It has been about 25 years since petroleum refining reactors used in hydrocracking and hydrodesulfurization process have been made of 2.25Cr-1Mo-V (F22V : SA-336(M)-F22V) steel. The F22V steel reactors are expected to operate continuously for decades, thus are required to have high reliability to withstand the high-temperature, high-pressure hydrogen environments. Although the assessments of minimum pressurization temperature (MPT) and remaining life of in-service reactors are required for efficient plant management, it is unclear whether the impact properties required for weld metals in the materials and fabrication standard such as API 934-A provide sufficient toughness for hydrogen assisted cracking (HAC) resistance in the welded joints of reactor under hydrogen service. It means that it is not clear how to deal with the assessments of MPT and remaining life of in-service reactors. Therefore, in this study, based on the results of finite element method (FEM) simulations of temperature, stress and hydrogen concentration distributions generating in the reactor wall under service, the HAC influence factors in each stage that is classified into startup, steady-state operation and shutdown were extracted. We then investigated the test method and conditions that simulated these influence factors for evaluating the HAC resistance in each stage of reactor operation. Furthermore, the HAC resistance tests using the welded test plate simulating the girth weld joint of actual reactor were conducted, and its characteristics were evaluated with the focus on the resistance to HAC initiation. In conclusion, the threshold stress intensity factor for HAC initiation, K_IH measured in the HEAC resistance test under reactor startup conditions was conservative, and it was found that the possibility of HAC initiation during reactor operation (K_I ≥ K_IH), i.e., the MPT for reactor to prevent it, can be evaluated by the HEAC resistance test under reactor startup conditions.

We investigated sedimentation in a colloidal dispersion composed of spherical ferromagnetic particles using Brownian dynamic simulations. We developed an algorithm to derive the maximum surface integral rate points, which is the area fraction that prevents the formation of a thin film under each condition, to clarify the conditions for forming a desired thin film in which spherical ferromagnetic particles settle down and exist separately. We derived the maximum surface integral rate points by dosing spherical ferromagnetic particles according to the formation of clusters under each of the 10000 total conditions, by varying the applied magnetic field, magnetic force between the particles, mass density, and temperature of the solution. The main results are summarized as follows. When the strength of the applied magnetic field is increased, the magnetic moment of the particle is strongly restrained in the direction of the magnetic field. Therefore, only the repulsive force caused by the magnetic moment of the particle is applied in the horizontal direction of the particle, leading to the formation of thin films with maximum surface integral rate points of less than 0.243. However, higher-density limiting points were undesirable. Therefore, increasing the random force improved the number of conditions in which the maximum surface integral rate points were greater than 0.243. In the condition where the strength of the applied magnetic field and the random force are large, the maximum surface integral rate points tend to be large; however, the force applied to the particle is large. Depending on the magnitude of the magnetic force between particles, there are conditions where the balance of the force is unstable and the maximum surface integral rate points decrease significantly. However, as the mass density increased, the maximum surface integral rate points increased. We found that by dosing particles according to cluster formation, we could reach a maximum area fraction of 0.322, even though we could not reach an area fraction of 0.403 when all particles formed a thin film.

One of the key functions of a mixer is to achieve uniform mixing of powders or liquids. In this study, a 3D motion mixer with a sealed container without mixing blades was employed. However, the internal flow behavior of powders or liquids within such sealed containers remains unknown, assessing the mixing state requires opening the container for evaluation. To address this issue, a 3D motion mixer with a transparent glass container was utilized, allowing the observation of powder flow within the container via a high-speed video camera. Additionally, a "non-contact dispersion evaluation method" was proposed. In this approach, were mixed with test powders, and the dispersion state was assessed based on the distribution of fluorescent particles visible through the transparent glass container walls. This evaluation method yielded almost the same results comparable to direct dispersion evaluation by boring, and also revealed the effects of powder filling rate and rotational speed in the container on mixing time.

