Applications of machine learning methods to combustion science and technology have been carried out in recent years. Combustion instabilities in a scramjet engine are significant problems to be solved, leading to devastating damage to the propulsion systems. In the previous study, although the dynamic mode decomposition (DMD) was applied to time-resolved CH^* chemiluminescence observed in a laboratory-scale scramjet combustor, there were many peaks including noise components, making it difficult to extract the features. Therefore, in the present study, the sparsity-promoting DMD (SP-DMD) with the L1 regularization term was applied to remove the noise and extract the dominant modes. The dynamic pressure and shadowgraph images were also measured with a pressure sensor and a high-speed video camera. The fast Fourier transform was used for the pressure data. The peaks around 1600 Hz were observed for the power spectral density of pressure. This instability was also explained by the SP-DMD modes and a low-rank expression of 8 modes with an observation of time-resolved images of a shadowgraph. The data-driven approach of time-resolved CH^* chemiluminescence and the shadowgraph with SP-DMD clarified the combustion instability mechanisms in detail.

Environmental problems have been getting serious at a terrific speed, and enhancement of thermal efficiency has been expected on engines. However, cyclic variation, a phenomenon whose output varies at each cycle in spite of steady operation, interferes with the enhancement of thermal efficiency. Not much has been revealed about the mechanism of cyclic variation and no specific ways of controlling it have been found. In this study, a neural network model with polynomial regression was constructed in order to try clarifying the mechanism of cyclic variation. Polynomial regression helps to visualize how much influence each input parameter has on a target parameter. In addition, the model was designed to eliminate unnecessary parameters in the regression equation all at once to enhance readability of the regression equation. Experiments were conducted on a 4-cylinder gasoline engine under stoichiometric conditions. Main data of those experiments were cylinder pressure, intake pressure and exhaust pressure of each cylinder. As input variables, these data are used in learning of the neural network model, and we get a regression equation, which has IMEP as an objective variable and input variables mentioned above as explanatory variables, as a model output. These input variables are the candidates which can be causes of cyclic variation. Analytical result showed that variables from other cylinder could affect cyclic variation of the target cylinder. Precision of the model was still quite low, but the model demonstrated its potential to show such a beneficial result. Therefore, it is necessary to improve the model precision. Considering other ways of selecting input variables may improve it.

A computational fluid dynamics simulation has been used to investigate reflow soldering process for a small electronics component. Isothermal immiscible two-phase flow of air and solder is considered with a VOF method using OpenFOAM. In this study, the effects of physical properties of liquid solder on the movement of air/solder interface are examined. A viscosity has been changed from the lowest value for molten solder to the highest value for solder paste. Lower kinematic viscosity causes the excessive deformation of the interface. The excessive deformation will result in an unrealistic movement of small electronics components. A reasonable modeling will be needed to simulate reflow process.

In this study, we measured the temperature distributions of Ag nano-wire network (diameter 120-150nm) using thermoreflectance imaging method, when a constant current was supplied. As a result, the temperature change gave a non-uniform distribution following the current path based on the variation of the resistance at the wires and connections. Weibull distribution was applied to the probability distribution of the nano-wire temperature to analysis and evaluate the uniformity and connectivity of the wire. The results showed some potential of the shape factor of the Weibull distribution to represent the uniformity of the temperature distribution.

Thermal management is more and more challenge since the development of modern electronic devices. This object of this paper is to investigate the thermal conductivity and viscosity of polymer composites with high packing ratios of alumina particles. Alumina/epoxy composites were fabricated by using planetary centrifugal mixer. Thermal conductivity of alumina/epoxy composites was measured by steady state method and the viscosity was evaluated by using tuning fork vibro viscometer. The effects of alumina volume fraction and addition of reactive diluent on the thermal conductivity and viscosity of the composites were discussed. As a result, the thermal conductivity and viscosity of alumina/epoxy composites increased since the alumina volume fraction increased. Moreover, the thermal conductivity of alumina / epoxy composites was affected by types and amounts of reaction diluent. 75vol% alumina/epoxy composites with 30wt% SR-8EGS reaction diluent in the matrix showed the highest thermal conductivity.

Polymer composites with high thermal conductivity are used in electric devices. The objective of this paper is to study the difference in thermal conductivity of polymer composites measured by steady state method and transient method. Finite element analysis of thermal conductivity measurement of steady state method and transient method was performed using two-layer models. The two-layer model consists of surface layer and thermal conductive layer. The effect of thickness and thermal conductivity of surface layer on the thermal conductivity of the model was discussed. The result showed that thermal conductivity obtained from transient method was higher than that obtained from steady state method. Moreover, decreasing thermal conductivity and increasing thickness of surface layer led to decrease in thermal conductivity of the model.

