Although bladed disks are nominally designed to be cyclically symmetric (tuned system), the vibration characteristics of all blades on a disk are slightly different due to the manufacturing tolerance, deviations in the material properties, and wear during operation. These small variations break the cyclic symmetry and split the eigenvalue pairs. Bladed disks with small variations are referred to as a mistuned system. Many researchers suggest that while mistuning has an undesirable effect on the forced response, it has a beneficial (stabilizing) effect on blade flutter (the self-excited vibration). However, almost all studies have focused on the deviation of the blade frequency, and few studies have discussed the damping mistuning. In a blisk (integrally manufactured bladed disk), it seems that the damping mistuning can be neglected because the material damping is dominant. However, in a bladed disk with the blade root inserted into the disk groove, the damping mistuning caused by the partial contact between the root and groove in rotation cannot be neglected in predicting the resonant response of the bladed disk. Therefore, in this study, incorporating the damping mistuning into the FMM (Fundamental Mistuning Model), the frequency response analysis of the bladed disk is systematically carried out. From the calculated results, the effect of the damping mistuning on the vibration response of the bladed disk is clarified.

The reversal windshield wiper system is a system that realizes the back-and-forth motion of the arms by reversing rotation direction of the motor. In this system, the wiper motor control has to control the reversal angle to meet the regulations that defines the wiping range of the wiper arms and to prevent collision between the wiper arms and the windshield frame. However, the control is often disturbed by the condition of rainfall, deterioration of the wiper blade, fluctuations of the battery voltage, etc. In order to compensate the influence of the disturbances and control the reversal angle of the arms precisely, this paper proposes the motor control and its design method to drive the wiper arms. The control mainly comprises the blocks of motor speed control and target motor speed set. The motor speed control is a state feedback controller and designed by an optimal regulator. The target motor speed set includes the manipulation direction set block that decides the appropriate reversal timing of the target motor speed. The reversal timing is decided based on the predicted motor angle calculated from the wiper system model and it is the most important feature of the control. The effectiveness of the control is demonstrated by experimental results with a wiper test rig.

Some examples of subsynchronous vibration occurred in the starved bearing condition are reported in the API 684 Standard Paragraph. It is introduced that the subsynchronous vibration was experienced in the high-speed balancing test facility in the vacuum condition and the cause is confirmed to be the starvation of bearing. Because in the vacuum condition of the test facility the pressure inside the bearing housing was low and estimated to be starved. In this example the installation of the proper end seals in the bearing housing can increase the inside pressure of the bearing housing and avoid the subsynchronous vibration. The authors experienced the similar operation in the same situation in the high-speed balancing facility. The rotor type is a back-pressure steam turbine for the turbo generator unit. The bearing types is 5 pad tilting pad bearing. The effects of starved oil film on the subsynchronous vibration was doubted. But generally it is difficult to perform the fluid film analysis in the starved bearing because it requires many complex steps in the analysis. Therefore, the analysis in this paper is performed with applying smaller size bearing pads to simulate the starved bearing film conditions with carrying same bearing load. The size of bearing pads is changed depending on the estimated starved oil film. The effects of the estimated starved film condition on the stability in the rotor and bearing system can be evaluated by applying dynamic analysis to the rotor-bearing system. From the results the excessive small damping ratios in the system could be obtained and the subsynchronous rotor instability was confirmed.

The purpose of this study is to identify natural frequencies and modes of a base-isolated building by modal analysis using finite element method. In general, multi-degree-of-freedom models are often used to analyze vibration properties of multi-story buildings; however, this modeling is hard to represent spatial behaviors of the buildings such as torsional vibration, and must be considered carefully in case to apply it to base-isolated buildings. In this study, we modelized a base-isolated RC building located in Meiji University as a frame structure combined with floor slabs and aseismic rubbers on the foundation, and computed the natural frequencies and modes of horizontal and torsional vibrations. The primary and secondary natural frequencies of the numerical models agreed approximately with those given by measurements of ambient vibration and seismic waves. The natural modes demonstrated that the aseismic rubbers restrained large deformation of the buildings and worked effectively for seismic isolation. In addition, we observed principal stress distribution of each natural mode. The maximum principal stress of the base-isolated building was relatively small and uniform on each floor, because the columns and floor slabs slid horizontally and did not deform largely. These findings could provide an effective feedback to architects and designers in design phase to consider possible damages on buildings.

