The presented work handles the compensation of unbalance forces for a 6/4-Switched Reluctance Machine (SRM), which generate significant radial magnetic forces in an eccentric motor shaft position. When the rotor is operating e.g. in its principal axis of inertia, these forces must be compensated by the bearings. This is of major interest for magnetically levitated long term storage flywheels where bearing efficiency is a key factor for the overall system performance. To minimize magnetic pull, the two opposing coils are controlled separately instead of a common serial connection. Three different methods to compensate unbalance forces are presented: The modification of the reference current of the opposing coils, a quasi-parallel connection and a combination of a current controller and a fluxlinkage controller. For the observation of fluxlinkage two different approaches are described. These methods are compared to each other by means of simulation. Due to the back electromotive force (BEMF) in high speed operation, precise current control is a major challenge. Hence the third method is leading to the best results, so it was also implemented on a test rig for validation of the simulation results.
This paper focuses on realization of both the one-degree-of-freedom (one-DOF) active magnetic suspension and motor operation of a three-phase permanent magnet motor using only one three-phase voltage source inverter. The suspension winding of the electromagnet for the one-DOF magnetic suspension is connected between the neutral point of the Y-connected three-phase winding and the middle point of the dc voltage sources, and the resulting zero-phase current is utilized to control the suspension force. The control method of the zero-phase current is theoretically derived to avoid interference with the motor performance. Experimental apparatus was built and tested, which consists of an iron ball magnetic suspension system and a three-phase interior permanent magnet motor. The experimental results demonstrated stable magnetic suspension and motor rotation using only one three-phase inverter.
In the last decade, ventricular assist devices (VADs) such as continuous flow rotary blood pumps have been successfully applied to advanced heart disease patients for bridge to transplantation and destination therapy. The success with VAD technology has recently increased significant clinical need for pediatric VADs. However, VADs which are applicable to infants and small children are still in a development stage because of severe design requirements such as long life expectancy, minimal blood damage and miniature device size. Magnetic levitation is an essential technology to enhance durability and blood compatibility of rotary VADs due to elimination of mechanical contact and wear. Miniaturized magnetic systems have strong potential to develop the pediatric VAD. In this study, a novel axial gap double stator self-bearing motor which can control five-degrees of freedom (5-DOF) of rotor postures have been developed for pediatric VAD. The motor has a top stator, a bottom stator and a levitated rotor driven as a synchronous permanent magnet motor. The rotor is axially suspended between the stators which have an identical structure. A double stator mechanism enhances a higher torque production and regulates radial position through active control. This paper proposes a concept of radial position active control and the magnetic suspension ability of developed self-bearing motor. The rotor is successfully levitated and rotated up to 6400 rpm without any physical contact with 5-DOF active control in the air. The radial oscillation amplitudes are actively suppressed with the proposed radial position control concept. The developed 5-DOF controlled self-bearing motor which is suitably miniature as an actuator for pediatric VAD indicates sufficient capability of the magnetic levitation and non-contact rotation.
From a designer point of view, optimizing the energy necessary for the control is an important element that could lead to downsize the control cabinet and to an increase of margins according to power amplifier capabilities. Also, reducing the energy used, generates a less significant environmental impact. The aim of this work was to develop and to assess a new approach using polar coordinates to observe and actuate the control of a flexible rotor supported by Active Magnetic Bearings (AMBs). Two fuzzy logic based controllers used to modulate the actuating forces were developed and assessed. The first is a SISO fuzzy PID controller for which the input is the displacement measured along an action line while the output is the force applied in the same direction. The second controller considers each bearing as a single MIMO system with the displacements in the two orthogonal directions as inputs, by managing two significant physical quantities, namely tangential and radial speeds. The “mechanical” performances are compared to those of an augmented PID controller. Then, the energy necessary for the control is compared. The study is first performed numerically and then validated experimentally. The results obtained show that the polar fuzzy controller presents the most suitable mechanical performances and energy costs
Highly efficient hybrid type magnetic bearing is proposed in this paper. It is intended to apply the proposed magnetic bearing to ultra low temperature fluid pump. Traditional low temperature pump is supported by a ceramic ball bearing which is damaged by the low temperature liquid. Non-contact support is highly requested. However, the standard electro-magnet type magnetic bearing consumes high electric power causing heat to the liquid. This paper proposes highly efficient hybrid type radial magnetic bearing which can operate with very low power consumption. Also the pump is canned by FRP pipe which has thick wall to separate the stator from the liquid. The designed magnetic bearing has 3 mm wide air gap. The fabricated magnetic bearing shows good dynamic properties with high force factor.