The aim of this study is to verify a detection method of defects in MEAs of a PEFC with non-contact. This study is based on a measurement of a current distribution inside the PEFC, which is estimated from magnetic field density around the PEFC by an inverse problem analysis using a combination of Tikhonov’s regularization and Sparse modeling. For the experimental data, only the direction in which the magnetic field density is high among the two-dimensional magnetic field formed by direct current from the magnetic sensor around the fuel cell was used. In addition, the used MEA was divided into 25 equal parts, and among them, the research was conducted with a fuel cell stack equipped with an MEA, in which No. 22 was defective.

For polymer electrolyte fuel cell (PEFC), control of water drainage is an important issue to supply oxygen to catalyst layer (CL) in high current density operation. The gas diffusion layers (GDLs) have a critical role of water drainage and gas distribution in PEFC. The author's previous studies have showed that hydrophilic treatment of GDL surface decreases water accumulation in GDLs. In this study, enhancement of water drainage was attempted by partial hydrophilizing of the GDL. The GDL surface that touch the rib part was hydrophilized by applying the Nafion solution and drying it. The effect of partial surface hydrophilization on the cell performance was discussed by observing condensed water behavior and measuring cell potential changes using a visualization cell. As a result of the experiment, it was shown that the partial surface hydrophilizing of GDL can improve the water drainage performance.

The power generation performance of polymer electrolyte fuel cells is known to be greatly deteriorated by the reduction in oxygen diffusivity caused by capillary condensation of water vapor in catalyst layers. In this study, we investigated the effect of lattice density functional theory (L-DFT) calculation parameters on the adsorption/desorption behavior by comparison with experimentally obtained sorption isotherms. In the sorption isotherm measurements, humidity-controlled gas was introduced into a catalyst layer, and the mass of adsorbed water was measured with a magnet suspension balance. Then, L-DFT calculations were performed using realistic structural data reconstructed from continuous cross-sectional images of the catalyst layer to investigate the effect of L-DFT wettability parameters. We found that optimization of the model parameters and consideration of the wettability of ionomer were required to better reproduce capillary phenomena in the catalyst layer using L-DFT calculations.

Nanoscale structural and diffusional properties of ionomer thin films, which are difficult to be investigated with experiments, are studied by Coarse-grained molecular dynamics (CGMD) method. The molecular models are established base on MARTINI force field, because the simulation system size could increase to the order of 100 nm and greatly improved than ordinary molecular dynamics simulations. Structural and water diffusional properties are analyzed with different thickness of ionomer thin film from 4.7 nm to 90.2 nm, and different wettability of the supporting substrate, i.e. hydrophobic or hydrophilic substrate. As the result, similar water diffusivities are found in bulk state and thicker (> 50nm) ionomer thin films since thicker film has lager ratio of central region that has similar structural properties to bulk state, but lower water diffusivities are found in thinner (< 20nm) films due to stronger effect of ionomer/substrate and ionomer/gas interfaces. Also the wettability of substrate affects greater in water diffusivities for thinner ionomer films due to different structural properties of ionomer/substrate interface.

In recent years, it has become important to take measures against greenhouse gases. Especially the effective capture and utilization technology of carbon dioxide. In this study, we produced on alkaline solution, using microbial fuel cell (MFC) to absorb carbon dioxide by aerial exposure. As a result, we confirmed that carbon dioxide can be immobilized by MFCs. And there are more carbon dioxide immobilized, when the fed gas had a higher concentration of carbon dioxide or in a more humid condition.

All-solid-state batteries are attracting attention as a useful power source to replace lithium-ion batteries, which may have a safety issue. Among them, oxide-type all-solid-state batteries using Li₇La₃Zr₂O₁₂ (LLZ) solid electrolyte are expected as a next generation energy device because of their high safety and high energy density. However, capacity degradation occurs due to crack formation between the cathode / solid electrolyte interface during cycles, leading to a serious problem. As a solution to this problem, a composite cathode consisting of the cathode and solid electrolyte materials is used to fabricate the electrode. It was reported that the battery performance was improved by using the composite cathode of LiCoO₂ (LCO) cathode and Li₃BO₃ (LBO) solid electrolyte. However, the capacity degradation has not been completely prevented, so it is necessary to further improve the positive electrode characteristics by controlling the internal structure of the composite cathode. In this study, we controlled the particle size of LCO and the mixing ratio of LCO and LBO in the composite cathode to manufacture LCO-LBO / LLZ / Li oxide-type all solid state batteries. The battery prepared was electrochemically evaluated by a constant current test. In addition, cracks in the composite cathode cross section were observed by SEM. As a result, the cycle characteristics were dramatically improved as the size of the LCO particles decreases and the area of crack was also reduced. By using LCO nanoparticle, we succeeded in creating an all-solid-state battery that has excellent cycle characteristics and suppresses the occurrence of crack formation.

