A simple and low-cost fabrication method for mesoscale fluidic devices with groove depths ranging from several tens to several hundreds of micrometers is proposed. In this method, grooves are formed by compressive relief transfer using a soft tool fabricated by additive manufacturing. Specifically, polylactic acid (PLA) is additively deposited on a polyethylene terephthalate (PET) film to fabricate a laminated soft tool, which is then used to compress a commercially pure aluminum (A1050-O) disk to transfer the patterned asperities. Transfer experiments and finite element analyses were conducted to clarify the effects of processing conditions on groove formation under relatively low forming loads. It was demonstrated that localized plastic deformation induced beneath the PLA regions successfully formed mesoscale grooves in the aluminum disk, with groove depths exceeding 200 µm achieved even at moderate average pressures. The thickness of the PET film had little influence on the transfer ratio, whereas the height of the PLA layer significantly affected the groove depth. Groove formation also exhibited a clear saturation behavior once full contact between the PET film and the aluminum workpiece was achieved. The effect of channel spacing was further investigated. A spacing of 1 mm caused mechanical interference and subsidence between adjacent channels, whereas such interference was effectively suppressed at a spacing of 3 mm, which is proposed as a practical design guideline. Finally, a water flow test confirmed smooth flow without leakage, demonstrating the feasibility of this low-load forming technique for mesoscale fluidic devices.

Although physics-informed neural networks (PINNs) are attractive machine learning-based methods for solving partial differential equations, they have a disadvantage referred to as “spectral bias” or “frequency principle”. This is a property where low-frequency components of the solution are learned faster than high-frequency ones. To solve this, sinusoidal representation networks (SIRENs) have been presented. Although PINNs employ hyperbolic tangent functions or Sigmoid Linear Units as their activation function, SIRENs use sinusoidal functions to preserve its higher-order derivatives. However, SIRENs suffer from instability problems during neural network training. To address this, we propose a novel formulation that employs a linear combination of sine and cosine functions as their activation function. The hidden layers of SIREN, which are represented with sinusoidal functions including weights and bias terms, can be converted to the linear combination of sine and cosine functions by using the trigonometric addition formula. Since, as a result, the bias terms can be moved outside of sinusoidal function, the weight and bias term determination problem of the original SIREN can be simplified. Numerical experiments on partial differential equations involving high-frequency oscillations demonstrate that the proposed method achieves comparable or significantly better performance compared to the conventional SIREN formulation using sinusoidal function.

Miniaturization of magnetically suspended motor has been strongly demanded to develop implantable continuous flow pediatric ventricular assist devices (VADs) with excellent blood compatibility. This study develops a compact axial gap type self-bearing motor for pediatric VADs. The magnetically levitated motor has a top stator, a bottom stator, and a centrifugal closed impeller which is sandwiched between both stators with identical shape. The impeller axial position is regulated with vector control. The inclination angle of the impeller is regulated using P+/-2-pole theory. The double stator self-bearing motor can enhance torque production with small device size. This paper investigates a miniaturized axial gap type self-bearing motor with height reduction considering anatomical compatibility for long-term in vivo evaluation of the magnetically levitated pediatric VAD. The developed self-bearing motor has an outer diameter of 30 mm and a height of 29 mm. The constructed magnetically levitated centrifugal blood pump was driven with non-contact operation in a closed loop circulation circuit. The maglev pediatric VAD prototype indicated that the flow rate can be regulated in the range of 0.5-2.5 L/min against a head pressure of 60-100 mmHg. The vibration amplitude of the levitated impeller was 0.02 mm in the axial direction, and the tilt angle was about 0.2 deg. The developed pediatric VAD demonstrated sufficient hydraulic and magnetic suspension performance for use in animal testing with device implantation.

Metal nanoparticle bridging across an electrode gap using dielectrophoretic force is expected to be applied to self-healing metal interconnects. However, reliable bridging requires simultaneously satisfying two conflicting conditions: applying a high voltage to the electrode gap before bridging and suppressing breakage caused by overcurrent after bridging. This study proposes an effective method to achieve both. Specifically, by introducing an external resistor and an inductance to the voltage application circuit and focusing on the circuit's frequency dependence, we modeled the difference between the pre-bridging and post-bridging states using an equivalent circuit. We derived the optimal conditions for the added resistance and inductance to achieve both pre-bridging voltage amplification and post-bridging current suppression. Analysis using the equivalent circuit model revealed two major effects: (i) the additional impedance provides a voltage amplification effect that allows a voltage exceeding the power supply's amplitude to be applied across the electrode gap, and (ii) this combination of resistance and inductance is particularly effective for both pre-bridging voltage amplification and post-bridging current suppression when the dispersion medium exhibits capacitive behavior. Therefore, we experimentally verified the proposed method using perfluoropolyether (PFPE), a non-polar dispersion medium with capacitive properties. By adding an external resistor (10 kΩ) and varying external inductances (10, 22, and 47 mH), we measured the voltage-current relationship. Under specific frequency conditions, the pre-bridging gap voltage was amplified up to 4.6 times. Furthermore, the post-bridging current was reduced to 1/20 or less by adding the 47 mH inductor, compared to the condition without an inductor (0 mH). Finally, metal nanoparticle bridging experiments in a circuit using the 47 mH external inductor demonstrated that post-bridging bubble generation was suppressed, achieving metal nanoparticle bridging with reduced bubble formation that inhibits bridge formation.

This paper is a study that predicted the spin vibration and noise of washing machines by applying blocked force and FBS (FRF-Based Substructure) techniques. FBS is a technique that predicts the vibration/noise of a system using the frequency response function (FRF) of each component, the physical properties of the connection, and the exciting force of the source. To apply the FBS technique, the unbalanced force generated by the laundry was defined as a blocked force and the washing machine was modeled by disassembling it into a sub-system (drum, tub, cabinet). Blocked force, transfer function and connection properties obtained through the impact test were applied to the FBS equation for a sub-system to perform a 1D simulation. To verify the simulation results, vibration/noise measurement experiments were conducted at the same location as the simulation. As a result of comparing the experiment and simulation, the forecast accuracy of vibration and noise in the 0Hz～1024Hz band was confirmed to be 89% and 85%, respectively. However, in the band above 0.1kHz, the harmonic component of the predicted value compared to the measured value was not clearly shown, and this is thought to be improved by separately modeling and applying the impact load applied by the damper to the cabinet.