The Ranque-Hilsch vortex tube (RHVT) is a device which can separate compressed gas into high and low energy flows only through its own fluid motion by generating a high-speed swirling flow inside a cylindrical vortex chamber. The energy separation phenomenon of RHVT is called the Ranque-Hilsch effect (RH effect). Although numerous studies have been conducted over the past 90 years since the discovery of the RH effect (1933), the energy separation mechanism has not been elucidated. Previous studies have suggested that the RH effect is caused by the unsteadiness of a high-speed swirling flow inside the cylinder, such as turbulence and acoustic phenomena. However, the unsteady characteristics have not been clarified because it is difficult to obtain time evolution data sufficiently guaranteed physical reliability. In this paper, high-speed unsteady swirling flow inside the cylindrical vortex chamber of RHVT is calculated by direct numerical simulation for the first time. Because the Reynolds number of the flow inside RHVT is high, it is difficult to use direct numerical simulation due to the high computational costs. To reduce the computational costs, we performed direct numerical simulation by setting the viscosity of the working fluid higher than the air. The viscous coefficient of the working fluid is set equal to or larger than 100 times that of the air. We find out that the turbulence intensity is axisymmetric, and the closer to the axial position of inlet, the stronger turbulence occurs. By comparing flows with different viscosities of the working fluid, we also find out that there is a relationship between the RH effect and turbulence in a high-speed swirling flow inside the cylinder, as the energy separation effect increases with increasing turbulence intensity.

The Karman vortex generated in the flow field around bluff bodies has various negative effects on our daily lives. Therefore, control of the flow field has been studied for suppression of the Karman vortex. Recently, advanced control of the flow field has become possible by the development of fluid analysis technology and computers, and the method using feedback control has been attracting attention. The previous study using model predictive control has the disadvantage that the computational cost is enormous because iterative calculations are performed while predicting future responses. In addition, the previous study using feedback control with the flow velocity cannot completely suppress the vortex because time delay occurs between the time of the flow velocity and the time of control due to detecting the flow velocity behind the bluff body. Therefore, the conventional method can only be applied to flow fields where regular vortices are generated, and is considered impractical. To solve these problems, we propose a method of preventing the occurrence of vortices by controlling at the moment of separation by feedback the pressure gradient on the surface of the bluff body near the separation point. The use of the pressure gradient enables real time detection of separation by its sign, and thus enables control according to the occurrence of the vortex, which may be applicable to flow fields where vortices occur irregularly. In this report, we report the results of control of the flow field around a circular cylinder at low Reynolds number to confirm the effectiveness of the proposed method as a basic study for control in the flow field of high Reynolds number. According to the results of numerical calculation, the proposed method effectively suppresses the Karman vortex generated in the flow field of low to high Reynolds number, confirming the effectiveness of the proposed method.

Heat exchangers are used in a wide range of applications and there has been a great deal of research and development on improving their performance, but in recent years there has been a growing demand for miniaturization, especially in energy systems. In the development of a compact heat exchanger with additive manufacturing, the thermal-hydraulic characteristics of a block type heat exchanger were experimentally evaluated. Three types of block type heat exchangers with different internal channel structures were manufactured by using powder bed fusion type metal 3D printer. In the tests, condensation heat exchange tests with cooling water and steam and single-phase heat exchange tests with cooling water and hot water were conducted. As a result, it was confirmed that the heat exchange rate could be improved by staggered arrangement of the channels and by reducing the diameter of the channels. In the condensation heat transfer test, the amount of heat exchange was the same regardless of the amount of cooling water. This is because the steam is fully condensed under all test conditions and condensation heat transfer is dominant. However, the distance at which steam condenses completely is reduced by arranging the channels in a staggered pattern and reducing the channel diameter. From these facts, the heat exchange rate can be further increased by increasing the steam flow rate. It has been confirmed that Shah's equation (Shah, 2022) can predict the results of condensation heat exchange tests with an accuracy of ±10%. In addition, the pressure drop was 0.5~2.0 kPa, indicating that the block heat exchanger has a high heat exchange performance but small pressure drop.

This paper proposes an assist-as-needed (AAN) controller for power-assist of a knee joint that takes into account differences in human and orthotic motion. The AAN control is one control method that adjusts the assist performance according to the wearer’s ability. AAN controllers are often designed to be time-independent so that the wearer can act whenever he/she wants. However, the speed of human motion changes frequently, nevertheless, most previous works did not actively change the desired speed of the orthotic motion. This paper tackles this issue, and we propose a region attractor that changes the speed of the reference motion based on the internal force acting on the lower limb attachment point. In addition, to reduce pain at the attachment point caused by a misalignment between the wearer and the orthosis, the center of rotation (CoR) of the knee is estimated in real-time. Using the estimated CoR, the motor torque that satisfies the desired knee torque is generated while minimizing the internal force in the direction that does not contribute to assist. Through experiments using a dummy robot of the human knee with various torque limits, the effectiveness of the proposed method was verified.