Monitoring structural integrity has been demanded to achieve a sustainable society against disasters, including seismic and extreme wind events, and thus Structural Health Monitoring (SHM) system is one of the significant technologies. The natural frequency and the yield displacement of structures will be the significant damage indices to assess the structural integrity. However, they would contain variances to the design values. The structural integrity assessments using the SHM should allow for the uncertainties in structural specifications, such as the yield displacement, to improve the assessment accuracy. Additionally, the earthquake ground motion will have uncertainties regarding the acceleration levels, spectral characteristics, and others. These uncertainties can lead to assessment difficulty, reducing the accuracy of the assessment. The authors propose the method using the deep Convolutional Neural Network to overcome this problem. This method using CNN has two features. The first is to suggest using the response spectra as the training data of CNN. The second is to create a simple CNN architecture with a high classifier accuracy using Bayesian optimization. This paper demonstrates that the image features detected by CNN enable us to accurately assess the structural integrity even when uncertainties in structural specifications and seismic motions are superimposed.

Several standards or manuals are available for the test methods of damping capacity of damping materials such as composites or laminated damping steel sheets; however, few documents focus on lightly damped materials like metallic alloys. This study intends to provide a simple method for loss factor identification for metallic materials using typical vibration measurement setups. Also, influences on the identified loss factor due to the shape changes of the specimen and the difference of fixing conditions such as tightening torque and inserting shims are investigated. Vibration measurement using slender beam-shaped specimens and data processing procedures are studied to obtain loss factors with high reliability. It has been shown that applying a 6 dB bandwidth method together with the spline interpolation technique on the measured mechanical impedance leads to good results rather than the commonly used half-power bandwidth method. As for the specimen's shape change, numerical studies using finite element (FE) analysis are carried out, and it has been demonstrated that the warp up to 15 mm versus beam length 500 mm does not affect the loss factor largely. According to the test results, tightening torque variations and inserting shims influence little on the loss factor, too. As an application example of the proposed method, the change of loss factor underwent different heat treatments have been investigated for two-types laminated metallic materials.

Recently, one of the authors proposed a multi-modal and multi-axis vibration reduction device against elastic vibrations of structures. The device is called eMDVA (embedded Mass Dynamic Vibration Absorber). The eMDVA consists of a ball-like mass embedded in a spherical viscoelastic material. Therefore the embedded mass can vibrate in every direction in the viscoelastic medium. The vibration characteristics of the eMDVA are investigated experimentally and numerically in this paper. Three eMDVAs with different sizes of spherical viscoelastic material containing a ball-like mass with a triaxial acceleration pickup are made, and vibration measurement tests are conducted. The measurement results show that the frequency response function (FRF) of the acceleration of the embedded mass versus excitation force has a single dominant peak. The peak frequency can be changed according to the size of the viscoelastic sphere. A finite element (FE) model of the eMDVA is developed, and investigations to determine the viscoelastic properties as Prony series expression are carried out. It has been shown that the ordinary procedure to adjust the Prony coefficients to fit the complex Young's modulus obtained by a dynamic mechanical analysis (DMA) leads to less accuracy in calculating FRF. On the other hand, good agreement between numerical and measurement results can be achieved by directly adjusting the coefficients to fit the FRF curves. In addition, FRF curves of the eMDVAs having non-spherical shaped viscoelastic material are calculated using the FE model with the viscoelastic properties mentioned above. The FE analyses have shown that by changing the shape of the viscoelastic material from sphere to spheroid or ellipsoid, the FRF has a different peak frequency according to the vibration direction.

In recent years, there is concern about decreased stability of the eye seal used for impellers due to the increase in pressure of the turbo compressors. On the other hand, the eye seal is required to have low leakage, and convergent stepped labyrinth seal are used in general because of good assembling in impellers and high sealing performance. However, there are few experimental studies on the destabilizing force of the convergence stepped labyrinth seal in the past studies, and its rotor dynamic characteristics have not been fully clarified. In this study, the dynamic characteristics of the convergent stepped labyrinth seal were evaluated by experiments and CFD analysis. The rotor dynamic coefficients that influence instability were in good agreement between experiments and analysis. It is known that the dynamic characteristics of the seal are affected by the pre-swirl, so we evaluated the dynamic characteristics due to the difference in the inlet swirling flow velocity and compared the characteristics with the straight labyrinth seal. The response of effective damping to pre-swirl was shown to be smaller with the step seal compared with the straight seal. It was suggested that the step seal is less unstable than the straight seal, even if the pre-swirl is large.