Computations of coupled phenomena between the fluid–acoustic interactions in the cavity flow and heat conduction in a cascade of flat plates, which is called stack, installed in the cavity were performed to investigate the flow conditions for an effective thermoacoustic heat pump driven by acoustic radiation in cavity flows. The effects of stack position and porosity on the thermoacoustic heat pump phenomena were investigated. As a result, the sound pressure level at the fundamental frequency was reduced due to the installation of the stack, while the reduction level became weaker with the installation position closer to the cavity bottom. The oscillation mode of the shear layer was also changed depending on the stack position. The effects of the installation of the stack with higher porosity on the sound pressure spectra were found to be weaker. The higher temperature gradient along the stack was obtained with the stack closer to the cavity bottom and higher stack porosity. The predicted velocity, pressure and temperature fields showed that the thermoacoustic heat pump phenomena occur in the acoustic oscillatory flow between the stack plates.

In order to investigate the sound generation from the flow field, the acoustic perturbation equation and the no-reflection boundary condition of the sound wave which is called Sommerfeld radiation condition are implemented in OpenFOAM, a set of computational fluid dynamics codes. Separated-reattaching flows between the blades of multi-blade fan model and the generated sound were calculated. Two types of blade models are used. One has a constant blade thickness, and is named original model. The other has a variable blade thickness, and is named filled model. The structure of the flow field contributing to the sound generation was extracted using the conditional average which refers to the calculated sound signal. As for the original model, it is difficult to extract the flow field contributing to sound generation because the space between the wings are filled with vortices. As for the filled model, it is suggested that sound is emitted at the moment when the coherent pressure fluctuation caused by the vortex exists near the reattachment point of the flow on the blade surface.

Acoustic impedance of a rectangular orifice of a resonator on a side wall of a duct was measured. Three types of resonator orifices were used: right-angled edge on the duct side (NN), round on the downstream side (NR), and round on the downstream and upstream sides (RR). The measurement was based on two-microphone-method, in which the acoustic impedance was calculated from the transfer function of sound pressures at the end of the resonator and at the outlet of the orifice. The acoustic impedance, especially the resistance θ_p at resonance frequency, depends on the sound pressure and the flow speed. When the flow speed in the duct grazing over the orifice U_g is low, the resistance θ_p mainly increases with the sound pressure. On the other hand, when U_g is high, θ_p increases with . The higher the bias flow speed U_b through orifice from resonator to duct, the higher the resistance θ_p. The orifice shape hardly changes the resistance θ_p of NN and NR, but the RR has a lower θ_p.

It is important to know the sound generation system that consists of an elastic nozzle and airflow for further understanding of biological sounds that are created by animals. The sound generation with an elastic nozzle made of konjac gel and an air compressor was experimentally examined. The exit of the nozzle voluntarily vibrated and generated sound when compressed air passed through the nozzle. The experimental result showed that the periodic open-close motion of the nozzle and the periodic bursts have the same frequency and occur synchronically. The pitch of the sound equals the burst frequency. To know how the tone burst such as the sound emitted from the elastic nozzle is detected by the human auditory system, the pitch of the tone bursts with various combinations of the burst frequency and carrier frequency (the frequency of the high-frequency waves that form bursts) were tested. It was examined whether the carrier frequency is clearly, ambiguously, or never heard in each condition of the frequency combination. This investigation suggests the pitch of tone burst is the same as the carrier frequency only when the burst frequency is less than 100 Hz.

A porous windscreen is used for wind noise reduction in a field measurement, but it cannot sufficiently reduce a low-frequency wind noise. Thus, we study a porous windscreen for low-frequency noise measurement. Low-frequency noise has less directionality, and this allows us to place a microphone with a hill-shaped porous windscreen on the ground, where a wind becomes slow. The flow around the windscreen is numerically simulated. The result shows that a part of the flow penetrates the windscreen and decelerates by the energy loss, and that the other part of the flow moves around the windscreen and separates near behind the top. The effect of the windscreen is experimentally examined with an open wind tunnel. The wind noise in the low-frequency regime reduces with an increase of the windscreen height, and this indicates that the flow separation near the top of the windscreen has dominant effect on the low-frequency wind noise.

In recent years, air floating systems are used in factories. However, a self-excited vibration occurs when supply flow rate or mass are increased. The self-excited vibration cause declining safety and performance. Therefore, confirming occurrence condition of the self-excited vibration is important. This paper presents analysis of the self-excited vibration of an air floating system. Instability conditions are examined considering translational and rotational vibration. Equation of motion and flow equation are separated into steady equation and nonsteady equation by method of perturbation. The nonsteady equation are Laplace transformed and system’s characteristic equation is obtained by substituting steady terms for the nonsteady equation. We used Routh-Hurwitz stability criterion and confirmed occurrence condition of the selfexcited vibration of an air floating system.

