unmeasurable input.

View full abstract

A nondestructive testing method is considered for a flexible one-link manipulator with a robot hand. First, a mathematical model is described by an Euler-Bernoulli beam with a tip mass at its free end, taking into account the dynamics of DC servo motor. The mathematical model which involves spatially varying coefficients due to material defects of the manipulator is presented by the finite Galerkin method using B-spline functions. Second, parameter estimation problem is solved for the model without material flaw. Third, the output of the model with flaw is evaluated using the parameter values of the model without flaw. A detection algorithm is proposed for estimating defective regions based on the evaluations mentioned above. Computational results are summarized with experimental data obtained by an optical encoder.

View full abstract

In this study, we examine the role of three force scaling laws in the performance, by human operators, of scaled teleoperated pick and place tasks. In the experiments we used a hybrid hardware-software telemanipulation system with force feedback. Human subjects were provided with visual and various types of force feedback in order to perform the desired task. The force feedback depended on the scaling law used. Our results show that impedance scaling and time scaling improved the performance with respect to constant (i.e., geometric only) scaling or no force feed-back.

View full abstract

This paper is a study on the hybrid force/position control of a two-fingered miniature gripper driven by piezoelectric bimorph cells. The system is composed of two flexible cantilevers and a compact force sensor attached to the tip of one finger. Control action is applied by two piezoelectric bimorph strips placed at the base of each finger. Both the classical PID control and the H_∞ control algorithm are examined for the control problem in which the gripper is commanded to grasp an object with a time-constant force 0.01 N at a prescribed fingertip position. Furthermore, a new PID controller based on the H_∞ theory is proposed and applied to the control of the gripper. It is found that the modified PID controller is simple and shows good robustness for controlling and driving the gripper.

View full abstract

In this paper we present mathematical models for hydraulic actuators which consist of a reciprocating pump driven by a multilayered piezoelectric element, a sigle-rod cylinder, a variable throttle valve and an accumulator, where the control input is the amplitude of rectangular wave voltages applied to the piezoelectric element. The mathematical model of the actuator was derived by considering the compressibility of the working liquid, the nonlinear characteristics of the piezoelectric element, the time delayed motion of the pumping valves, and the viscous and coulomb frictions of the cylinder. In order to confirm the validity of the proposed model, various simulations were carried out for the output power of the piezoelectric pump, the dynamic behavior of the actuator and the performance of the position control system, and good agreement with the experimental results was obtained. These simulations show that the precise description of pumping valve characteristics is crucial to the evaluation of actuator performance.

View full abstract

Robot for use in welfare or agriculture should be light in weight and have redundancy of freedoms. A rubber artificial muscle (Rubbertuator) is an actuator with the above properties. We propose a two-degree-of-freedom artificial rubber actuator. The proposed actuator is composed of two types of actuator: two Rubbertuators (R-R type) which move a joint antagonistically, and a combination of a Rubbertuator and a spring (R-S type) which also moves antagonistically. The effectiveness of the control algorithm is examined using on R-R type or R-S type actuator. To determine the basic characteristics of the above state actuators, experiments were executed (for R-R type and R-S type). SAC with a PID compensated algorithm and that with an adaptive disturbance compensated algorithm were applied to the state actuators and several evaluations were made based on the experiment results.

View full abstract

This paper deals with a real-time planning method for transportation/regrasp of a large object by multiple robots (a robot group). Real-time performance can be realized by iterating two-step sub-planning in one sampling interval: (1) motion planning of the object, and (2) that of the robot group. The virtual impedance method, which was proposed by the authors, is extended to sub-planning (1). For sub-planning (2), nonlinear programming is used to determine the optimal distribution of the robot group with consideration of obstacle avoidance and stable grasping. Simulations are performed for transport of a rectangular object by three robots. The effectiveness of the proposed method is shown in terms of evaluating distance between the robots and obstacles and grasping stability.

View full abstract

This study proposes a method for obtaining a feedforward input for motion control of a brachiation robot which has a nonlinear property in its dynamics. We deal with the brachiation robot which has three links with two actuators and no actuator at the root of the first link. We regard the brachiation robot with a nonlinear property as a time-varying system and apply the error learning method of final-state control. We especially adopt an iteration method which uses a learning coefficient less than 1 to improve the convergence property for errors of final state. By simulation taking into consideration the nonlinear property of the brachiation robot, we verify that the desired final state is achieved using the feedforward input obtained. As a result, we see that the iteration method using error learning is useful for motion control of the nonlinear system.