Advances in image processing technology have made it possible to measure the surface shape of the road ahead while driving. We propose a new semi-active suspension control method considering the forward road surface shape. A vehicle model equipped with a semi-active suspension can be expressed as an MLD (Mixed Logical Dynamical) model. When the shape of the road ahead can be measured accurately, it can be considered that the information on the future disturbances is available before the vehicle undergoes. In this paper, we formulate the finite time optimization problem of an MLD model to consider future disturbances as an MIQP (Mixed Integer Quadratic Programming) problem in the same way as the conventional optimal control problem without future disturbance. The solution to an MIQP problem can be obtained by a commonly available solver software. The performance evaluation of the proposed method was carried out by a simulation, and the simulation was performed on the road surface shape generated based on ISO 8608. In the simulation study, the proposed method achieved better ride comfort, i.e., equivalent suspension stroke, compared to the traditional MLD predictive control and the Skyhook approximation methods.

Mechanical snubbers and inertia mass dampers are known as vibration control devices that have inertia mass elements and generate the inertial resistance force. By adding the devices with the inertia mass elements to piping systems, we aim to reduce various responses of piping systems subject to seismic disturbances. In this study, we propose a design method to optimize design parameters of inertia mass dampers installed on a piping system that is modelled by a 3-dimensional beam finite element model with some tangent and curved pipe elements to achieve good dynamic responses while maintaining the structural integrity of the piping system and the inertia mass dampers. A performance index considering various specifications related to mitigation of responses of the piping system for earthquake disturbances and the economic cost for installation of the devices with some inequality constraints is defined. Design parameters to optimize the performance index are the number and the model size (capacity) of the inertia mass dampers, the value of its inertia mass, and its placement (node and attitude) on the piping system of the inertia mass dampers. The performance index is optimized with a genetic algorithm. With a simulation study of a piping system represented by the 3-dimensional beam model, we show the effectiveness of the proposed design methodology.

Damping characteristics of the mechanism consisted of a double-cone, travelling on straight and curved rails, have great influence on the hunting motion of railway vehicles, and on the efficiency of wave-powered electrical generators. In this work, damping associated to reciprocating movement of a double-cone travelling on straight or curved divergent-convergent rails, is evaluated. Concretely, rails are attached to the hull of a reduced-scale ship, which, by using a pendulum, is excited to roll along its longitudinal axis. By employing three laser displacement sensors, the rolling angle of the ship and the position of the double-cone on the rails, are simultaneously measured. From the recorded response signals, the damping ratios of the pendulum and double-cone are determined by using the logarithmic decrement method, both in the free and coupled oscillatory movements. Replacing the straight-rails with curved-rails, the total travelling time of the double-cone was largely increased, by reducing the energy loss at the middle point, where the rails change from the divergent to the convergent zone. In order to explain these results, the opening angles of the straight and curved rails, as well as the angles and radii of contact of the double-cone with the rails, are explicitly given. Then, by deriving relations between the damping ratio and the friction coefficient, and also between the damping ratio and the restitution coefficient related to collisions of the double-cone at the middle point of the straight-rails, phenomenological explanations are presented. Compared to the monotonical reduction of the vibration amplitude, known for oscillations under constant damping, a heaving variation of the amplitude was observed in the case of this two-degrees of freedom vibration system, this suggesting a periodically variable damping ratio.

In this paper, a convenient evaluation of an intensity of singular stress field (ISSF) at an interface corner in a single lap joint (SLJ) with a fillet is examined. The analysis method is focused on the ratio of the stress values at the interface corner obtained by applying the finite element method (FEM) to the SLJ model and the reference model which are subdivided by the same mesh pattern. It is shown that the square-shaped inclusions in the infinite plate under bi-axial uniform stress and pure shear stress can be analyzed by the reciprocal work contour integral method and are suitable for the reference model. The numerical simulations are performed on the SLJs with the fillet which are composed of the steel adherend and the epoxy adhesive. It is confirmed that the ISSF for the SLJ with the fillet can be evaluated conveniently and accurately by comparing the asymptotic solution with the FEM stress distributions. Then, the critical ISSF values are calculated from the experimental results of the SLJs with the fillet having overlap length L = 12.5, 20, 25, 40, 60mm in the earlier study. When L ≥ 40mm, the fracture occurred from interface corner, in which case it is shown the critical ISSF values are constant independent of L.