A solar-powered magnetic suspension carrier is fabricated which achieves noncontact power supply to the carrier. There are two conventional methods of supplying power to a carrier. One is through wires or contacts such as pantograph and the other is an on-board battery in the carrier. However, the contactless property is lost in the former while stations for charging the on-board battery are necessary in the latter. These problems are overcome by the proposed carrier system using solar magnetic suspension that can achieve continuous levitation. It is assumed for the system to operate indoors like inside a factory building. The generation of power by solar cells is low so that the magnetic suspension needs power saving. A zero power control system is applied to the suspension of the floator. In addition, low-power peripheral devices are also fabricated. The developed carrier system achieved a 24-hour levitation without external power supply in a bright environment.
Characteristics of flow fields produced by a dielectric barrier discharge plasma actuator in quiescent air are numerically investigated. A time-dependent localized body-force distribution is utilized to mimic the effect of the plasma actuator with modulated bursts. The computed time-averaged and instantaneous flow fields are compared with the experimental results by using high-speed schlieren photography and particle image velocimetry. The computed flow fields are in good agreement with the experimental results when the nondimensional parameter (Dc) is within the appropriate range. With an appropriate choice of Dc, the location and size of the induced flow structures, computed with respect to the maximum flow velocity parallel to the wall, are quantitatively in agreement with the experimental results. Also considered are the effects of the burst frequency (non-dimensionalized by the chord length and the free-stream velocity of assumed separated flow control experiment) on the induced flow. The results show that changes in the burst frequency cause insignificant changes in the magnitude of the time-averaged flow parallel to the wall, but they cause significant fluctuations in the amplitude and power spectral densities of that flow.
Water solution of volatile organic compounds (VOCs) in air, especially explosive materials, is required to detect VOCs by biosensors efficiently. We developed a collecting device like olfactory mechanism of animals: thin film of water including odorant binding proteins (OBPs), which play a major role in transport of hydrophobic odor molecules to olfactory receptors. Thin liquid film was generated using rimming flow on the inner surface of a rotating glass cylinder. Two kinds of OBPs derived from Bombyx mori (Pheromone Binding Protein 1: PBP1, General Odorant Binding Protein 2: GOBP2) were used as additive in pure water. Eugenol was used to simulate an explosive substance for collection. The sample gas including dilute eugenol was collected by the rimming flow device. The concentration of eugenol in solution was measured by gas chromatography. The experimental results showed that the efficiency for the collection of eugenol increased with the revolution rate of the cylinder. Increase in the efficiency was mainly attributed to the spread of the liquid in the cylinder. Furthermore, as the revolution rate was increased, the mole fraction of eugenol near the inside wall was higher than at the axis of rotation by centrifugal force. This might enhance the dissolution of eugenol in water. The collection efficiency further increased by the adding of both proteins and PBP1 exhibited higher affinity for eugenol.
An anode gas recycle (AGR) system that uses an ejector for a 1 kW solid oxide fuel cell (SOFC) was developed to increase the electrical efficiency of combined power generation. The effects of anodic recirculation on the SOFC performance were determined experimentally with a variable flow ejector at a fuel utilization of 0.80 to 0.87 and a steam feed rate of 0 to 1.5 g/min. A quadrupole mass spectrometer was used to identify the recirculation ratio, the gas composition of reformed gas at the SOFC inlet, and that of the recycle gas at the recycle plenum located downstream of the SOFC. The recirculation ratio could be changed from 0.67 to 0.71 by moving the needle inside the ejector nozzle, and thus the SOFC showed stable performance with no steam supply, due to the large recirculation ratio of steam produced via the electrochemical reaction that was supplied to the injected fuel for the steam reforming process. The results indicate a slight decline in output power with an increase in the recirculation ratio. In addition, the fuel utilization limitation for SOFC operation is estimated to be as much as that for a one-path SOFC system without AGR to avoid a local lack of hydrogen as fuel.
Recently, it has become necessary for rotors of centrifugal compressors, industrial rolls, and other devices to be less weighted, have a long span, and operate at high rotation speed to increase productivity. Such rotors are more likely to have a high order mode than other types of rotors at the same rotational speed, and may become unstable. To solve this problem, active magnetic bearings are used to support the flexible rotor. However, because a general active magnetic bearing uses bias current to control the bearing force, it consumes a large amount of electricity. To reduce power consumption, a hybrid active magnetic bearing (HAMB) that consists of electromagnets and permanent magnets is used. The HAMB fabricated in this work has negative stiffness in both radial and angular directions. Two HAMBs, one at each end of a flexible rotor arranged horizontally, are installed to levitate it. A PID controller with a notch filter is used to control levitation. With only a PID controller, levitation of the flexible rotor becomes unstable because the flexible mode stimulates vibration. To settle the flexible mode vibrations, the notch filter is designed at the frequency of the flexible mode. Applying the notch filter to the displacement control loop has little effect on vibration decrement, whereas applying it against the angle control loop produces a remarkable effect. Therefore, the notch filter is applied only to the angle control loop. A rotation test is then implemented. The experimental results show that the HAMB-based control system stably passes through the second flexible mode rotation speed. Then, by using the advantage of the HAMB to reduce power consumption effectively, the ability of the system to reduce the control current to use zero power control is demonstrated, and, the results are compared with PID control when levitating the flexible rotor. Furthermore, a levitation test is conducted with linear quadratic regulator control.