View full abstract

Cooperative control approach between human and robot plays an important role in carrying out diversified tasks which vary according to the work situation. The degree of participation of the human operator must be determined according to the work situation; therefore, a flexible cooperating process which allows varying degrees of direct human control and autonomous robot control is required in order to achieve successful cooperative control. This study proposes such a process using fuzzy reasoning to achieve the cooperative control. The degree of human control and autonomous robot control is determined according to the working strategy and working circumstance in this system. Some contacting tasks and avoidance motions are executed for various object walls using a two degrees of freedom Cartesian robot. The results indicate the validity of the proposed cooperative method for cooperative control between human and robot.

View full abstract

This paper deals with a modeling and controlling strategy for multirobot group behavior. While the distributed autonomous robot system has the advantage of flexibility, it also has difficulties in coordinating global system such that the designer expects. We present a methodology for the designing issue for group robotics. Self-organization is a natural principle in a dissipative system. Utilizing this principle, we present the top-down and bottom-up hybrid approach and a basic model for coordinating macro-scale group behavior as a framework for the self-organizing control of group behavior. Then we present the model of strategy formation in the system based on the opinion formation model. The analytical predicted group behavior is verified by simulation results. In particular, we realize self-organization of temporal behavior pattern in group-level cooperation.

View full abstract

This paper reports the 3D force feed back device for CAD system with the adaptable control for human characteristics. For the problem of unstable motor action of force feed-back device caused by disagreement of impedance between man and machine, the human friendly device is proposed with adaptive control using parameter tuning for device controller. One hypothesis of proportional relation between grasping force and muscle tension is discovered. The experimental result shows the unstable motor actions caused by higher muscle tension. From the hypothesis and experimental results, the system parameters are unfolded on grasping force, and an adaptive control method is described. As an evaluation experiment of proposed device, a comparison is examined among tracing-resistances of virtual surface and real surface (ice, steel, and wood), and shows smooth feeling rather than ice plate.

View full abstract

We have developed an intelligent mobile robot for service use, which can be utilized as a "secretary-or helper-robot" by daytime and "robot for security guard or maintenance including floor cleaning" by night in office buildings. An autonomous locomotion control system of this service robot is organized by four functions, i.e., path planning, recognition of absolute position, locomotion control and learning navigation controller. In this paper, in order to move more appropriately and smoothly in environments including human beings and obstacles, we have examined the learning function by means of fuzzy neural network which tunes up membership functions for fuzzy locomotion control. By this method, the intelligent robot can learn the moving area autonomously. Results obtained by computer simulation show that the proposed method is very useful for autonomous locomotion control of the intelligent robot.

View full abstract

This paper deals with a practical method for the realization of nonlinear and linear PID controls using fuzzy control methods. In this method, not only nonlinear PID control but also linear PID control can be easily realized by applying the simplified indirect inference method to the fuzzy control rule. Therefore, it is first shown that nonlinear PID control based on the PID controller can be realized using this method. Secondly it is shown that linear PID control can also be realized simply by replacing an arctan type with a line type in the antecedent part of the fuzzy control rule.

View full abstract

The effects of an electric field on lubricating ability of synovia constituents were examined using a reciprocating friction tester for a sliding pair of a conductive silicone rubber cylindrical surface and a stainless steel plate. Water solutions of sodium hyaluronate (HA) with different concentrations of human serum protein (γ-globulin) were used as lubricants. For 0.5wt% HA solution with 0.3 wt% γ-globulin, marked decrease in friction is observed by means of extenally applied low-frequency alternating current (AC3V, f=10 Hz, sinusoidal wave) in comparison with high-frequency alternating current. Based on atomic force microscopy (AFM) observation for stainless steel plate after testings, the adsorbing behavior on rubbing surfaces seems to be related to the line of the electric force and the electric power. The increase in concentration of γ-globulin changes the frictional behavior and imposes restrictions on lubricating-ability control by means of an electric field. These results suggest the electric field affects the adsorption to frictional surfaces and the movement of constituents of the synovial fluid between surfaces. In particular, the application of a low-frequency alternating current to rubbing surfaces may produce the optimum lubricating condition.

View full abstract

The penalty rigid plastic FEM program that was developed in our previous report is applied to reduce the number of stages in a three-stage forging process of an axisymmetric part, which was simulated in the previous report. With the aid of numerical simulation by the use of this program, a two-stage forming process is designed. Furthermore, practical forming is carried out in order to confirm that forging accuracy and long tool life can be obtained using this process. Results of the investigation show that the FEM-aided design of multistage forging processes is effective in reducing the number of stages and optimizing the process.