The effect of the layer thickness of prepreg on the tensile properties and failure behavior of quasi-isotropic (QI) laminates of thermoplastic CFRP consisting of carbon fiber and polyamide 6 was investigated. The layer thickness of each ply was set to 40, 80 and 120 μm by using thin-ply prepreg with the thickness of approximately 40 μm made by tow-spreading technology. As a result, that so-called "Thin-ply Effect" reported for thermosetting CFRP was not remarkable in this study of thermoplastic CFRP composed of CF/PA6. The tensile stress linearly increased with tensile strain up to the maximum tensile stress independent of the layer thickness. The tensile strength and modulus hardly changed with the layer thickness, although the delamination was interestingly inhibited with decrease of layer thickness. However, although the effect of the layer thickness of the prepreg was slight until the maximum tensile stress was reached, the failure behavior at the final breaking was significantly different depending on the layer thickness. Thinner layers inhibited delamination, and as the layer thickness increased, delamination became the dominant failure mode. Therefore, it was clarified that the effect of the layer thickness of prepreg appeared in the fracture behavior at the final failure after reaching the maximum tensile stress in the QI laminates of CF/PA6.

In this study, we develop a new theoretical model for the quantitative prediction of atomic diffusion in a crystalline material. Concentration of the solute atom is expressed by a probability density function whose time evolution is governed by the Fokker–Planck equation. The stochastic differential equation describes the two competitive processes; drift motion by the stress field and thermal diffusion by the Brownian motion. Non-singular stress field by a lattice defect is obtained from nonlinear elastoplasticity on Riemann–Cartan manifold. Numerical integration is conducted using the finite difference method which satisfies the numerical stability criteria. Present analysis demonstrates the formation of Cottrell atmosphere around a straight edge dislocation. The probability density reaches to an equilibrium distribution and the time evolution satisfies the Cottrell’s theoretical prediction quantitatively. The mean diffusion path is predicted from the streamline which includes the surface effects. We also demonstrate the dislocation pipe diffusion, i.e., accelerated diffusion along a dislocation line.

This study aims to establish a loading test method on test tracks composed of a single rail fastening at rail joints. The authors construct FEM models for railway tracks which represent the rail joint parts accurately and perform the FEM analysis. The FEM results are in good agreements with the results of a loading test on a test track composed of plural rail fastenings. Based on the FEM model, the authors proposed the loading test method on test tracks composed of a single rail fastening and compared the test results of a single rail fastening and plural rail fastenings. As a results, it becomes clear that rail head displacements and rail tilting angles obtained from a single rail fastening are in good agreement with those obtained from the plural rail fastenings. Therefore the proposed method of loading test on a test track composed of a single rail fastening based on the FEM model is available to evaluate the performance of the rail fastenings systems of rail joints.

A practical structural health monitoring has been proposed for evaluating the structural health of a whole mechanical asset by using digital twin with data collected during the operation of the asset. Digital twin can be utilized to predict the remaining useful life by estimating the variation of the physical quantity that dominates the life, even though any records of failure do no exist. However, a mechanical asset includes huge number of local hot spots where structural health should be evaluated, and accordingly, huge man-hours are required to integrate a monitoring system that evaluates the health at all the hot spots by using digital twin. A hierarchical structural health monitoring has been therefore developed to overcome this drawback. In the first stage of the health monitoring, the overview of the mechanical damage of the components included in a asset is evaluated according to a metric, D factor, that defines the cumulative damage of the components, and the assets having relatively large damage are extracted. The extracted assets are then evaluated in detail in the second stage; that is, structural health is checked at the local hot spots. The monitoring system that employs digital twin and the hierarchical health monitoring was applied to the health management of wind turbines. As the result of evaluating the structural health of the main components of wind turbines, about a hundred wind turbines can be prioritized according to the D factor. In this first stage, a surrogate model based on a machine learning was utilized for evaluating the overview of the damage with low computational cost; the approximation error of the D factor was less than 3 % by using the surrogate model. It can be therefore concluded that this practical structural health monitoring is useful for the decision making of fleet health management of mechanical assets.