A stable supply of pharmaceuticals is indispensable in daily lives. Currently, various pharmaceutical products are being developed to support healthy living conditions. However, the monitoring of manufacturing processes is essential to ensure the consistent production of pharmaceuticals of stable quality. Monitoring is currently performed by independent systems based on their respective operational units. Therefore, to respond to the problem of human resource shortage, which could become severe in the future, a process monitoring method utilizing the latest technology and various types of data is essential. This study therefore presents a method for the real-time, efficient, and effective monitoring of equipment status using simple wireless sensors, focusing on continuous granulation as a representative part of the continuous formulation process. Vibration acceleration sensors were installed at three critical points of a twin-screw granulator: the powder feeding, center, and discharge sections. The collected data were analyzed using a methodological approach involving standardization, dimensional compression, and hierarchical clustering, using Ward’s method to visualize equipment status on a two-dimensional plane. Our results demonstrated the feasibility of this method which requires no high-precision sensors. As such, the application of this method is expected to control the deterioration caused by persistent unfavorable conditions, thereby reducing the risk of production interruptions due to equipment malfunction. Overall, the proposed method is considered effective for monitoring not only the equipment itself, but also the status of the process, which is an important outcome of the present study.

In quantum information processing, control of quantum entanglement is of central importance, particularly for applications in quantum computing. One established method for generating entanglement is squeezing, which exploits the uncertainty principle: fluctuations in one of a pair of conjugate variables are suppressed at the expense of enhanced fluctuations in the other. A representative approach for squeezing is the parametric amplifier. Depending on the parametric excitation amplitude, the system exhibits regimes of stability and instability of the trivial solution. Conventional studies of squeezing have focused on the stable regime, where the squeezing is maximized near the instability threshold. When the parametric excitation amplitude is increased further and the system enters the unstable regime, the system can reach a steady state due to cubic nonlinearity. However, the squeezing properties of this state have not been fully clarified. In this study, as a first step toward a quantum-mechanical treatment, we investigate squeezing in a classical parametric amplifier in the unstable regime. To enable analytical treatment, we approximate the white noise by a narrowband resonant cosine component and neglect stochasticity. The analysis shows that the steady-state amplitude depends on the phase difference between the noise and the parametric excitation, as in the stable regime, indicating a squeezing effect. Moreover, the conditions under which the steady-state amplitude is maximized and minimized with respect to this phase difference are derived. In addition, we analytically determine how the parametric excitation and noise modify the cubic-nonlinearity-induced bifurcation structure. Based on these findings, we propose a squeezing mechanism that exploits the properties of the unstable regime, unlike conventional approaches. Numerical simulations incorporating stochastic noise reveal the emergence of squeezing, although the effect is weaker than in the stable regime. These results suggest that the unstable regime may offer a possible route for utilizing previously unexplored characteristics.

Critical het flux (CHF) in an inverted U-bend tube is an important design parameter for thermal-hydraulic equipment such as heat exchangers and boilers. In the bend section, bubbles travel through the curved tube while the flow direction changes from upward to downward, resulting in a complex two-phase flow governed by the combined effects of buoyancy and centrifugal forces. However, the CHF characteristics under such conditions have not yet been fully clarified. In this study, CHF experiments were conducted using a forced convective boiling system. In addition, flow observation experiments were carried out using an air-water two-phase flow system. As a result of CHF experiments, two types of dryout were identified: liquid film dryout in annular flow at the test section exit, and incipient dryout in the bend section. For the former, liquid film dryout in annular flow within the inverted U-bend tube can be treated by that in an equivalent straight channel with the same heated length, as the Katto-Ohno's CHF correlation showed good agreement with the experimental results. For the latter, the incipient dryout was attributed to phase separation in the downward flow region of the bend section, where stagnant large bubbles were observed during the flow observation experiment. A criterion for large bubble stagnation in the inverted U-bend tube was developed based on Bendiksen's correlation, and it successfully predicted the occurrence region and condition for incipient dryout.

A magnetic suspension and balance system (MSBS) was developed to measure accurately the aerodynamic forces acting on an aircraft. The MSBS can levitate an object using magnetic force, eliminating the need for mechanical supports. This study investigates a mass measurement method using the MSBS and measures the mass of two spherical models. The calculation formula for the mass is derived by transforming the equation of motion for forced vibration with a single degree of freedom. In this experiment, the MSBS levitates a spherical model in the air. Then, the model is excited within a range of 1 Hz to 10 Hz using a sinusoidal wave generated by electromagnetic force. The mass of the models is calculated using the derived formula by substituting the values of the applied external force, the amplitude of the model, the natural angular frequency, and the excitation frequency during oscillation. The experimental results showed relatively stable mass measurement with an error of approximately 5 %. This method is expected to be applicable to measuring the added mass of objects in water because it can estimate the mass of the objects without mechanical support. These findings experimentally demonstrate that mass measurement using the MSBS is feasible, and this method achieves relatively high accuracy regardless of the excitation frequency.