Recently, importance of hydraulic power generation is more increasing owing to its availability of flood prevention and decarbonization. Accordingly, demand for dam discharge valves has been increasing. However, there remains problems to improve reliability of the valve systems because of complicated two-phase flows take place in the valve. Therefore, in this study, we examine a link-sleeve valve (LSV), which is usually used for public water supply, as a dam discharge valve. In general, to prevent cavitation erosion, vibration, and noise, air is injected from atmosphere into the downstream pipe of dam outlet work. Since LSV has been used in the public water supply where the air aspiration cannot be used due to consideration of sanitation, the air aspiration methodology for LSV has not been investigated. In order to clarify the characteristics of the air entrainment by water jets in the LSV, experiments were conducted. As a result, there was similar analogy with a jet pump at high water flow rates, while air suction characteristics shows complicated characteristics at low water flow rates.

Water channel tests of fluid-elastic vibration occurred at circular cylinders subject to cross flow have been conducted. The purpose of this experiment is to separately analyze the influence of the longitudinal pitch ratio (L/D) and the transverse pitch ratio (T/D) in an in-line square array on fluid-elastic vibration. The array consists of 9 cylinders in 3 rows and 3 columns, and only the central cylinder vibrates in the direction perpendicular to the flow. The pitch ratio is changed by 0.2 from 1.3 to 2.1 and, when array patterns were set, the pitch ratio in one direction was fixed at 1.3 and the other was changed. The pitch ratio of 1.3 is small enough to occur fluid-elastic vibration. Natural frequency of cylinder is 3.1 Hz, and the Reynolds number is 2.0×10⁴. As a result, we found that the larger the T/D was, the larger the amplitude became at the same flow velocity and L/D. On the other hand, the amplitude rapidly increased when the L/D became larger than 2 at the same flow velocity and T/D.

The vertical direction conveyance control of the magnetic levitation system was investigated in this study. The deviation of the levitation object from the target position, which is caused by the vertical direction conveyance motion, was suppressed effectively by applying the feedforward control. The feedforward controller was designed by the filtered-ε LMS (Least Mean Square) algorithm, which is constructed by the following three-step. The first step is the estimation of the FIR (Finite Impulse Response) model from the target levitation position to the sensor signal which detects the position of the levitation object. The second step is the inverse modeling of the FIR model which is estimated in the first step. And the last step is the derivation of the feedforward controller using the inverse model constructed in the second step. All of the three-step are conducted off-line. Both models of the first and second steps, and the feedforward controller are all derived by the method of the steepest descent iteration process. The goal of this study is to convey an assembly part or an industrial product in a three-dimensional space by using electro-magnet. The levitation control system is controlled by PID, and control of the horizontal direction conveyance was already achieved successfully by the robust PID control system. In this study, the effectiveness of the vertical direction conveyance control system is shown through the experimental results.

In order to improve the safety of transporting art works, the presenters are investigating the vibration characteristics of art works. First, we focused on the shape of the folding screen, which is one of the most frequently transported art objects, and conducted vibration experiments on a replica of the screen. The Japanese paper on the top surface of the folding screen has a coloring layer, and is the part that must be protected the most. It is thought that the vibrations generated on the washi during transportation affect the washi in the direction of damaging at least the coloring layer on the washi. Damage to the coloring layer includes, for example, cracking and peeling. Although such damage rarely occurs in a single transport, it is thought that there is a risk of gradual progression as a result of repeated transport and exposure to a large number of vibrations. Therefore, reducing the vibration of washi as much as possible during transportation will ultimately lead to the protection of the coloring layer. The presenters thought that the vibrations generated in the washi paper on the surface of the folding screen would greatly affect the vibrations of the underlying wooden framework structure. Therefore, they conducted a vibration experiment using a replica of the wooden frame base only. In this presentation, the results of the vibration experiments will be discussed. As a result of the experiment, the existence of a resonance frequency around 70 Hz was observed at almost all measurement points when the replica was vibrated with a sinusoidal wave in the range of 10-100 Hz. At the resonance frequency, the acceleration value increased toward the center of the replica.

In recent major earthquakes in urban areas, pathways and stairways that serve as evacuation routes were cut off due to falling pipes and falling elements of equipment. In this way, there is a risk that the disruption of the evacuation route may lead to the occurrence of secondary damage. Therefore, it is necessary to evaluate the characteristics of earthquake damage and prepare for future earthquake disasters comprehensively and quantitatively. In this study, damage events to mechanical structures related to the selection of evacuation routes were investigated, and the events were classified according to the degree of damage so that they could be confirmed quantitatively. In addition, for efficient evacuation behavior, pedestrian simulation will be used to study the impact on traffic depending on the extent of damage.

In forestry work, vehicles such as heavy industry machines or tractors are used when they carry tools. It is necessary to grasp the behavior of the vehicle when they run on uneven ground. Gravel path is one of non-even ground, which can be modeled as group of particles. When the vehicle run on group of particles, vibration occurs on the vehicle that causes deterioration in safety and ride comfort. It is important to clarify the characteristic and occurrence condition of vibration. In this study, the particle method is used to analyze the vibration of a vehicle. Non-smooth DEM is used in among the particle methods. One of the reasons is short calculation time compared to DEM. By changing the speed of the vehicle and the number of particle layers, vibration characteristics, the occurrence conditions, and mechanisms are considered. Calculation result show that the amplitude increased as the rotation speed increased.