View full abstract

It is very difficult to machine the side surface of a thin wall workpiece with precision using an end mill because the deflection of the thin wall workpiece due to the cutting force is larger than that of a thick wall workpiece. There are a number of unclarified points regarding to the machining accuracy of a thin wall workpiece, and an error avoidance method has not been established. Therefore, in this study we analyze deflection of the workpiece and tool using FEM and present basic data on how to realize precision end milling of a thin wall workpiece. The main results are as follows. (1) The loads acting at the nodal points of the FEM model workpiece and tool are calculated based on the cutting mechanism of a helical end mill. (2) The deflections of the workpiece and tool at the surface generating point on the cutting edge are calculated using the FEM under forces obtained in (1). The machining error of the side surface is obtained from the relative displacement of the workpiece and tool. (3) Calculated values of machining error almost coincide with the experimental values; thus, the appropriateness of the meshed elements, assumptions of cutting force distribution and the calculation inethod for the nodal load are verified. (4) The displacement ratios of the workpiece and tool to the calculated values of machining error are shown. Under experimental conditions, the former ratio exceeds 0.5 in the region of the thin wall workpiece with height exceeding 15 mm.

View full abstract

This study deals with the development of a method for ball screw generation. The method is based on a generation process that uses a standard-shaped grinding wheel, instead of a form-grinding process that uses a form-dressed grinding wheel, and requires no extensive form-dressing process. In the previous study, the wheel axis is set three-dimensionally at the computed optimum angle of inclination consist of the tilt and swivel angle. However, it needs a complex working process, because the shape generated by the method is asymmetric. In this study, in order to acquire a symmetrically generated shape, the wheel axis is set two-dimensionally at the computed optimum tilt angle, and the swivel angle is set as 0. The optimum tile angle that gives a minimum geometrical error is determined by a numerical simulation of the generation process with the aid of a computer. The results of grinding experiments show that the present method reduces processing steps and working time. Through the grinding simulation and generation experiments, it is shown that the ball screw generation method with two-dimensional setting of the wheel axis has great advantages in manufacturing.

View full abstract

The surfaces of six ground specimens with an arithmetic mean roughness Ra ranging from 0.05 to 1.6μm were illuminated with a halogen lamp. The intensity distribution of the scattered light measured with a CCD camera was approximated by a two-dimensional (2-D) Gaussian function and its broadness was evaluated using the standard deviations (s.d.) σ_x and σ_y of the function. The S.d. σ_x perpendicular to the grinding direction increased with increasing Ra. On the other hand, the s.d. σ_y parallel to the grinding direction remained almost unchanged regardless of Ra. The distribution of the surface angle θ was calculated from the surface roughness profile. The s.d. σ_x increased with increasing s.d. σ_θ_x of the 2-D Gaussian function approximating the distribution of θ as σ_x=2.02+1.60 lnσ_θ_x This indicates that the intensity distribution of the scattered light represents the distribution of the surface angle θ.

View full abstract

This paper deals with the identification of alignment errors in five-axis machining centers. There are thirteen alignment errors in each five-axis machining center. Three alignment errors in the three translational slideways are identified by circular tests using a Telescoping Ball Bar. In this paper, assuming the use of a Telescoping Ball Bar, identification methods for the other ten alignment errors in five-axis machining centers are proposed. The theoretical background for identification is given by a form-shaping theory. The validity of the proposed identification method of alignment errors is confirmed by experiment.

View full abstract

A number of test codes for performance and accuracy evaluation of machine tools have been standardized worldwide. There are, however, few standards for five-axis machining centers. The standards for evaluating the motion accuracy of five-axis machining centers so far available are National Aerospace Standard 979, which was standardized by Aerospace Industries, and ASME B5. 54. In our research, a method for testing the alignment accuracy of five-axis machining centers, which assumes use of a Telescoping Ball Bar, is proposed. Three parameters, which are the range of measurement values, the mean, and the standard deviation, are defined in our evaluation method. The validity of the proposed testing method is confirmed by computer simulation of the three typical types of five-axis machining centers.