The high-pressure hydrogen tank for fuel cell vehicles is made by the prepreg winding method using carbon fiber reinforced plastics (CFRP). Unexpected destruction may occur in the high-pressure hydrogen tank due to internal pressure if an impact load is applied from the outside of it. Therefore, in this study, failure aspects occurred in the cross-ply and quasi-isotropic CFRP laminates by an impact penetration test and residual tensile strength were evaluated. In the impact penetration test, a steel ball was collided with the CFRP laminate using a gas gun. Then, fiber breakage and splitting failure occurred on the impact and back surfaces of the laminate were observed and interlaminar delamination was investigated by a cross-sectional observation. After the impact test, the CFRP laminate was cut into a strip-shaped specimen, and the residual tensile strength was measured. The residual tensile strength tends to decrease as the impact energy increases, and the strength of the penetrated specimens are about 27% (cross-ply) and 14% (quasi-isotropic) lower than that of the non-impact specimens. Finally, the damage propagation behavior of the penetrated specimens including fiber breakage, splitting failure and interlaminar delamination was analyzed by a finite element method. Damage initiation and propagation behavior of the specimens could be simulated, but the maximum loads obtained from the numerical results were lower than the experimental ones. From the results, the prediction method for residual tensile strength of penetrated CFRP laminates was discussed.

This paper describes an experimental study on inlet distortion in the flow ingested to an intake of a ducted fan. Recent high bypass ratio engines tend to shorten the nacelle length for better fuel consumption. In return, the shorter the nacelle intake, the more intensive the inlet distortion inside the intake, which leads to deteriorating fan aerodynamic performance and intensifying the fan noise. This study aims at understanding the influence of intake geometries on the distortion ingested to a ducted fan and thus the fan noise caused by the interaction of the distortion with the fan blades. A subscale fan (fan simulator) coupled with intake models were used in the experiment. The intake models with different axial lengths were prepared to compare the distortion in front of the fan. One is a nominal intake as a reference. The other intake has a short non-dimensional distance from the lip to the fan surface. Crosswind conditions were imposed to realize a clearer influence of the distortion on the acoustic and aerodynamic parameters. According to the experimental results, the intake models provided a distinct increase in sound pressure levels inside the intake. It is anticipated that the sound amplification is attributed to the distortion associated with the intake lip geometry, intake length, and crosswind speed. Non-dimensional total pressure distribution revealed the high correlation between the fan noise and the inlet distortion.

For a small axial fan in a duct, the effects of acoustic resonance occurring in the duct on the flow around the fan are focused on. To clarify the condition for the intense acoustic resonance, the effects of the rotational speed and duct width on the flow and acoustic fields around the fan were investigated by compressible flow simulations with a volume penalization method. The computational methods were validated by comparing the predicted static pressure rise and aerodynamic sound by the fan with those measured. The sound pressure level became most intense for a specific rotational speed, where the acoustic resonance occurs at the blade passing frequency in the upstream duct. Although the overall tendency of the increase in the static pressure coefficient was found for a higher rotational speed with the thinner momentum thickness around the blade surface, the static pressure coefficient drops at the rotational speed for the acoustic resonance. Moreover, the acoustic resonance became more intense for a narrower duct width. In this condition, flow disturbances occur near the wall of the bell-mouthed inlet of upstream duct due to the acoustic resonance, which prompts the inflow turbulence to the fan. To clarify the effects of the acoustic resonance on the flow around the fan, the computation with the artificially suppressed acoustic resonance was also performed and the predicted flow fields were compared with those with intense acoustic resonance for the same rotational speed and duct width. The incoming flow to the fan is confirmed to become more turbulent when the acoustic resonance occurs, which promotes the spreading of the tip vortices between rotor blades. As a result, the blade loading decreases and the turbulence in the blade wake leading to mixing loss becomes intense, causing the static pressure rise by the fan to decrease.