For over a hundred years, the classical plate theory has been an undisputed cornerstone in many rigorous solutions for different thin-walled plate/shell models and widely applied in the design specifications and numerical analysis tools of such structures. However, one underlying theoretical barrier has not received sufficient attention: the binary linear simultaneous equations in Hooke’s law relations for traditional plate model. The influences can be categorized into three primary aspects: (a) the simultaneous equations suppress deformation in the thickness direction between free surfaces; (b) although the mid-surface is unstretched under pure bending, each layer parallel to it inevitably becomes thinner or thicker owing to stretching, and both aspects affect the value of anti-deformation stiffness; and (c) it mathematically obstructs the flexible application of the plate model beyond linear elasticity. This barrier can be unblocked by introducing the assumption of uniform thickness deformation. Thereby, the plane-stress Hooke's law can be characterized by a set of mutually independent linear relations to release constraints on the thickness deformation. Then, by utilizing Nádai’s expressions to introduce nonlinear deformation relations into the new relations, many related solutions traditionally applied in linear elasticity can be extended to the nonlinear analysis, which is impossible in the corresponding classical theories. An illustrative application is the rigorous analysis of the critical buckling of a uniformly compressed rectangular thin-plate model examined within the deformation range encompassing both linear elasticity and nonlinearity. Finally, some in-depth theoretical and technical inferences are presented.

This study investigated impact oscillations in a railway wheelset using an essential discontinuous dynamical model that retains only the velocity discontinuity caused by flange contact. The proposed system is formulated as a double-sided impact oscillator, and periodic solutions are characterized as fixed points of a Poincaré map defined on the switching surfaces. A multiple shooting method is employed to compute periodic impact motions and evaluate their stability. Numerical bifurcation analysis revealed the coexistence of symmetric and asymmetric impact oscillations within a certain range of running speeds. The asymmetric oscillation underwent a period-doubling bifurcation, leading to the emergence of a period-2 motion. Time-domain simulations confirmed both bistability and the doubled structure of the asymmetric orbit. The findings of this study demonstrate that impact-induced velocity discontinuity alone is sufficient to produce complex bifurcation phenomena in railway wheelset dynamics, even in the absence of additional nonlinearities.

This paper presents the theoretical formulation and experimental demonstration of multiple mass sensing using self-excited coupled resonators with a bandpass filter (BPF) in the feedback loop. To construct a single-input single-output system, multiple resonators with different natural frequencies individually coated with different chemical acceptors are connected to a common shuttle mass. A BPF is employed to individually address each resonator, selectively enabling mass sensing of a target analyte adsorbed to a specified resonator. We propose forcing the system under self-excited oscillation rather than periodic external excitation. This approach is particularly advantageous in viscous environments, where frequency response curves obtained from external periodic excitation often fail to exhibit distinct resonance peaks, thereby limiting the effectiveness of conventional detection methods. Furthermore, linear velocity feedback is applied to the control loop to compensate for viscous damping and induce self-excited oscillations at the natural frequency of the system. A BPF is used to selectively generate self-excited oscillations specifically at the natural frequency of the resonator as it adsorbs a target analyte. Consequently, the change in the added mass adsorbed by each resonator can be identified by monitoring shifts in the eigenfrequency of each resonator. The resonator behaviors are formulated through a discrete mathematical model. For the entire system including the BPF dynamics, a stability analysis is conducted depending on the change in the BPF center frequency and the feedback gain. By investigating the root loci for every system mode including the shuttle mass and BPF, we verify that the self-excited oscillation can be produced with the natural frequency of the specific resonator with the adsorbed target analyte. Furthermore, the method is experimentally verified using a macroscale apparatus comprising five cantilever resonators connected to a shuttle mass.

As the demand for vehicle quietness increases, transfer path analysis (TPA) for identifying vibration sources is becoming increasingly important. Recently, TPA using principal component analysis (PCA) was proposed for surface contacts; however, verification regarding moment inputs, a potentially dominant factor, was insufficient. This study compares the proposed method with the conventional matrix inversion method to verify its validity under moment input conditions. By extracting multiple principal components from response data, the proposed method can appropriately evaluate complex surface inputs, including moments, without depending on specific physical input points. Verification was conducted via Finite element method (FEM) analysis, simple structure experiments, and actual vehicle experiments. The analysis and simple structure experiments compared the accuracy of both methods under conditions with intentionally applied moments. The actual vehicle experiment evaluated the separation accuracy under complex vibration inputs, assuming an actual development environment. In the analysis and simple structure experiments, the matrix inversion method, defining the input area as a single point, failed to accurately separate the contributions of complex surface inputs involving moments, deteriorating in accuracy around natural frequencies. Conversely, the proposed method estimated contributions with high accuracy across almost the entire frequency band. In the actual vehicle experiment, the proposed method maintained good separation accuracy under complex input conditions, whereas the matrix inversion method's accuracy decreased below 300 Hz due to practical measurement constraints. These results demonstrate that the proposed method enables highly accurate transfer path analysis under complex vibration inputs involving moments and actual vehicle development conditions, indicating practical superiority.

Ductile fracture is widely understood to result from void nucleation, growth, and coalescence within a material. Although void nucleation and growth have been investigated in previous studies, only a limited number of studies have systematically quantified the strain-dependent evolution of the void volume fraction throughout plastic deformation up to failure in structural steels. To obtain material samples at various strain levels, we conducted interrupted tensile tests on notched round-bar and flat-plate specimens made of two types of carbon steel. These specimens were unloaded at predetermined strain levels approaching failure, then sectioned and polished. We measured the void volume fraction and area number density at multiple locations on the notched cross-sections using SEM observation and image analysis and determined their in-plane distributions. The total initial volume fractions of inclusions and voids in the undeformed initial materials were less than 0.04%. Discernible increases in the measured void volume fraction were observed at half or less of the failure strain. The increase in void volume fraction during plastic deformation was slight, on the order of ~1%, even just before material failure. The void volume fractions in the notched flat-plate specimens, which exhibited dominant shear failure, showed lower void fractions than round-bar specimens. Our findings suggest that substantial void growth and coalescence, which produce the fibrous fracture surface characteristic of ductile failure, occur within a narrow strain range exceeding 95% of the failure strain. Furthermore, the void volume fraction just before failure is very small and may be considered negligible in elastoplastic analyses based on J2 plasticity theory. These results provide useful quantitative reference data for deformation and damage modeling in carbon steel and may contribute to future studies on damage evolution and model validation in steel.