This paper presents an adaptive controller for a position tracking problem of a skid-steer mobile robot. First, a non-adaptive control law is designed by the backstepping method, where the reference signals are generated by a non-holonomically constrained kinematic model in order to make it easy to construct a Lyapunov function candidate. Then an estimator for unknown cornering stiffness is designed based on the Immersion and Invariance (I&I) approach, where the manifold is properly designed to reduce the effect of the estimation error on the tracking performance. Finally it is shown that the closed-loop system can be uniformly ultimately bounded, and numerical simulation shows the effectiveness of the proposed controller.
This paper proposes a trajectory planning method for a transfer system on a two-dimensional (2-D) surface. The 2-D transfer system must ensure safe and rapid transfer of an object. Safe transport of objects containing vibrational elements requires vibration suppression and obstacle avoidance. Meanwhile, the transport time should be short and should satisfy the constraints of the transfer system. Our approach adds a reference trajectory to the transfer control system. The reference trajectory is derived by minimizing the integral square errors of the transferred object and the goal positions, and the energy of the frequency bands which include the natural frequencies of the vibrational elements. This optimization problem constrains the acceleration, velocity, and position of the transfer system. Angularly postured obstacles impose additional ellipsoidal constraints on the position of the transferred object. The trajectory planning problem is formulated as a quadratic problem with quadratic constraints. The effectiveness of the proposed trajectory planning is verified in simulations of an omni-directional mobile vehicle carrying the liquid container.
In this paper, we present the benefits of bond-graph analysis for mechanical-electrical systems, which are energy-harvesters based on structural vibrations and electric loads. The bond-graph is an energy-based approach to describing physical-dynamic systems. It shows power flow graphically, which helps us understand the behavior of complicated systems in simple terms. Energy-harvesting involves conversion of power in mechanical form to the electrical one and the bond-graph is a good tool to analyze this flow. The bond-graph method can be used to calculate the dynamics of the combining mechanical and electrical systems simultaneously. The biggest advantage of the bond-graph technique is that it can be used with the systems that are subject to component alternations, such as inserting, removing and swapping. The bond-graph method involves solving simultaneous algebraic equations, instead of differential equations. On the other hand, in common simulation methods, such as solving differential equations, it is difficult to change the number of components because the differential equations will have to be reconstructed. Because the bond-graph has not been used for harvesting analysis, bond-graph models for harvesting need to be created in advance of numerical analysis. In this paper, we first proposed a piezoelectric model that matches the bond-graph method. We also propose a diode-bridge model and a harvesting controller model that are suitable for bond-graph analysis. We then analyze a self-powered energy harvester that has multi-bifurcated and looped flow in the mechanical-electrical coupled dynamics.
Reaction pathway analysis was carried out to investigate the activation energy barriers of Shockley partial dislocation mobility in 3C-SiC. For each partial dislocation, there are two types of dislocations according to which kind of atom, Si or C, comprises the core edge of the dislocation line. In this paper, the partial dislocation is simulated by Vashishta potential functions. Moreover, the activation energy of kink pair nucleation and kink migration are investigated by reaction pathway analysis. The dependence of the activation energy on the driving shear stress is also discussed. The results show that during kink migration, 30° partial dislocations have a lower activation energy barrier than 90° partial dislocation. And, C-core partial dislocations have a higher activation energy barrier than Si-core dislocations for both degrees of partial dislocations during kink migration and nucleation. This conclusion is consistent with the experimental result that Si-core dislocations migrate more readily than C-core dislocations. Furthermore, we found that partial dislocations with larger distance between the dangling bond atoms along the dislocation line have higher activation energy barriers. Based our calculation results, we propose new models to account for the morphological differences in the dislocation lines.