Humans have whole-body viscoelastic connections called anatomy trains (ATs), which include multiple muscles and connective tissues. ATs are expected to realize coordinated motion in passive dynamic walkers with a large number of joints. However, the details of how the coordinated motion is realized are not fully understood. In this paper, we propose a mathematical model of a quasi-passive dynamic walker with whole-body viscoelastic connections inspired by the AT theory, and investigate how the connections improve the walking performance in numerical simulation. We introduced a tension transmission line that connects from the foot to the head, modeled after one of human ATs, the superficial back line (SBL), into the mathematical model of a conventional passive walker. The important features of this whole-body line are that the tension transmission function of the line can be instantly switched on/off by a controller that contracts and relaxes the line, and that the line in a contracted state mechanically interacts with the whole-body joints, depending on the spring constant of the line and the posture of the whole body. The simulation results indicate that the walking performance, i.e., the step count, can be improved under the specific combinations of the spring constant and the on/off timing of the whole-body line. These results suggest a method for tuning the SBL to improve the step count, based on the step count and how posture is disrupted. By accumulating insights through both simulations using this mathematical model and experiments with physical robots, while cross-referencing these approaches, it is anticipated that this will eventually contribute to theoretical analysis in the future.

Digital twins are considered to be one of the most promising technologies for both green transformation and digital transformation. To facilitate the early adoption of digital twin technology, it is crucial to reduce the development time of models that can simulate with low computational cost and high accuracy. However, verifying the accuracy of these models can be time-consuming due to the significant variations in boundary conditions caused by changes in environmental conditions, design requirements, and specifications of products. In this study, we proposed an evaluation method for evaluating the accuracy of reduced order models built through parametric model order reduction. We utilized the conceptual idea of the Monte Carlo method with Latin Hypercube Sampling to randomly and comprehensively generate boundary conditions. Based on these evaluation methods, we have developed an algorithm to determine the order of the projection matrix, which is obtained through singular value decomposition of the original projection matrix generated by parametric model order reduction techniques. We applied this algorithm to a 3D FEM model with linear parameter varying convective heat transfer. The results show that the simulation time of the reduced order model, with an accuracy of 0.1 ℃ or less, can be reduced to 1/200 compared to the original FEM model (full order model), and to 1/12 compared to the original reduced order model generated by conventional parametric model order reduction techniques. This significant reduction in simulation time would enable real-time simulations in the digital twins environment.

Strain engineering is a crucial approach in the engineering field to optimize various physical properties of materials by applying mechanical strain loading. However, it is extremely challenging to find out the best conditions of strain with unprecedented physical properties in the vast strain space consisting of six components. Here, we developed a technical framework that enables efficient exploration of physical properties in the vast strain space based on machine learning (i.e., artificial neural networks), active learning, and high-throughput first-principles calculation. We demonstrated the active learning technique to successfully and efficiently construct an accurate machine learning model for ferroelectric PbTiO₃ with minimal first-principles datasets (only 3.7% of the vast strain-space). Our machine learning model can accurately predict the nonlinear mechanical deformation and electromechanical response in the three components of normal strain loading. We also carried out strain optimization of piezoelectric response using the machine learning model and found that a large piezoelectric response is five times larger than without strain loading. We showed that the physical property explorer framework constructed in this study makes it possible to optimize strain for various material properties in a vast strain space by calculating only a small number of data points. These results suggest paving the way for constructing nonlinear piezoelectric constitutive equations for novel piezoelectric devices via strain engineering.

Future vehicle design will require lightweight, crashworthy, and stiffness, yet it will balance these requirements. A promising approach, multi-material topology optimization, can reasonably place solid materials with different mechanical properties within the structure. In this study, we propose a multi-material topology optimization approach that maximizes energy absorption performance to achieve better results in large deformation scenarios. Unlike conventional methods that assume constant external loads, the objective function of this study is defined as that which maximizes the corresponding reaction force for fixed displacement. The proposed method is demonstrated and discussed through several numerical examples.

During the powder feeding process in the manufacture of cosmetics, powder is dispersed into the air as dust. If inhaled by workers, the dust can have adverse effects on health, such as dust lung and asthma, and can also lead to contamination of the work environment and of the product. Therefore, dust control measures are essential in the powder feeding process. However, production efficiency is declining due to longer feeding times, and new dust control measures are necessary to ensure worker safety and improve production efficiency. The purpose of this paper is to develop a new dust collection system that can prevent dust generation and improve production efficiency in the manufacturing processes of cosmetics. Specifically, we propose a bladeless dust collection system designed to prevent powder dispersion and enhance the adsorption of powder particles onto the liquid surface. This is achieved by elucidating the forces exerted on the powder particles by the airflow. The effectiveness of the proposed system is demonstrated through powder feeding experiments. The powder feeding was carried out under conditions identical to those in the actual cosmetics manufacturing process. It was observed that dust control was achievable, confirming the absence of any practical issues. Furthermore, because the dust collection system can be installed in accordance with the agitation location, it is possible to implement dust control measures anywhere, and this mobility should improve production efficiency. In conclusion, the effectiveness of the bladeless dust collection system was confirmed.