View full abstract

Refrigerant pipes in outdoor air conditioning units link the compressor and heat exchanger. The pipes vibrate due to the compressor vibration, and high vibration may cause pipe breakage or other problems. This is why study of ways to reduce pipe vibrations is so important at the design stage. Vibration analysis technology (CAE) can be used to calculate the vibration amplitudes and stresses of pipes in order to design low-vibration pipes. In order to conduct vibration analysis, a 3D shape model must be defined in a computer, and then a 3D analytical model broken down into finite elements must be created. Unfortunately the process of creating a 3D analytical model is so labor-intensive that vibration analysis is not currently a part of design work. In order to automate the process of creating a 3D shape model, an algorithm fo 3D shape model auto conversion based on a 2D drawing of a pipe was developed. The pipe targeted for this study has two features. First, the 2D drawing shows only the central axis of the pipe. Second, the 3D pipe model is expressed by one continuous line without branches or other breaks. By limiting this study to such a simple pipe shape, an algorithm for 3D shape model auto conversion based on a 2D drawing of a pipe can be developed.

View full abstract

To realize actual designs in computer aided design, we introduced a design method by the attainability and simultaneous equations to determine structural dimensions in order to make all requirements accomplish their respective objective values. We also introduced an identification method of sensitivity coefficients without theoretical design formulae. We present a multiobjective satisfactory design method that includes non-technical objective functions. Design formulae of the functions are described by orthogonal polynominals, using panel evaluations by the semantic differential method and the regression analysis. Partial derivatives of the formulae give sensitivity coefficients. The iterative solving of the simultaneous equations with the coefficients yields converged or preferred solutions of the design. In a bath unit design with 6 objective functions and 9 design variables, this method produced more precise solutions than the current models in 2∼25 seconds.

View full abstract

An optimal design problem is considered for a flexible-manufacturing system (FMS), which consists of a group of numerically controlled machines, automatic guided vehicles (AGVs) and an auto-storage. The processing times of materials (jobs) at each machine form a sequence of identically, independenly distributed (i.i.d.) random variables (r.v.s.). The transferring times of jobs by each AGV from the auto-storage to a machine or from a machine to the total storage also form a sequence of i.i.d.r.v.s. The decision variable is the number of AGVs in the system when the number of machines is fixed. We propose a method for obtaining a lower bound N_L on the optimal numbers of AGVs under some important economic criteria which have been used extensively for optimal machining speed problems: This means that at least N_L AGVs should be installed in the system. Some numerical examples are presented that demonstrate the practical usefulness of the proposed method.

View full abstract

Closed-loop production systems are used in many factories, e.g., machining, adhesive manufacturing, plastic forming and assembly factories. In these systems, many pallets, jigs or molds circulate and are used repeatedly, and the number affects the system production rate. In this paper, we analyze the effect of a number of pallets and system parameters on the production rate, and present the existence of the optimum number of pallets which yields the maximum production rate. Next, we propose an approximation method for estimating the optimum number of pallets, for which the estimation error is very small and the estimated number gives almost the same production rate as that yielded by the optimum number of pallets.

View full abstract

In this paper, we propose a method of facility layout which allows scheduling flexibility corresponding to variations in production at different periods. The critical transportation is defined as the transportation between facilities which has zero idle time in the production schedule using only operation time. To determine the layout in the first period, the facilities with critical transportation are positioned so that they are as close as possible. After the second period, facilities whose critical transportation quantity is greater than that of the preceding period are rearranged, taking into account the re-layout cost, and the layout is determined such that the total flow time of rescheduling, including transportation time, is minimized. The final period layout is the one which can be utilized in all periods.

View full abstract

A valve lifter is composed of two main parts, a body and a plunger. However, machining accuracy distributions of these parts are considerably different, so the machining dimension must be controlled and adjusted. This paper deals with the behavior of part flow in valve lifer manufacturing systems where a matching method is applied. First, we analyze the effect of combination of machining accuracy distributions and the effect of buffer capacity on the production rate. Next, we analyze the effect of machining lot size and setup time required in dimension control on the production rate, and present the existence of the optimal lot size which yields the maximum production rate. Finally, we propose an approximation method for estimating the optimal lot-size.

View full abstract

In order to obtain product designs which enhance the flexibility for changes of the working environment of a product, a flexible design optimization method is proposed in which the changes of the working environment are clearly evaluated at the design stage. Here, a criterion of maximization of the flexibility for the product working environment is added to the conventional criteria such as product performance and manufacturing cost and a multiobjective optimization problem is formulated. Influence of enhancing the flexibility for the product working environment on other characteristics is integratedly evaluated by the tradeoff analysis of the Pareto optimum solution set. Then, the preference structure of the decision maker is formulated by the preference function for each evaluative characteristic. The preference functions are interactively modified based on the tradoff analysis between the evaluative characteristics. The solution most sufficiently reflecting the preference structure of the decision maker is finally determined as the design solution.

View full abstract