When seismic wave is input into the models of high-speed-moving connected vehicles such as the bullet train, the models are too complex to analyze. So, by using sparse estimation, we could extract high-risk modes for overturning and reduce order of the models. In this study, we examined the extracted modes in terms of participation factor and mode shapes. In conclusion, the larger absolute value of participation factor was, the more modes tended to be extracted. And, by observing the mode shapes, we found that the mode which contributes to overturning were extracted. Also, the displacements of the extracted mode were mainly associated with the body of the vehicle. Furthermore, we investigated how much the low-dimensional models could reproduce the original models. As a result, the reproducibility was improved when the weight of each vehicle was made uniform.

In recent years, there has been a growing demand for terrain measurement systems using 3D laser scanners. There are two types of topographic measurement systems in use: Mobile Mapping System (MMS) and Unmanned Aerial Vehicle (UAV). MMS has lower survey error than UAV. In addition, both MMS and UAV are very expensive. One type of MMS is a 3D scanner for marine topography measurement, which is mounted on a ship. The accuracy of the marine scanner is between that of a land-based MMS and a UAV, but due to the special nature of its application, its utilization rate is low and it is not fully utilized. In this theme, we analyze the mechanical properties of the Merlin 3D scanner for marine use and develop a vibration-isolation mechanism for use on land, aiming to develop a system that can be used more easily than existing MMS for land. This system will contribute not only to the efficiency of surveying at civil engineering sites but also to the development of disaster prevention such as the reconstruction of hazard maps.

The author had previously reported the optimal design formulas for a series-type double-mass dynamic vibration absorber (DVA) to be installed in a damped primary system. At that time, only one formula giving the exact optimal design conditions for the DVA was reported for each optimization criterion. Each formula was obtained by solving a set of simultaneous algebraic equations consisting of four or five equations. The usual method for solving simultaneous equations is to reduce the number of unknowns in these equations one by one, and finally derived a single higher-order (cubic or quartic) equation consisting of a single unknown. Depending on which variables are left at the last equation, completely different solutions calculating the optimal designing values of the DVA can be obtained for each criterion. Because these are the solutions to the same optimization problem, numerical calculations will return to us the same values. However, the derived multiple solutions cannot be derived from one solution to the other through algebraic operations. In this article, some of those solutions will be presented.

A Sodium-Cooled Fast Reactor (SFR) operate at high temperatures and low pressures. Accordingly, the primary component has a thin-walled structure to reduce thermal stress. The design seismic loads in the horizontal directions and vertical directions have increased due to recent earthquakes. The seismic design of the component is important to severe design loads. The 3-dimensional seismic isolation system, which consists of laminated rubber bearings, the coned disc spring units and oil dampers, is investigated to reduce the horizontal and the vertical seismic loads acting on the components. The characteristics of elements in the 3-dimensional seismic isolation system has been grasped by static tests. The analysis model with the characteristics of the seismic isolation elements has been suggested. In this study, the seismic response analysis with the proposed analytical model was carried out to investigate the effectiveness of the 3-dimensional seismic isolation system. As a result, the acceleration in the vertical direction is reduced in the 3-dimensional seismic isolation system compared with the 2-dimensional seismic isolation system. The effect of rocking vibration on the structure is small.

Two issues when performing seismic design for a piping system using elastic-plastic analysis were discussed using a simple model. First, elastic-plastic analysis of the piping model using elbow elements that can compute the elastic-plastic behavior was conducted. By compared with the results calculated using shell elements, it was shown that the elbow elements can be used for design analysis of the elastic-plastic region basically. Next, though the popular piping design is performed using the nominal outer diameter and nominal wall thickness, the dimensional tolerances of piping and the variations in material properties are not included in the design condition. This study showed that it is necessary to be careful about using the nominal dimensions for the elastic-plastic analysis of piping due to the variation in the analysis results for the model considering the tolerance of piping.

In order to understand the actual strength and shape of pipe joints, and the difference from the nominal values determined by the code, material certification data and shape measurement data in the previous studies were collected. It was found that the yield stress and ultimate strength were approximately 20-40% larger than the values determined by the Japanese industrial standard. As for the shape data, the outer diameters and the wall thicknesses were investigated. The outer diameters were almost same with the nominal value. It was confirmed that the wall thickness of the elbow pipe tended to be slightly thicker on the intrados side and slightly thinner on the extrados side, and the variation from the nominal value is up to about 8%. As for the tee pipe, the wall thicknesses were distributed unevenly in the pipe joint, and the thickest position was the bottom of the tee pipe. The maximum wall thickness was 38% larger than the nominal value.