To efficiently operate complicated production systems such as mixed-flow production, production system simulation is a promising tool. However, constructing accurate simulation models requires significant effort. In this study, we proposed a data-driven and multi-scale modeling approach that optimizes model configuration by combining multiple modeling methods. In order to verify the usefulness of the proposed approach, we organized the possible model configurations from the perspective of available data from target system and conducted computational experiments for flow shop production systems to compare various model configurations. The experimental results show that the optimal model configuration strongly depends on the complexity of target system and models to be identified, and it is important to divide the model so as to reduce the complexity.

In the maintenance service business, introducing IoT solutions is an expected strategy to improve the maintenance efficiency. However, it is difficult to predict in advance the quantitative effects of IoT solutions on the overall business. This study proposes the simulation approach to quantitatively evaluate the improvement effect of IoT solutions using the multi agent modeling for maintenance business. This model includes the behavior of maintenance workers and is characterized by evaluating the resource workload. The effectiveness of the proposed approach is verified in the case study of a dispatch-type maintenance service for wind turbines. The availabilities of wind turbines are compared in two prognosis solution introduction stories with six cases. The change trends in availability by introducing prognosis solutions can be explained by both the reduction in asset failures and the resource status of maintenance workers. In the simulation result, the introduction of anomaly prediction without changing the current maintenance policy resulted in an increase in workload and rather worsened availability, which indicates that the proposed method can assess the risk of reduced availability. Also, the availability improved significantly when combined with a change in maintenance policy, such as omitting periodic inspections. These results indicate that the proposed method can evaluate the synergistic effects of prognosis solutions and maintenance policy changes.

In many areas of manufacturing, there is a growing need for imaging technology that non-destructively inspects the surface of products. However, detecting micro-defects on curved surfaces is challenging because the appropriate imaging conditions can vary depending on the shape of the curved surface. To address this problem, we have developed a method for visualizing micro-defects on curved surfaces that uses an imaging system to obtain a color map of light directions reflected from the surface. This imaging system is equipped with a stripe-patterned multicolor filter and a radially diffusing illumination light. The unique characteristics of this imaging system enable it to capture drastic spatial changes in the direction of light, which are caused by micro-defects on the curved surfaces, as differences in image color. Furthermore, we have also constructed an image-processing algorithm that automatically detects these micro-defects. The key point of this image-processing algorithm is that it automatically detects areas where drastic color changes have occurred due to micro-defects by using spatial frequency filtering techniques. As a result, the developed method was shown to have the ability to automatically detect micro-defects on curved surfaces with depths of several tens of micrometers.

Automobile body design is performed using structural analysis by numerical calculation, but to make it efficient, it is important to establish a technique that can quickly derive the structure satisfying the target value. Although optimization methods are used in many design studies, problems sometimes arise in the actual design stage, and conventional methods in which designers make iterative design changes based on trial and error are still widely used. However, when designing the mutual average compliance defined by the sum of the weights of displacements including displacements other than the external force acting point, it is difficult to derive a design plan because the strain energy, which is a general rigidity index, is not proportional to the mutual average compliance. Furthermore, even if the design sensitivity is calculated, it is difficult for the designer to estimate the type of required stiffness from the sensitivity of the scalar quantity without the displacement shape, as a result, structure examination requires more time. In this paper, we propose a method to reduce the absolute value of mutual average compliance using eigenvalues and eigenmodes obtained by eigenvalue decomposition of the stiffness matrix. First, the eigenvalues and eigenmodes of the stiffness matrix are explained, and the relationship between the mutual mean compliance is clarified. Using a cantilever beam model as an example, it was confirmed that the designer could efficiently carry out the conventional iterative design change by replacing the design problem of mutual average compliance with the design problem of eigenvalues of the stiffness matrix.