In this paper, we present an algorithm for frictionless contact problems of linear elastic bodies with multi-point constraints. Our algorithm is based on an interior point method and is developed for large scale stress analysis of electronic device models. Electronic devices consist of dozens of thin parts such as liquid crystal displays, printed circuit boards and covers and these parts are placed layer by layer. Therefore, the finite element models contains so many discretized contact constraints and multi-point constraints that make convergence of contact states difficult to achieve. In our algorithm, multi-point constraints are removed by a quadratic penalty method at first, then a primal-dual interior point method is applied. We implemented our algorithm into FrontISTR, which is open-source and large scale finite element structural analysis software, and investigated its performance from simple contact models to actual electronic device models. The numerical experiments show that our algorithm is more efficient than an active set method with an penalty method for large models, although the convergency strongly depends on the parameter settings of a primal-dual interior point method.
Yoshimoto had been presented a hypothesis concerning the fatigue strength of a bolt in bolt/nut assembly using Ishibashi's hypothesis on the relationship between the fatigue notch factor and the local stress distribution on the first thread root of a bolt mated with nut. By using Yoshimoto's hypothesis, the effect of the manufacturing process (sequence or order of thread rolling and heat treatment processes) on the fatigue strength can be explained in connection with the axial residual stress at the first thread root of a bolt as a change of mean stress level. However, this hypothesis has not yet been verified quantitatively since the axial residual stress which may exist locally around the thread root could neither be measured nor estimated with sufficient accuracy. This study aims to quantify the axial residual stress on the thread root generated by thread rolling process. For this purpose, a method for simulating thread rolling process by using 3D elastic-plastic FEM was proposed to estimate the axial residual stress distribution by using precisely determined material property. Calculated results for two types of test specimens, one is grooved specimen with larger root radius and the other is leadless bolt specimen having the same thread profile as M10×1.25 bolts, show that the axial compressive residual stress of 1000 MPa level is generated at the root of the specimens. Finally, the validity of the simulation was confirmed through the X-ray stress measurement for grooved specimen, and the fatigue tests for leadless bolt specimen.
The objective of this study was to clarify the optimized lower limb motion during throwing in water polo by the simulation. The flexion/extension angles of the hip joint around the ball release were optimized in order to maximize the ball velocity, which corresponded to the hand velocity at the ball release, under the constraint in terms of the hip joint torque and power. The following findings were obtained from the results of optimization: The ball velocity in the case of original joint torque and power limit was 20.8 m/s. The increase in velocity from that of the original motion was 7%. In the case of doubled joint torque and power limits, the increase reached 15%. The characteristic motion in the optimized cases was the kicking forward (flexing) of both lower limbs just before the ball release. The preceding extending motion that was needed to increase the flexing velocity was also found. The increase in the ball velocity in the optimized motions can be explained as the sum of two components. One is the velocity relative to the center of mass of the whole rigid body, and the other is the absolute movement of the rigid body itself. With respect to the relative velocity, the flexing rotation of the lower half of the body produces the counter-rotation of the upper half of the body because of the conservation law of angular momentum. This effect contributed 70% of the total increase in the ball velocity. With respect to the absolute velocity, the rotations around the longitudinal and lateral axes of the inertia were induced by the reaction of kicking the fluid forward. This effect contributed the remaining 30% of the total increase in the ball velocity.
This study investigates the robustness of a space reflector structure consisting of radial ribs and hoop cables by using the multiobjective optimization method. The radial ribs are deformed into a parabola shape by cable tensions applied to the hoop cables that are arranged concentrically around the central hub and to the tie cables that are connected to the deployable structure. The design problem is to achieve the ideal deformation shape for the radial rib under the prescribed cable tensions through the determination of the rib height distribution. In addition, the ability to adjust the shape by changing the cable tension is required for handling uncertainty under actual environment condition. A simplified structural model with only one radial rib is used for structural design, where the cables are replaced by equivalent tensions. This study adopts the multiobjective optimization method to verify the structural design by investigating the trade-off between the deformation error and its sensitivity with respect to the cable tensions. Robustness corresponds to lower value of sensitivity, that the RMS error is difficult to deteriorate by changing the cable tension. On the other hand, design with higher value of sensitivity is called adjustability, because such a design is easy to adjust the deformation shape by the cable tension. The primary objective of the design problem is to minimize the RMS error between the ideal and the deformation shape of the rib under the prescribed cable tension in terms of the rib dimensions. The other two objectives are to accomplish the robustness and the adjustability of the rib deformation shape by adjusting the cable tension using the tie cable and the outermost hoop cable. This multiobjective optimization problem is evaluated by the satisficing trade-off method (STOM). Through investigating Pareto solutions obtained from the two-objective and then the three-objective function problems, the effects of cable tension variations on the surface shape error and the robustness are discussed.