Sea-cage aquaculture is important for ensuring a stable supply of fishery resources. However, mass mortality events occur repeatedly in sea-cage aquaculture and cause substantial economic losses, while the underlying mechanisms remain poorly understood. Because hypoxic water masses are suspected to contribute to these events, long-term, simultaneous measurements of local currents and dissolved oxygen (DO) around cages are required to better understand the underlying mechanisms. However, such monitoring remains challenging in practice, as conventional waterflow sensors are limited by power availability and are prone to failure due to clogging by marine debris. Here, we propose a real-time monitoring system for waterflow and DO at sea cages by integrating a Pitot-type waterflow sensor optimized for underwater operation with a commercial DO sensor. The proposed Pitot-type waterflow sensor features a membrane-covered inlet and an incompressible liquid sealed inside the housing to improve resistance to clogging and hydrostatic pressure. It employs a simple circuit centered on a differential-pressure sensing element, enabling low power consumption and long-term operation under the limited power conditions typical of aquaculture cages. By integrating sensors, a battery, a solar panel, and a control unit, we developed a platform for real-time monitoring with cloud-based data transmission. In a seven-month field deployment at an aquaculture cage, the system continuously recorded waterflow and DO data. A mass fish mortality event was observed during a period of weak waterflow and reduced DO, suggesting that decreased water exchange near the cage and stagnation of low-oxygen water may be risk factors for mortality events. Overall, this study provides a practical, low-power platform for cage-scale monitoring and supports the development of on-site risk alerts to mitigate aquaculture losses.

In this study, we propose a hollow isotropic electromagnetic metamaterial that enables monitoring of the subsurface, along with a measurement method utilizing this material. The device consists of a hollow cubic frame (3D-printed resin) and an aluminum metal pattern. Its symmetrical structure allows for the acquisition of responses that are independent of orientation within the soil environment. First, simulations using the finite element method were performed to confirm a simulated resonance peak around 2.5 GHz. Furthermore, by measuring transmission responses under various orientation conditions and reflection responses within the ground while varying the distance between the antenna and metamaterial, we demonstrated that it is possible to detect isotropic responses independent of orientation and clear responses even at a depth of 200 mm underground. The results of this study contribute to the realization of soil property monitoring by enabling measurements within the ground, which was previously difficult with conventional technologies.

In this study, a three-dimensional isotropic metamaterial device with a lowered resonance frequency is proposed toward in-body wireless sensing. Wireless sensing using isolated cuboid metamaterial devices has been proposed as a passive sensing approach; however, operation in the several-GHz range is disadvantageous in high-water-content biological environments because of dielectric loss caused by water relaxation. In addition, in such isolated devices, the response from a single element is often too weak to be clearly distinguished in transmission measurements. To address these issues, an arrow-shaped resonant structure is arranged on the six faces of a cubic substrate. The introduced geometry increases the effective capacitance, thereby lowering the resonance frequency, while the modified resonator configuration improves the visibility of the transmission dip. Finite element simulations show a transmission dip at approximately 1.49 GHz, and measurements of the fabricated device in air show a clear resonance dip around 1.6 GHz, even for a single device. Stable resonance behavior is also maintained under random orientation conditions. Transmission measurements through water layers that mimic high-water-content environments also show an identifiable resonance dip, although the clarity of the dip decreases under thicker water-layer conditions. These results indicate that the proposed device is a candidate for wireless sensing in environments relevant to future in-body applications.

Synchronization can occur between the nonlinear autonomous oscillators due to their interaction, and the oscillators are entrained into the same frequency without any control. Synchronization has been applied to various vibration machines. However, in-phase synchronization occurring in rotors installed on the structure sometimes becomes a problem due to the large amplitude of the structure, and it is desirable to reduce the amplitude of the structure or to avoid the occurrence of in-phase synchronization. A dynamic vibration absorber is often used to reduce mechanical vibration, but its effect on synchronization as well as other nonlinear systems is not clear. Therefore, this paper investigates the effect of the dynamic vibration absorber on synchronization between two unbalanced rotors. This paper treats a system in which two unbalanced rotors are installed on a rigid body and the dynamic vibration absorber is added on the body and applies the energy-based analysis from the previous report using modal analysis with the assumption of stiffness proportional damping. The validity is examined by comparing the solutions by the energy-based analysis with the results by the shooting method, and two energetic conditions for the analysis are derived in the physically understandable form with modal analysis. Then, the effect of the dynamic vibration absorber on synchronization is analyzed by energy-based analysis considering the effect of two modes on synchronization with the viewpoint of energy. In addition, applicability of the analysis with the proportional damping is examined by comparison between the results by the analysis and that by the shooting method.

Small insects exhibit unique locomotion capabilities, enabling them to walk not only on horizontal surfaces but also on inclined planes and vertical walls. Understanding these locomotion mechanisms requires quantitative measurement of ground reaction forces (GRFs), which can also contribute to the development of micro-robotic systems. Although various types of micro force plates have been developed for measuring insect GRFs, their application to measurements on walking surfaces with different inclinations remains challenging. Here, we developed a micro force plate capable of measuring both vertical and propulsive forces at the micronewton scale and applied it to evaluate the total GRFs of a fruit fly. The force plate was fabricated simply on a glass substrate using laser processing, which simultaneously formed the plate body and supporting spring structures to enable biaxial displacement. The displacement at the center of the plate was measured using two laser displacement sensors tilted at ±45°, and the corresponding forces were calculated based on Hooke’s law. The developed force plate achieved a force resolution of less than 1 μN in both directions, corresponding to about one-tenth of the body weight of a fruit fly. Using the fabricated force plate, the propulsive and vertical components of GRFs were measured during horizontal walking, inclined walking, and vertical wall walking. On level ground, vertical GRFs were dominant, whereas during wall walking, propulsive forces—likely associated with adhesive interactions—exerted forces comparable to the body weight. The experimental results demonstrate the proposed force plate's ability to quantify multi-axis GRFs in small insects and provide a useful tool for investigating insect locomotion mechanics.