Instabilities are commonly manifested in our daily lives, such as in water jets breaking into isolated drops, wave patterns at the sea surface, sand dunes, cloud waves, and wrinkles on human skin. On the other hand, the actuation phenomena of materials upon external stimulus have attracted much attention for the development of excellent sensors and devices. In this study, hierarchical concentric circular wrinkling patterns were prepared by out-of-plane stretching method with UVO and plasma treatments. Moreover, the obtained wrinkled-PDMS was modified by the temperature-responsive polymer.

This paper reports a study on the effect and mechanism of a method to refine grains on the surface layer of steel plates by burnishing and subsequent heat treatment. In this method, plastic strain is applied to the surface layer by the burnishing, and the accumulated plastic strain energy works as the driving force of the static recrystallization in the subsequent heat treatment. It is shown that the plastic strain is effective for the formation of fine recrystallized grains. A physical mode of the burnishing and static recrystallization process is proposed, and effects of burnishing condition on the size of recrystallized grains are discussed. It is shown that refinement of the average grains size is limited even though the burnishing load and burnishing time are increased. In addition, a method to apply strong shear strain by burnishing is proposed to achieve further grain refinement.

Thermal switching and thermal diodes have been actively studied from the viewpoint of effective energy utilization and temperature control. Compared to heat conduction and convective heat transfer, the radiation heat transport is more easily accompanied by temperature gradients which can produce strong thermal non-equilibrium. The amount of heat transport is proportional to the fourth power of temperature. In this study, we developed a wavelength-selective thermal switching device consisting of a blackbody heater and a wavelength-selective emitter. For wavelength selective emitter, we fabricated silver-alumina cermet films by the MOD method. The fabricated cermet films showed a normal spectral absorptance of about 0.8 at short wavelengths and 0.1 at long wavelengths. Results show found that the device is adiabatic when the heater temperature is below the operating temperature, but the heat flux increases exponentially when the heater temperature is above the operating temperature. It was also found that the operating temperature can be controlled by controlling the cutoff wavelength.

Re-entrant texturing may potentially improve the hydrophobicity of a surface. The purpose of this study was to establish a direct-micro-fabrication method of reverse-tapered hole by using a femtosecond-pulsed laser to produce hydrophobic surfaces. An optical unit was fabricated including (i) an expander, (ii) a pair of step mirrors, and (iii) a condenser lens. The distance of the laser beam from the center of condenser lens can be controlled by the angle of the step mirror, thus the reverse-taper angle can be adjusted arbitrarily. A test-piece with re-entrant structures was prepared with the reverse-taper angle of ± 20°. The reentrant structures on an aluminium had an apparent contact angle of 150°.

The solid-state ion exchange method is one of the processing methods of glass materials. Silver precipitates are formed inside the glass when an additional voltage is applied to silver-doped surface by solid-state ion exchange. Since the silver precipitates have high electrical conductivity, they are being investigated for application as internal electrical conduction channels of glass. In this study, we tried to form miniature / multilayered wiring using silver precipitates. As a result, we succeeded in producing fine interconnections with a minimum line width of about 90 μm by using printed silver nano-ink as a silver ion source. In addition, we achieved double-layered wiring in glass by alternate doping of silver and sodium ions followed by additional voltage application.

Engineering of complex 3D structures is a highly challenging task for development of materials with novel optical properties, tissue engineering scaffolds, elements of micro- and nanoelectronics devices. Three-dimensional materials can be fabricated using various methods including two-photon photolithography, interference lithography, and molding. Recently, capillary force (i.e., surface tension) has been utilized to fold millimetric elastic membranes into 3D structures (termed as capillary origami). The capillary force appears to be ideal for 3D mircofabrication since it becomes dominant at microscale compared with other forces such as gravitational or magnetic force. Making use of capillary origami, we developed a facile strategy to fabricate complex 3D hydrogel structures (3D wrinkled imprinting structures) with programmable shape and size. This method is simple and spontaneous, involving placing a hydrogel droplet on a wrinkled-PDMS sheet prepared by one-push stretching and subsequently crosslinking it at any moment during hydrogel evaporation. Moreover, we succeeded 3D hydrogel structures covered with metallic nanofilm.

Recently, soft material structures, in which liquids and droplets are introduced into solids in various shapes, have been attracting attention. These structures can be used in the fields of soft electronics, soft robotics, and wearable devices, etc. In this study, we focus on the auxetic structures, which are patterned structures with negative Poisson's ratio. The material with auxetic structure is expected to have mechanical properties such as high energy absorption and fracture resistance. Herein, we fabricated an auxetic channel structure in elastomer using 3D printer, and aimed to develop a novel wearable device with liquid metal inflow. This room-temperature liquid metal has been widely used in flexible and stretchable sensors, focusing on embedding liquid metal in microchannels, liquid metal microdroplets formation, captive sensors, and liquid metal nanoparticles, etc. Moreover, we prepared conductive elastomer film with liquid metal through the direct composition.