Gas atomization, in which high-pressure gas jet impinges on molten metal stream from the melt tube to atomize it, is a popular method of producing fine metal powders. In this study, we focused on a swirling flow of the injection gas as a factor affecting the produced powder. Computational analyses and experiments were carried out to investigate the effect of the swirling gas flow on the atomization process. First, a single-phase flow analysis was performed considering only the injection gas. The simulated aspiration pressure agreed well with the experimental results. The behavior of the aspiration pressure against the gas injection pressure depended on the presence of the swirling flow. The simulation results clarified that the aspiration pressure was decreased by backflow into the melt tube in the case of the non-swirl nozzle. The backflow was caused by a vortex just below the melt tube tip, which was generated by the injected gas. On the other hand, in case of the swirl nozzle, the swirling flow decreased the pressure at the center of gas flow and inside the melt tube, which caused a high and stable aspiration pressure. Next, the effect of the swirling flow on the melt behavior and particle atomization was investigated numerically. A low-velocity region was generated at the center of the gas flow as a result of the mass loading of the melt, which was larger in the swirl nozzle. Consequently, the particles in the swirl nozzle were coarser than those in the non-swirl nozzle because some of the particles flying in the low-velocity region could not be atomized enough. Finally, metal powders were produced using a gas atomization plant, and the powder sizes and the surface images were compared. The powder size distribution showed that the non-swirl nozzle produced finer powders than the swirl nozzle did, as in the simulation results.

Large Eddy Break Up (LEBU) device has been tested experimentally for the wall shear stress manipulation in a fully developed two-dimensional turbulent channel flow. Measurement of the skin friction coefficient states that LEBU reduces the local wall shear stress. With the wall shear stress reduction, mean velocity gradient and Reynolds shear stress are reduced near the wall. From the wall static pressure drop, no net drag reduction occurs in the channel flow with LEBU manipulation. Influence of LEBU device on the streamwise pressure gradient near the wall is suggested as one of the reasons why velocity gradient and Reynolds shear stress near the wall are reduced in the manipulated flow with LEBU.

In a direct-injection gasoline engine, strict control of pollutant emissions at cold start is a critical point to comply with new regulations, and the fuel film formed by spray impingement on the piston wall is closely related. In this context, the purpose of this study is to clarify the fuel film formation process of spray impinging on a low temperature wall. Total internal reflection laser induced fluorescence (TIR-LIF) method was applied to the fuel film formed by wall impingement gasoline spray. TIR-LIF method can measure the fluorescence from fuel film without the influence of the spray droplet. The fluorescence intensity from the fuel film depends on the film thickness and temperature. In this paper, the temperature dependence of the fluorescence from the fuel film was investigated under the condition that the laser light is totally reflected from the top surface of the film. The order of magnitude of the heat transfers due to the mixing of spray droplets and fuel film, the heat transfer from the ambient gas, and the heat transfer from the wall were compared and modeled. A method to simultaneously calculate the unsteady change in fuel film thickness and fuel film temperature was developed. Then, the model analysis method was applied to the measured experimental data and verified.

To develop a high-performance three-way catalyst, it is necessary to construct an accurate catalytic reaction model. Previously, we have developed the three-way catalyst model under dry simulated exhaust gas conditions. Water, which is contained in the real exhaust gas, however, complicate the phenomenon encountered in the three-way catalyst. Thus, in this study, the effect of water on the catalytic reaction on a palladium-based three-way catalyst was investigated. Moreover, a catalytic reaction model was developed based on the experimental results. As a results, it is found that the constructed model well predicted the light-off behavior of the palladium-based three-way catalyst.

Compression auto-ignition engines have a high potential to achieve both high thermal efficiency and low emission of harmful exhaust gases, but they have a problem of difficult control of ignition and combustion. To realize such control, model-based control has attracted much attention, which requires an ignition model with high prediction accuracy and light computational load. In this study, we developed a machine learning method to enhance the prediction accuracy of an existing empirical model of ignition delay period. The method has multiple existing empirical models, each with different model parameter values, and a neural network that weights the prediction values of them considering the driving conditions and in-cylinder gas states. To obtain data for training and evaluation, transient operation experiments of a diesel engine were conducted under conditions that relatively premixed combustion occurred. The neural network and the model parameters were trained by the stochastic gradient descendant algorithm using the training data, and the proposed method showed much better prediction accuracy for the test data than the single empirical ignition delay period model adapted using the training data. Investigation on each ignition delay period model of the proposed method after training revealed that the prediction values of each model are highly correlated, indicating that the proposed method does not take advantage of the availability of multiple models. However, the proposed method successfully realized the state that multiple ignition delay period models make different predictions, and the neural network weights them based on the driving conditions and in-cylinder gas states.