A new Piston-mounted Impact Damper Utilizing an Elastomer Particle Assemblage (PMID-EPA) is proposed, and an experimental study on its vibration suppression performance based on a beam setup is presented. The damper structure combines an impact damper and a damper using a particle assemblage by installing a piston inside the impact damper, and its container moves freely along a rod. The damper is fixed to the free end of a beam, which is excited by a hammer. The particles used in the experiment are elastomer particles. Results show that the damping behavior varies depending on the packing ratio of the damper. At low packing ratios from 10% to 40%, the damping waveforms of the beam are similar to those at a packing ratio of 0%, and the vibration damping does not change. However, at higher packing ratios at 50 and 60%, the damping response improves, with faster convergence of waveforms. Specifically, at a 60% packing ratio, the damping ratio increases by approximately 60% compared to the 0% case, clearly demonstrating the effectiveness of the damper. Frequency response analysis reveals that while resonance amplitude decreases with increasing packing ratio, the resonance frequency remains nearly unchanged.

This study experimentally investigates the frequency characteristics of aerodynamic noise generated by an air-cooling system model for compact electronic devices, in which a heat sink and a louver are combined. Experiments were conducted using a suction-type wind tunnel equipped with an anechoic chamber over a flow velocity range of 1–10 m/s. Each heat-sink fin had a thickness of 1 mm and a length of 19 mm, while the louvers had square cross-sections with a side length of 1 mm. The relative alignment between the heat-sink and the louvers in the streamwise direction was varied among in-phase (aligned), out-of-phase (staggered), and half-phase configurations, together with the axial spacing between them. The results show that the noise levels remained low at a spacing of s = 1 mm for all configurations, although the out-of-phase alignment produced slightly higher noise than the others. At s = 7 mm, pronounced spectral peaks appeared, and their frequencies increased with flow velocity. In contrast, at s = 14 mm, broadband noise dominated regardless of alignment configuration, and no distinct velocity-dependent peaks were observed.

A superconducting magnetic levitation system is susceptible to the sudden onset of large amplitude oscillations due to low damping and nonlinear magnetic forces, leading to transitions from stable to unstable states. Furthermore, the eigenfrequency of the levitated body can change with variations in mass or cooling position, highlighting the critical need for accurate identification of parameter conditions that occur resonance. Numerical calculation based on modeling faces challenges in achieving high prediction accuracy because analytical representations of magnetic forces deviate from actual forces. To address these limitations, recent advancements have introduced a data-driven approach, based on bifurcation theory, for predicting resonance in nonlinear systems. This method does not require system modeling. It enables accurate resonance prediction by analyzing changes in damping and angular frequency values, which are derived from response measurements under perturbation as the system approaches a resonant state. In this study, this data-driven approach was applied to superconducting magnetic levitation systems and specifically extended to auto-parametric resonances, which involve the transfer of vibration energy between degrees of freedom. We investigated how excitation amplitude influenced the prediction accuracy of this methodology through a stability analysis of the solutions of the equations of motion, utilizing the method of multiple scales. As a result, we confirmed that increasing the excitation amplitude accelerates the decrease in the frequency component of the perturbation as the excitation frequency approaches the resonance band. Nevertheless, since the tendency for the frequency component difference of the perturbation to decrease is observed even when the excitation amplitude changes, we theoretically demonstrated that this index remains effective for resonance prediction.

Characterization of ion distribution in liquids is essential for understanding microscale transport phenomena in electrolyte solutions. However, effective methods for direct measurement of electrical potential and ion distributions have not yet been established. In previous reports, we developed glass microelectrodes that clarified electrical potential distributions in electrolyte solutions and quantitatively evaluated electrolyte concentrations. The optimally designed test section allowed direct analysis of potential gradients and concentrations without the need for prior calibration. On the other hand, evaporation of liquids and dissolution of gases through the reservoirs seemed to deteriorate measurement accuracy, especially for dilute solutions. In this study, improving the test cell, the spatial resolution of the glass microelectrode was investigated for scanning intervals of 20 and 100 µm. Suppressing evaporation and concentration fluctuations, the measurement error achieved at most 0.57% for a scanning interval of 100 µm. The spatial resolution was improved to 20 µm, maintaining a mean error of 7.2%. Although the measurement accuracy has been limited by the scanning resolution, electrode tip size, and signal-to-noise ratio, the present method is expected to push the limits of direct measurement of electrical potential and ion concentration in liquids.

Understanding the vibration characteristics of the human–seat system is essential for evaluating ride comfort, and model‑based simulation provides a practical means to analyze such characteristics under controlled conditions. In this study, a three‑dimensional finite element model of the human–seat system was developed, representing major human body segments and the seat structure using concentrated masses, rigid beam elements, and spring–damper elements. The unknown mechanical parameters, including joint rotational stiffnesses and the stiffness and damping properties of the human–seat interfaces, were identified so that the transmissibility at key locations—including the buttocks–seat‑cushion interface, the back–seat‑back interface, and the head—agreed with experimental measurements below 50 Hz. The identified model successfully reproduced the dominant features of both the local transmissibility at the human–seat interfaces and the global dynamic behavior of the system, including modal characteristics observed in experiments. Using the identified transmissibility, responses during actual driving were predicted by combining the model‑based transmissibility with the measured floor‑input spectrum, and the predicted responses captured the major peaks in the measured spectra. In addition, responses estimated directly from the experimentally obtained transmissibility reproduced the measured driving responses with higher accuracy, demonstrating the practical effectiveness of transmissibility‑based response prediction. These results indicate that the developed model is capable of predicting the key vibration responses relevant to ride comfort under automotive running conditions and suggest a practical approach for estimating local ride comfort by applying the ISO 2631‑1 weighting curves to these responses.