A typical anti-wear additive, zinc dialkyldithiophosphate (ZDDP), has been utilized for many types of lubricants. ZDDP is known to form reaction film by direct contact between sliding two substrates and protects adhesion wear. However, there are few reports about in situ observation of tribofilm formation due to difficulty in investigating the formation process of the tribofilm in nano-scale. Recently, Gosvami et al. observed the growth process of the tribofilm formed by ZDDP at the contacts of single asperities using a novel in-situ atomic force microscopy (AFM) method and suggested that increasing contact pressure and temperature promoted the growth of ZDDP tribofilm. In addition, they confirmed that the ZDDP formation started at scratch marks. From these backgrounds, we considered that the growing amount of tribofilm could be controlled by changing the surface morphology. In this experiment, we conducted in-situ AFM in ZDDP containing oil at 120 ℃ for two different roughness samples. We confirmed that a rough surface (Sa = 2.2 nm) promoted the growth of tribofilm compared to a smooth surface (Sa =1.4 nm), and the tribofilm had the pad-like structure known as the typical ZDDP shape. In addition, the formation process showed that the tribofilm grew from each asperity of tribofilm-selves and rough surface formed more shape tribofilm. Based on the above, the promotion of tribofilm growth is considered due to the increase in contact pressure by decreasing the radius of single asperities.

Ice formation and accretion present serious economic and safety issues for many essential infrastructures such as aircraft, power lines, and marine vessels, as in many commercial and residential refrigerators and freezers. Therefore the surface to reduce adhesion between ice and the surface is required from the industrial field. The effect of contact angle on water droplet freezing process on silicone film with spherical curvature structure was experimentally investigated. A series of hydrophobic surfaces with different contact angles were prepared by spincoating.The experimental results showed that the contact angle has a strong influence on the water droplet freezing time. The larger the contact angle is, the longer the freezing time of water droplet. The measured frost crystal growth velocity on the hydrophobic surfaces is larger than that on the plain glass surface. The larger the contact angle is, the lower the adhesion force between the frozen water droplet and the cooling surface.

The micro-surface structure influences the contact angle of droplets. Recent improvements in microfabrication technology have led to many industrial applications based on the perspective of biomimetics. On the other hand, there are not enough theoretical considerations. Only two states, Cassie-Baxter and the Wenzel states have been presented. Experimentally, the contact angle is expected to take an intermediate state between the two states, or to mutually shift between the two models. However, most of the previous studies based on biomimetics, have focused on simple mimicry of biological surfaces that exhibit superhydrophobic properties or aiming for a contact angle closest to the ideal value, and have not yet proposed an industrially appropriate micro-surface structure. In our experiments, we fabricated a simple micro-surface structure on a specimen of polydimethylsiloxane (PDMS). The micro-surface microstructure consists of micropillars arranged in a plane-fillable pattern. The distance between the pillars can be continuously deformed by stretching the specimen. By carefully investigating the change in contact angle with the change in elongation ratio, we aim to find an experimental formula to compensate for the deviation of the experimental value from the theoretical value. Our final goal is to elucidate the mechanism of water repellency on micro-surface structures based on the comparison of theoretical models and experimental verification.

In order to clarify friction characteristics by a combination of periodic surface patterns, friction tests were carried out using micro/nano groove patterns prepared on cylindrical surfaces. For the combinations between microgroove surfaces, the friction coefficients were affected by not only the groove geometry but direction of sliding. For the combinations between two nanostripe surfaces, the maximum friction coefficients were measured for the configuration that the grooves of the nanostripe on both surfaces are perpendicular to the sliding direction, and for the configuration that the grooves of the nanostripe on both surfaces are orthogonal to each other. For the former configuration, the sticking force generated by the slope of the microgroove was dominant in the measured friction force.

Popular Machine Condition Diagnosis Techniques are to monitor changes in vibration velocity (RMS) over time or to detect fluctuations in vibration frequency and peak value by frequency analysis. Another technique is to visualize the machine vibration in three dimensions by using a Lissajous curve which is not widely used due to the difficulties in synchronizing the measurement data in three axes, reducing the influence of background noise, and adjusting the triaxial calibration at the installation site. To address the above issues, we developed a prototype of a three-axis digital vibration sensor and a pre-processing algorithm for precise Lissajous curves. Both hardware and software improvements enabled both simple measurement and precise visualization of machine vibration. Changes in the symmetry of vibration, which is difficult to be monitored with conventional time/frequency analysis, are successfully visualized. In this paper, we describe the 3-axis digital vibration sensor and the pre-processing algorithm for precise Lissajous curves, and introduce the extracting and indexing of features of the resulting Lissajous curves.

Anomaly detection for predictive maintenance of cutting tools is one of the challenging problems in shop floor and tools management. A modern machine learning approach, including deep learning, has been widely studied for the last decade. This study focuses on the multimodality of various cutting time-series data for extracting features of cutting tool status and proposes a multimodal variational autoencoder (MVAE) method. As an extension of our previous work, we newly consider a time series of vibrational acceleration of a cutting tool and a main spindle motor load in addition to a cutting temperature near the cutting edge. MVAE learns a so-called generative model, which is implicit but stochastic, capable of reproducing original time series data. Euclidean distance is employed to evaluate the normality of a given cutting status on the latent space acquired by MVAE. We demonstrate the applicability of the proposed MVAE method in anomaly detection for the endmill cutting tools by comparing it with conventional machine learning methods such as Autoencoder.