NO generation mechanism and NOx formation of the inversed diffusion flame (IDF) and MILD combustion formed in the secondary combustion area of waste incinerator were experimentally investigated in order to clarify the NO generation mechanism in the waste incinerator. NO generation mechanism of secondary combustion near the nozzle exit was characterized by NO planar laser induced fluorescence (PLIF), temperature distribution. Besides, ammonia was used as the additives in fuel for understanding the effects of flame structure on NOx formation. The amount of NOx in the exhaust was sampled and analyzed. In order to eliminate the effect of ammonia decomposition, ammonia was replaced with formamide. Results showed that in the case of MILD combustion without ammonia, the NO-LIF signal intensity was weaker than that in IDF without ammonia because the temperature of MILD combustion was more uniform and lower. Furthermore, the NO-LIF signal intensity in downstream of the flame is stronger than that the upstream flame. On the other hand, MILD combustion with ammonia showed larger region of NO-LIF signal intensity than that of IDF with ammonia because the spatially homogeneous reaction region was formed in the condition of MILD combustion. the NO-LIF signal intensity in downstream of the flame is weaker than that the upstream flame. By comparing NO-LIF and OH-LIF in IDF, we found that NO-LIF signal intensity is stronger on the outside of the flame zone than inside. Additionally, IDF and MILD combustion reverse the tendency of NO-LIF signal intensity in the height direction. By comparing with the NO-LIF signal intensity when formamide and ammonia is added, ammonia is not decomposed because there is almost no effect of NO-LIF signal intensity.

For sit-riding type wheeled inverted pendulum personal mobility vehicle (WIP PMV), a rider operates it in a sitting posture and a weight-shifting motion is achieved by only rotating an upper-body around a waist. This causes a drop of acceleration and deceleration ability of the PMV, as more, a rider feels anxious when the PMV inclines back and forth while accelerating and decelerating. To cope these issues, it is important to know rider’s behavior in a sitting posture on an unstable seat and to adopt the characteristics in a driving controller of the PMV. In this paper, we investigated a balancing ability and a rapid weight shifting behavior of a rider sitting on an experimental device that imitated a sit-riding type WIP PMV. In the experiment for evaluation of balancing ability, it was revealed that 8 subjects could keep their balance when a rotational stiffness of a seat was 2.45 Nm/deg. In the experiment of the rapid weight-shifting motion, it was confirmed that there was a dead time of about 0.35s from the time when the subjects visually recognized a start trigger signal, and that a reverse reaction motion where the subject’s body was once tilted forward occurred. In addition, it was clarified that the reaction force of the lower limbs against footrest did not affect the behavior of the rapid motion, and that the lower rotational stiffness of the seat might cause the time delay that the weight started to tilt backward. These results can be expected to be useful index when designing inverted driving controller in sit-riding type WIP PMV.

In noise control, the Filtered-x LMS algorithm is generally used as an adaptive algorithm. This algorithm can achieve stable noise reduction with a simple control system. However, it has a disadvantage that the control cannot be performed when the secondary path characteristics change with the movement of an evaluation point. In this paper, we use a noise control system using neural networks with the learning ability to follow the change of the secondary path characteristics when an evaluation point moves. Conventional studies of noise control using neural networks have problems with their control performance and control success rate. To solve the problems, here we propose a novel structure of neural networks and a method to bias the learning rates according to the type of input signals for noise control system. The effectiveness of each method was investigated by numerical simulations. The results show that the proposed structure using the method to bias the learning rates has the control success rate of more than 90%, which is 30 points better than a conventional structure and that the noise reduction performance is approximately 16 dB, which is approximately 4 dB better than a conventional structure.

Diesel or gasoline particulate filters are included in almost all vehicles to clean emission gas from internal combustion engines. On the other hand, a new regulation for vehicle noise emission has been into operation, and noise from exhaust system must be reduced. Although mufflers with expansion chambers and sound-absorbing material are mainly applied to reduce noise emission, diesel or gasoline particulate filter can reduce noise transmitted through the filter. In this study, sound transmission loss of diesel or gasoline particulate filter is measured and some amount of sound transmission loss is found to be obtained. Then, we propose a method to predict sound transmission loss of diesel and gasoline particulate filters from a microstructure model of ceramic porous material that is applied for filters. Homogenization method for sound-absorbing poroelastic material based on the asymptotic expansion approach is applied in this study. This method was proposed by one of the authors of this study. Microscopic structure images are obtained by scanning electron microscope, and the size of pores are identified by image analysis. Homogenized properties, such as equivalent density and bulk modulus are calculated by performing microscopic analysis with unit cell finite element model. Finally, sound transmission loss is obtained by macroscopic calculation with homogenized properties. Numerical sound transmission loss agrees well with sound transmission loss measured by acoustic tube and simplified exhaust system.