Pulse waves are important physiological signals that are widely used for health monitoring. Currently, optical measurement methods using wearable devices such as smartwatches are commonly employed for pulse wave sensing. In recent years, pulse wave sensors based on piezoresistive strain gauges using laser-induced graphene (LIG) have attracted increasing attention. LIG is a porous graphene material formed by laser irradiation on polyimide (PI) films and can function as a strain-sensitive element. However, conventional LIG-based pulse wave sensors require strong contact pressure between the PI film and the skin, which limits measurement sensitivity and stability. Here, we propose a stretchable pulse wave sensor based on a kirigami-structured LIG strain gauge. The kirigami structure enables three-dimensional deformation and stretchability through regular cut patterns introduced into a planar film. A PI film was processed into a kirigami structure using UV laser machining, and an LIG strain gauge was subsequently fabricated on the surface by CO₂ laser irradiation. The sensor element was embedded in an elastomer in a stretched state, resulting in a stretchable sensor with a simple fabrication process. The effects of three-dimensional deformation induced by kirigami stretching on sensor sensitivity were evaluated through vertical loading and tensile tests. The results indicate that sensor sensitivity strongly depends on the kirigami design parameters and that appropriate structural optimization significantly enhances sensitivity. These findings demonstrate that the proposed kirigami-based stretchable LIG pulse wave sensor is suitable for stable and reliable pulse wave monitoring.

An application of Au-Ag alloy nanoparticles to glucose detection in the liquid phase was reported in this paper. The excellent biological compatibility of gold and silver has attracted attention to their use in glucose sensors. Alloy nanoparticles exhibit different electronic and optical properties from those of single metal nanoparticles. Although Au-Ag alloy nanoparticles can be synthesized using thermal liquid-phase reduction, precise temperature control is required for the synthesis. To overcome this issue, Au-Ag alloy nanoparticles were synthesized using a non-thermal liquid-phase reduction with a mixture of citric acid and tannic acid aqueous solution as the reducing agent in this study. The synthesized Au-Ag alloy nanoparticles exhibit a mean diameter of 9.1 nm and a coefficient of variation of 0.24, as determined by TEM image analysis. Additionally, a glucose detection test was conducted using the absorption spectra of a mixture of the synthesized Au-Ag alloy nanoparticles and a glucose aqueous solution. The glucose sensitivity was calculated based on the relation between the peak absorbance reduction and glucose concentration. The measured absorption spectra confirmed that the peak wavelength of Au-Ag alloy nanoparticles decreased with increasing glucose concentration, and the glucose sensitivity showed 0.107 /M for the Au-Ag alloy nanoparticles and 0.068 /M for the Au nanoparticles. These results found that the Au-Ag alloy nanoparticles exhibited 1.6 times higher sensitivity compared to the Au nanoparticles.

In this paper, we report a low-cost printing process for poly(3,4-dioxythiophene) (PEDOT): polystyrene sulfonate (PSS)-based all-organic microelectrodes suitable for in vitro neural stimulation and recording. Conventional microelectrode arrays (MEAs) are primarily composed of expensive metals and have been fabricated through high-cost and complex lithography processes, which limits their application to neuroscience experiments. Here, we present a method that combines a printing-based fabrication method for microelectrodes using a PEDOT:PSS composite ink, the formation of an insulating layer having a single-cell-sized sensing opening, and the improvement of cell-electrode contact by adding a PEDOT:PSS/poly(ethylene glycol) diacrylate (PEGDA) composite conductive hydrogel to the opening. Through a simple microfabrication process using cheap and easily obtainable inks, cost reduction and improved accessibility of MEAs are expected. Microscopic observation revealed that the fabricated microelectrodes possess transparency suitable for the observation of cells. The biocompatibility of the device was suggested by a Live/Dead assay of cultured cells. Also, cyclic voltammetry revealed that it has a large electric double-layer capacitance of 0.842 F/cm². This is important for preventing electrolysis accompanied by cytotoxicity during electrical neural stimulation. Furthermore, this electrode exhibited a sufficiently low electrical impedance of 0.300 Ω·cm², making it suitable for neural recording with a high signal-to-noise ratio, and it successfully acquired model electrical waveforms in physiological saline. The localized opening required hydrophilic treatment to demonstrate performance; however, by adding the gel, these excellent performances were achieved without hydrophilic treatment, and they did not significantly degrade even after a washing process. These results indicate that easily fabricable and low-cost printed all-organic microelectrodes can be utilized for neural stimulation and recording, and we believe that the application of MEAs can be expanded into various neuroscience research.

Eddy-current displacement sensors are commonly used for rotor position control in bearingless motors; however, these sensors are typically installed outside the motor, which increases the overall system size. To address this issue, a bearingless motor with an unequal-tooth-pitch stator is proposed, in which the pitch of some stator teeth is enlarged to allow eddy-current displacement sensors to be installed between the teeth. However, when eddy-current sensors are integrated in this manner, the sensor output is influenced by the motor magnetic field, resulting in apparent displacement components synchronized with the rotor rotation, referred to as sensor run-out. This paper investigates the characteristics and physical origin of this sensor run-out and proposes a compensation method for its suppression. Levitation experiments were conducted on a 10-pole, 12-slot, two-axis actively positioned bearingless motor with an unequal-tooth-pitch stator. The run-out was compensated by combining a function with pre-identified harmonic coefficients and speed-dependent notch filters. Without compensation, touchdown occurred at rotational speeds below 1000 rpm, whereas the proposed compensation enabled stable levitation at rotational speeds exceeding 5000 rpm. Furthermore, two-dimensional finite element analysis revealed that the run-out originates from an induced voltage in the sensor coil caused by magnetic flux linkage from the rotor permanent magnet, providing a physical explanation for the observed harmonic components and the effectiveness of the proposed compensation scheme.