The pharmaceutical and food industries are facing various problems due to metallic foreign substances, and as a countermeasure, they are removed by a magnetic separator. Rare earth magnets used in the magnetic separator, it is necessary to operate qualification for demagnetization by aging or environment. In this study, we were static magnetic field analysis using the Maxwell equation for the relationship between the desired surface flux density and demagnetized. Calculating a surface flux density and the permeance coefficient of the magnetic array of a plurality of patterns, it issued a proper solution to magnetic separator. As a result, it is possible for operation qualification during operation, thereby ensuring the quality of the product and risk hedge.

This paper presents the design optimization result of anchor shape with design of experiments (DOE) and magnetic-fluid-system co-simulation technology for designing a quieter high-pressure pump. Since the impact noise level increases with increasing valve impingement velocity, the objective of design optimization is to reduce the valve impingement velocity. In this study, we executed the DOE to simulate the anchor/core impingement velocity with oblique model and four design parameters. In addition, we also compared the impingement velocity results between which obtained by DOE and which obtained by original shape, evaluated the four design parameters with factor effect diagram and obtained following results. 1)The anchor impingement velocity which obtained by on two of nine models is slower than the original shape and. 2) The flat surface length(l2) of the anchor mostly affected the fluid force acting on the anchor in the oblique model. The length also mostly affected the solenoid force acting on the core and the anchor-core impingement velocity.

In this study, structural optimization design process using both algorithmic design method and finite element analysis is proposed. Proposed process can automatically carry out from generating structures to the evaluation of objective functions. As algorithmic design method is executed by fewer design parameters for complicated structure modeling, the direct search method was used to generate optimized structures. Bayesian optimization process is also addressed to reduce the number of simulations. Stiffness maximization problem was solved to check the validity and utility of the proposed process.

While double-precision real numbers have been used for CAE simulations based on the finite element method, low-precision real numbers are used for deep learning for achieving much faster learning and inference. Due to overwhelming demand for deep learning, low-precision arithmetic is beginning to be introduced into general-purpose processors as accelerators. By using low-precision real arithmetic for finite element analysis, we can effectively use the arithmetic unit for deep learning and get advantages in terms of calculation speed and power consumption. In this study, in order to investigate the applicability of low-precision computation to finite element analysis, low-precision computation is firstly applied to the element integration process, and its effects are quantitatively evaluated.

In the design of pump turbine runner, both performances in turbine operation and in pump operation must be considered. As for Francis pump turbine, which is the most widely used type of pump turbine, the decrease of runner inlet diameter improves the turbine efficiency; it moves the best efficiency point closer to the operation head range. However, such runner decreases the circumferential velocity at the runner outlet in pump operation and the total head is also decreased, so there is a trade-off between the performances in turbine operation and in pump operation. Additionally, since Francis pump turbine runner consists of fixed blades and the blades have three-dimensionality, the design with high performance runner needs plentiful design experience. However, the design dependent on experiences may not exploit the freedom of runner geometry, thus more efficient design system which utilizes the wide design space of the Francis turbine runner has been needed. In this study, we developed the automatic high-speed design system for Francis pump turbine runner with multi-objective genetic algorithm and response surface method.

Generally, a peltier device is controlled by PID control method. The performance of the temperature control depends on its PID parameters, and the optimum parameters have been explored through many try and errors. Those process have required huge effort of the engineers. So, we have used Bayesian Optimization method which is based on Bayesian statistics to optimize the parameters for PID control automatically. We adopted Bayesian Optimization method which was based on Bayesian statistics to optimize the parameters for the temperature settings automatically. We had constructed the temperature control unit to realize the rapid and the precise temperature control. Those process have required only 20 trials to improve the fast response and stability of the temperature control.

Surface texturing is a design technique to generate irregularities on the surface of a product for changing touch feeling or friction properties. In this research, we propose a numerical method for designing a macro-size surface texture recognized by touching with a finger with the aim of achieving any tactile sensations. Based on the response surface method, the shape design variables are assigned to the orthogonal array after parametrizing the surface texture. Then, the surface texture models are created based on orthogonal array. The sliding and contact FE analyses are performed using the surface texture models and a finger model with the mechanoreceptors inside. The time-dependent stress data of the mechanoreceptor elements are used to evaluate the dynamic characteristics between the surface texture and the finger. Using the response surfaces of the dynamic characteristics, the optimum shape design parameters are determined for optimization problems with an objective and constraints condition for designing a macro-size surface texture. The effectiveness of the proposed method was confirmed through the numerical results.