This study proposes a methodology to generalize and reduce the computational cost of sensitivity analysis for static linear elastic systems in topology optimization. The process is fully generalized by applying sensitivity analysis in conjunction with the adjoint variable method and automatic differentiation. The design sensitivities can be computed independently of any evaluation function and the material interpolation method for static elastic systems. Furthermore, the issue of excessive computational memory required by automatic differentiation is resolved by utilizing the adjoint variable method. To demonstrate the generality of the proposed method, we focus on multi-material topology optimization for isotropic linear elastic systems, evaluating compliance and maximum stress based on the L_p-norm. Additionally, the extended SIMP and Discrete Material Optimization (DMO) methods are investigated as interpolation schemes for material properties in multi-material problems. Finally, the finite difference method and automatic differentiation are benchmarked against other sensitivity analysis methods, and the proposed method is evaluated in terms of accuracy, computational cost, and memory usage. Moreover, examples of topology optimization for a 3-dimensional problem are presented to demonstrate the applicability of the proposed method to large-scale analyses.

According to a survey conducted by the Ministry of Health, Labour and Welfare in Japan in 2023, the numbers of fatalities and injuries at factories and other production sites have been decreasing, but the numbers of injuries have been increasing. For this study, musculoskeletal model analysis was used to elucidate the starting motion of cart pushing as a factory task. For 10 healthy male study participants, body coordinate data and floor reaction force data were obtained using a three-dimensional motion analysis system and two floor reaction force plates during cart pushing without a weight and with a 100 kg weight. Then, lower limb joint angles, lower limb joint moments, lower limb muscle forces, and lumbar muscle forces were evaluated using OpenSim musculoskeletal model analysis software. The following results were obtained. The amplitudes of the pelvis list moment and pelvis rotation moment were significantly higher when pushing the weight. The maximum values of muscle force in the gluteus maximus muscle and tibialis anterior muscle of the left leg were significantly higher. Also, the integral value of the rectus femoris muscle was significantly higher. The maximum and integral values of the external abdominal oblique muscle right were significantly higher. Results demonstrated that moving a cart with a 100 kg weight on it required greater force in the hip and knee joint extension muscles, the ankle joint dorsiflexion muscle, and the external abdominal oblique muscle right.

Most rolling stocks in Japan have tachometer generators on their axle ends. The only function of the tachometer generator is measuring rotational speed of the axle based on the frequency of the output signal. On the other hand, the amplitude of the output signal reflects not only the speed but also the gap between the tachometer-generator sensor and the axle-end gear. First, this paper shows that the amplitude of the output signal can change during running and the amplitude has a strong correlation with the position of the rolling stock. Generally, the sensor is fixed on the axle box and there is a small allowance between the axle box and the axle journal bearing. Therefore, the relative position between the axle box and the axle can change due to the interaction among the bogie, the axle, and the rail. This fact indicates the possibility that the tachometer generators can be used for some sort of anomaly detections. Based on this, this paper shows some experimental results of the anomaly detection on the rail by means of the tachometer generator.

In this study, we focused on hanging sleepers and rail-head surface irregularities, which are considered to affect the fatigue life of aged rails, and statistically analyzed their occurrence on commercial lines. The occurrence of these conditions was expressed by the probability density function, which confirmed that the gamma distribution fits well. In addition, fatigue tests were conducted on aged rails, where these conditions were significantly present. Aged rail bending fatigue tests (cumulative passing tonnage: 800 million tons) were conducted, and significant hanging sleepers and rail-head surface irregularities were observed. It was confirmed that the fatigue strength of the rail at the end of the current replacement cycle is equivalent to that of the previously tested aged rail (cumulative passing tonnage: 540 million tons). It is expected that the replacement cycle can be further extended if the hanging sleepers and rail-head surface irregularities are below a certain value. The study also proposed a probabilistic safety evaluation method and control values for the long-term use of rails. The rail health index, calculated from the rail bottom bending stress and fatigue limit, was proposed as an index for preventing rail breakage at locations where significant hanging sleepers and rail-head surface irregularities occur when the cycle replacement is extended or eliminated.

A method for evaluating snow accretion around a train bogie was investigated using a newly developed model vehicle running apparatus. The model train running apparatus is capable of running a 1/11 scale model train on a 384 m track at a maximum speed of 120 km/h. The phenomenon of snow accretion due to train running was experimentally simulated by setting up a simulated snow condition in which simulated snow particles were deposited on the track. Comparison of snow accretion on the end blocking vertical panel downstream of the bogies of an intermediate car between real cars and model running tests confirmed that the general characteristics of the snow distribution adhered on the panel can be reproduced. As a result of investigating the growth process of the snow accretion from the model experiment using image processing and 3D scanning, it is found that the initial snow accretion starts from the under side of the vertical panel downstream of the bogie, especially near the wheels. As the travel distance of the snowed section increased, the snow accretion on the under side of the vertical panel extended over the bogie cavity, and snow accretion also occurred on the top and sides of the vertical panel. Eventually, the entire vertical panel was found to be covered with snow accretion.