Flexible and mechanically compliant power sources are increasingly required for wearable electronics and soft robotic systems, where stable operation under bending and other deformations must be combined with safety, comfort, and long-term durability. Although hydrogel-based electrolytes are attractive because of their softness, ionic conductivity, and low flammability, practical implementation remains limited by weak gel–gel and gel–electrode interfaces, deformation-sensitive electrical output, and insufficient environmental stability caused by drying, freezing, and swelling. In this study, we propose a thin, highly flexible battery based on an integrated five-layer acrylamide-gel cell composed of a high-salinity gel, an anion-selective gel, a low-salinity gel, a cation-selective gel, and a second high-salinity gel. These layers are monolithically bonded using gel prepolymer solutions to suppress interfacial failure and structural collapse while maintaining ionic conductivity. Flexible carbon fabric electrodes are implemented at both ends, and the entire device is encapsulated with a thin silicone elastomer to maintain mechanical compliance while providing packaging protection. In addition, glycerol is introduced into the gels to form an organohydrogel, thereby improving resistance to dehydration and enhancing environmental robustness for long-term use. The proposed design also enables lateral series integration of cells, allowing system voltage to be increased without increasing device thickness. Experimental evaluations focusing on long-term stability, discharge behavior, and flexibility are performed to examine the feasibility of the proposed device for thin, soft, and wearable power applications.

Active Magnetic Bearings (AMBs) offer significant advantages for high-speed and high-precision applications due to their contactless operation, low mechanical wear, and active control capabilities. Conventional control strategies often rely on constant bias currents to linearize the electromagnetic force-current relationship. However, this approach increases energy consumption and introduces non-differentiability in the control inputs, which hinders flatness-based motion planning and reduces overall control performance. This paper proposes a variable bias current allocation strategy formulated as a multi-objective optimization problem. A complementarity function is used to nonlinearly distribute the electromagnetic forces across paired coils, while satisfying physical constraints and reducing energy consumption, without compromising dynamic performance. Experimental validation on an axial AMB test bench demonstrates significant reductions in steady-state power consumption compared to existing approaches. Beyond local performance gains, the proposed strategy ensures a continuous and differentiable force-current relationship. This regularity is essential for integration into a global control architecture combining flatness-based control, state observation, and trajectory planning. This architecture allows further enhancement of system dynamics through accurate trajectory tracking and robust disturbance rejection. The results highlight the dual benefit of the proposed approach: improved energy efficiency and compatibility with advanced AMB control frameworks, grounded in system physics rather than empirical tuning.

Magnetic resonant coupling-based wireless power transfer (MRC-WPT) has gained considerable attention as a viable approach for high-power energy delivery, offering enhanced safety and operational flexibility—particularly within industrial automation and advanced transportation infrastructures. This technology is especially relevant for applications such as autonomous electric vehicles, electric trams, and next-generation high-speed transit systems like hyperloop. Traditional energy transmission techniques, including wired interfaces through pantograph mechanisms, tend to escalate infrastructural expenses—most notably in enclosed environments such as tunnels—and may adversely affect the dynamic performance of systems employing magnetic levitation, including magnetic levitation (MAGLEV) and superconducting propulsion platforms. Despite its advantages, WPT systems designed to supply power to high-temperature superconducting (HTS) levitation modules face fundamental challenges, primarily due to the significant inductive time constants associated with superconducting coils, which hinder efficient energy storage and rapid power transfer. In addition, the deployment of such systems incurs high capital and operational expenditures due to the reliance on complex auxiliary subsystems, including high-frequency power amplifiers and high-current rectification modules. Recently, supercapacitor technologies have emerged as promising candidates for energy buffering applications across various domains—including portable industrial equipment, detachable battery systems, and embedded sensors—owing to their high-power discharge capability and suitability for low-voltage operation. In this context, the present study investigates the potential of a fast-charging wireless power charging system (FC-PCS) that integrates super-capacitive energy storage. The research includes experimental assessments utilizing a 58 F supercapacitor module energized by a 200 W, 370 kHz of the RF power source under variable magnetic loading conditions. Based on the experimental investigations, the current charging durations by the supercapacitor discharging performances are calculated by the LC theoretical equations under different magnetic conditions.

Self-powered biosensors that operate independently of external power supplies are gaining significant attention for wearable and point-of-care applications. In this study, we report the development of a totally organic, self-powered glucose sensor based on glucose enzyme electrodes integrated with wireless energy harvesting. The glucose enzyme electrodes were fabricated exclusively from organic and polymer-based materials. The electrodes exhibited stable redox behavior and generated electrical power through the glucose oxidation, showing clear glucose concentration dependent power generation within the 0.0–1.0 mM range. The electrodes were encapsulated in a polydimethylsiloxane (PDMS)-based case and lid, forming a compact glucose sensor. While the self-powered glucose sensor exhibited a reduction in power output compared to electrode-only measurements, sufficient electrical output for self-powered operation was obtained through employing appropriate electrical configurations. To enable battery-free wireless data transmission, the self-powered organic glucose sensor was coupled with a commercial energy-harvesting wireless transmitter. By employing a series-parallel configuration of multiple sensors, we achieved wireless transmission driven solely by the electrical power generated from glucose oxidation. The wireless transmission rate directly reflected the electrical output of the glucose sensor, allowing glucose concentration-dependent wireless transmission without the use of batteries. Furthermore, the feasibility of glucose sensors for non-invasive glucose sensing was investigated using artificial urine as the electrolyte. The glucose enzyme electrodes exhibited stable electrochemical performance and sustained power generation in artificial urine. The self-powered organic glucose sensor connected with the wireless transmitter demonstrated glucose concentration-dependent responses in the range of 0.0–1.0 mM, consistent with results obtained using buffer solutions. These results indicate that the proposed self-powered organic glucose sensor can function effectively in complex biological media and achieve wireless glucose sensing without external power sources. While further optimization is required to improve power output and robustness, this work provides a foundation of sustainable, battery-free and wearable glucose sensing systems based on organic materials.

An investigation of a vibration-optimized arrangement of individual permanent magnets for radial magnetic suspension of a machine rotor is presented. The individual permanent magnets are arranged to form a ring on the rotor which is aligned with an adjacent ring formed by individual permanent magnets on the stator. In contrast to previous work, a spacing between the individual permanent magnets is explicitly considered to facilitate the incorporation of the magnet arrangement into a 3D printed machine design. The investigation is performed with finite element analysis using open source software GetDP and Gmsh. The analysis results are compared to test bench measurements for two selected bearing configurations.