The purpose of this study is to investigate a machine learning method to support upstream of design process by taking a motorcycle design targeting riding comfort as an example. In our previous report, a machine learning decision tree model was trained using temporal factors with the riders’ subjective evaluation as the target variable while the physical and dynamic states of the bike and the riders were used as predictors. In the present study, a high performance deep-learning regression neural network was trained to predict the real-time subjective evaluation of riders using physical state data captured over a certain time period. Deep learning algorithms, however, often yield black box models which are inadequate for the purpose of acquiring design knowledge. Hence, the feature importances of the models were extracted using explainable artificial intelligence (XAI). Relationships between physical states and changes in rider’s pleasure were elucidated by applying XAI on the deep learning model and bike riding data. As the result, two design guidelines to increase riding comfort were obtained as (1) improve the throttle feel and responsiveness, and (2) improve vibration control at the rear of the motorcycle and rear suspension.

The purpose of this paper is to understand the load state of pinion bearings used for a gear unit in an electric railway vehicle. For this purpose, a system capable of measuring the rolling element load distribution of the pinion bearing in an actual gear unit was constructed, and the measurement was conducted under the same conditions as in an actual railway vehicle. Generally, tapered roller bearings are used for pinion bearings in a face-to-face arrangement, which has an axial clearance (endplay value). Therefore, the effect of the endplay value on the rolling element load distribution was investigated by changing the endplay value at specified values, larger values, and smaller values. As a result of measuring the rolling element load, it was found that the rolling element load tended to increase as the transmission torque of the gear unit increased. In addition, large rolling element loads occurred only when the endplay value was smaller than the specified value. The reason for this is that the axial thermal expansion of the pinion shaft exceeded the endplay value, and the bearing became pressurized. The bearing loads (radial and axial loads) were determined from the measured rolling element loads, and it was found that under conditions where the endplay value was smaller than the specified value, the axial load was significantly higher, which could cause a reduction in bearing life and wear of the bearing. Finally, the load distribution factor of the bearing was determined for each condition. It was observed that the load distribution factor decreased as the endplay value increased, and that for the same endplay value, the load factor tended to be smaller for the conditions with lower torque. It was also found that the load distribution of the rolling elements does not change significantly except under conditions where the endplay value is smaller than the specified value.

Spatiotemporal local changes have been extracted for evaluating physiological phenomena by an image reconstruction algorithm of electrical impedance tomography (EIT) using sparse Bayesian learning (SBL). The proposed method identifies a region of interest (ROI) by a priori information on conductivity distribution σ of each biological tissue and automatically learns block-sparsity and temporal-correlation in the identified ROI in the form of blocked column vector (BCV). Two types of numerical simulations are conducted to model lymphedema (LE) and venous edema (VE) that cause spatiotemporal local changes in σ of subcutaneous tissue fluids in human calf phantom which consists of three parts: great saphenous vein (GSV) as ROI and subcutaneous adipose tissue (SAT) and muscle as background. From the results, spatiotemporal local changes in σ are clearly extracted only near GSV by the proposed method even in a field where the time-varying σ in the background is large. Furthermore, the accuracy of the proposed method is evaluated under the variant conditions of conductivity ratios of SAT and muscle relative to GSV, i.e., ρ^GSV/SAT and ρ^GSV/muscle, respectively, and area ratio accuracy ARA^GSV is the highest in the case where ρ^GSV/SAT = ρ^GSV/muscle, which achieves ARA^GSV = 2.241 regardless of the values of ρ^GSV/SAT and ρ^GSV/muscle.

In order to improve handling and stability not only in autonomous vehicles but also in manual vehicles, research had been conducted to appropriately control the two state quantities of vehicle body slip angle and yaw rate. By constructing a state feedback system that centrally controls these two states and applying the walking control technology of a biped robot, natural, smooth and good handling and stability was realized regardless of the driver. In addition, by setting the feedback gain that does not depend on the coefficient of friction of the road surface, it was possible to construct a relatively simple control system without considering the case classification according to the condition of the road surface.