This paper presents the Lorentz force generation model for bearingless Surface-mounted PM Synchronous motors. The Lorentz force varies according to the different combination of pole numbers of the motor windings and the suspension windings of bearingless motor. The mathematical model of bearingless motors with generalized pole number are analyzed. The derived models are compared with FEM simulation using the bearingless motor models with the combination of 2-pole motor and 4-pole suspension windings, 4-pole motor and 2-pole suspension windings, 4-pole motor and 6-pole suspension windings, 6-pole and 4-pole bearingless motors. The analytical model and FEM simulation agree mostly in the y-direction, however, more than 10% error appears in x-direction. From result of the additional investigation, it is considered that the cause of the analytical calculation error is the effect of the reflective magnetic field by the eddy current.

A technology that estimates heartbeats using only sensors from a magnetic levitation system of an impeller has been developed that could eliminate the need for additional ECG sensors and applying it to heartbeat synchronization control. The heartbeat-synchronized control using self-sensing is that heartbeat applied to the ventricular assist device generates a fluid force acting on the impeller. This fluid force is estimated from the impeller displacement and the current flowing through the electromagnets by using disturbance observer in the magnetic levitation system. The flow rate through the ventricular assist device (VAD) is then calculated based on the estimated fluid force, and the heartbeat is subsequently estimated from the calculated flow rate. However, one of the issues of this method is that it was difficult to estimate weak heartbeats and hence, it is required to improve the power spectrum density (PSD) ratio of estimated fluid force and flow rate through the VAD. To improve the PSD ratio, we established a control system that allows for high-sensitivity heartbeat estimation while maintaining levitation stability by lowering the stiffness of the support impeller for heartbeats (0.7-2 Hz) and increasing the stiffness for unbalanced forces (10-50 Hz) using the loop shaping method based on H_∞ control theory. We successfully applied the designed control system to a magnetically-levitated VAD and the estimated force's PSD ratio was doubled.

Solving the problem of friction-induced vibration (FIV) in catenary–pantograph systems is crucial for improving reliability and reducing costs. Previous studies have reported that FIV is a type of self-excited vibration caused by mode coupling and that standing waves on a contact wire can suppress FIV. However, the underlying mechanisms by which the standing waves can suppress FIV have not been sufficiently investigated. Therefore, this study analytically and numerically investigates the band gaps caused by periodic catenary structures using two simplified models. For simple catenaries, an exact solution is derived and the parameters that govern the band gaps are explicitly identified. Furthermore, it is theoretically demonstrated that under certain conditions, one of the simplified models can be separated into two submodels. For compound catenaries, the band gaps are numerically computed, and the influence of each parameter, especially droppers, is clarified. Additionally, the band gaps corresponding to a heavy compound catenary are presented, and their relationship with FIV is examined. The numerical trends obtained from the simplified model are further compared with those from an established catenary–pantograph model. This analysis suggests that the standing waves capable of suppressing FIV, as reported in previous studies, can be attributed to band gaps. Based on these analytical and numerical results, implementation-oriented design guidelines for tuning wire tensions and dropper stiffness are discussed. These findings are expected to not only aid in preventing self-excited vibrations in catenary–pantograph systems, but also provide practical guidance and insights into the general suppression of self-excited vibrations using band gap mechanisms.

In this article, we examine the potential of employing scientific machine learning techniques to develop an accurate dynamic real-time model of an axial electromagnetic actuator. We focus on one side of the axial Active Magnetic Bearing (AMB), specifically an electromagnet with a single coil, and aim to model the magnetic flux as a function of the input voltage to the coil. Real-time capability is defined by the model's ability to operate at a frequency of at least 14kHz on a standard computer CPU. To achieve this goal, we compare two methods for modeling dynamical systems: Neural Ordinary Differential Equations (NODE), a purely data-driven approach, and Universal Differential Equations (UDE), which integrates physics-based structures into the model. In this study, the physics-based structure employed is a Cauer Ladder Network (CLN), representing the system as a linear electrical equivalent circuit. A finite element model is constructed, and the parameters of the CLN are directly extracted from the finite element matrices. The UDE discussed in this article seeks to adjust the resistances and inductances of the CLN using a neural network based on the states of the ordinary differential equation. Training and testing datasets are generated from the finite element model. Both NODE and UDE are trained using stochastic gradient descent with a quadratic loss against the training datasets, and their accuracy is evaluated using testing datasets. Our findings demonstrate that the UDE provides an accurate and stable dynamical model, delivering superior accuracy and convergence speed compared to NODE.

Analytical stiffness models are most effective in the preliminary design of permanent magnetic bearings (PMBs). While semi-analytical models based on Coulombian and Amperian formulations provide high computational efficiency, they cannot directly account for the finite relative permeability of magnetic materials, potentially leading to an overestimation of bearing stiffness. This paper presents an improved semi-analytical approach that considers the non-unity relative permeability of rare-earth magnets. First, a correction model iteratively estimates the effective magnetization of the magnets, which is then used as input to idealized semi-analytical stiffness models. Two correction methodologies are proposed and compared: (i) the load line correction method (LLCM), based on magnet demagnetization curves, and (ii) the surface charge correction method (SCCM), based on a surface charge density formulation. These models build on established frameworks and are tailored for PMBs with a focus on reducing the computational cost. The improved semi-analytical model is verified against finite element simulations to assess its accuracy for axially magnetized radial PMBs. Results show that the LLCM and SCCM improve radial stiffness and axial force estimation. For single-layer radial PMBs, the SCCM achieves lower modeling errors, while the LLCM offers superior computational efficiency. For multi-layer radial PMBs, the LLCM yields lower modeling errors while preserving its computational advantages for the investigated cases. Overall, the improved model offers a computationally efficient and robust design tool, well-suited for preliminary bearing design, particularly in high-precision applications.