This study concerns a method for detecting the structural anomaly of F type support information board from acceleration measurements. The evaluation of structural anomaly is performed by the autoencoder method．By the proposed method, training of the autoencoder is conducted from data of normal conditions, and diagnosis of the structure is conducted from reconstruction error of the autoencoder. To validate the effectiveness of the method, the method is applied to long-term measurement data from actual equipment. To evaluate the detection accuracy and reliability of the method, the method is applied to sites with fluctuating natural frequencies and sites with stable natural frequencies. As a result, the proposed method with the autoencoder is able to correctly evaluate anomalies of the structure with more than 80% detection and a low false positive rate of 5%, and is able to evaluate the normality of the structure stably over a long period of time.

Young’s modulus is one of the most important mechanical properties of materials, and several methods, such as the indentation method, ultrasonic pulse method, and laser-induced surface acoustic wave method, have been developed to measure it both statistically and dynamically. However, these methods have characteristic advantages and disadvantages, including accuracy, specimen shape, environment temperature, and equipment simplicity. In this study, a simple identification method for Young’s modulus based on ball impact tests was developed. In this method, two hard balls with different Young’s moduli were impacted on a flat homogeneous isotropic target material to make their maximum contact radii equal. Young’s modulus of the target material was identified by measuring the rebound velocities and using the predetermined material properties of each ball. The fundamental equation was derived based on Hertzian contact theory. The proposed method was verified by ball impact simulations using the finite-element method and the ball impact tests. The simulations confirmed that Young’s modulus of the target identified by the proposed method agrees well with the input value, irrespective of the target material properties. In addition, the ball impact tests revealed that the proposed method exhibits the potential for identifying Young's modulus, although improvements in identification accuracy are necessary.

Critical resolved shear stress (CRSS) determines the ease of slip deformation of metals. In crystal structures with low symmetry, such as the HCP structure, non-equivalent slip systems exist. CRSSs are different between the non-equivalent slip systems. The CRSS ratios between slip systems is an important factor governing the plastic deformation mechanism. Thus, methods for easily evaluating the CRSS ratios will play an important role to understand the deformation mechanism. In this study, we developed and validated a system to predict CRSS ratios for metal materials using spatial distributions of normal strain in the tensile direction and crystal orientation. The validation was performed by predicting CRSS ratios of HCP metals using the strain distributions which was obtained by crystal plasticity finite element analysis (CPFEM). The results showed that CRSS ratios were successfully predicted under conditions where multiple slip systems were activated, but in some cases the prediction failed. Investigating the cause of failure, we found that predictions failed under the following conditions: (i) only one slip system was active, (ii) distributions of Schmid factor for individual slip systems were similar to each other, and (iii) multiple slip systems were activated in the same region. We also proposed methods to solve the problems.

A magnetic non-destructive inspection method based on measurement of the magnetic flux leakage near the material surface has been studied for carbon steel pipes used in various plants. However, in actual plants, stress fluctuations caused by changes in the temperature and internal pressure of the fluid passing through the pipes may affect the residual magnetization state due to magnetomechanical effect and prevent accurate inspection. In this paper, we applied cyclic stress to pre-magnetized STPG370 carbon steel up to 10⁷ cycles and measured the change of the magnetization state as the change of the magnetic field near the specimen surface. As a result, under the environmental magnetic field, the magnetization (or corresponding magnetic field near the specimen surface) of the specimen decreases significantly after the first few cycles of stress loading, which suggests the magnetization state approaches the anhysteretic (stable) magnetization state. Furthermore, the magnetization gradually decreases further with higher cycles of stress loading.

Numerical simulation was conducted in this study to calculate the agitating torque of the high viscosity Newtonian fluid by the explicit-MPS method using virtual boundary particles for the agitating container and the agitating flapper. At first, a calculation of the resting hydraulic pressure was carried out to confirm the pressure of fluid particles in the vicinity of the flapper. As a result, it was confirmed that the average of the pressure distribution of the fluid particles corresponded with a theoretical value of the resting hydraulic pressure. Considering this result, trial calculation of flapper agitating torque was calculated from the pressure of the fluid particles in the vicinity of the rotated flapper. The flapper agitating torque was calculated from the torque caused by the differential pressure between front and back faces of the flapper and the torque caused by the shear stress of the fluid particle near the flapper edges. From this calculation, stable agitating torques were obtained for each time step at each flapper speed, and consistent with the high viscous Newtonian fluid agitating test results which showed the linear relationship between the number of flapper speed and agitation torque. Furthermore, it was confirmed that the agitation torques were dominated by the torque caused by the pressure between the front and back faces of the flapper.

This study focused on the biogas-air premixed flames formed on a flat burner under the conditions of oxygen enrichment and hydrogen addition to improve the combustion characteristics of biogas-air premixtures. The objectives of this study were to study the effects of oxygen concentration in the oxidizer and hydrogen concentration in the fuel on the intrinsic instability of premixed flames, and to study the effects of flame instability on the flame shape and behavior. Photographs of flames were taken by a digital single-lens reflex camera, and the emission of light from flames was acquired by a photodiode. Cellular structures of flame fronts were identified by photographs taken. The light emission was analyzed to clarify the fluctuating characteristics by time series analysis. Cellular flames were observed at equivalent ratios near the lower flammable limit under the conditions of oxygen enrichment and hydrogen addition. The cellular-flame range decreased with increasing oxygen concentration in the oxidizer and increased with increasing hydrogen concentration in the fuel. Chaotic time series analysis was performed on the light emission to reconstruct the attractor. As a result, regardless of the oxygen concentration, the attractor at the lower flammable limit exhibited quasi-periodic behavior slightly. On the other hand, the attractor near the lower flammable limit became more complex with increasing hydrogen concentration in the fuel. These results indicated that the oxygen enrichment and hydrogen addition affected the intrinsic instability and flame shape and behavior.

In this paper, we propose a spatial measurement method based on object geometry and velocity information using past measurement information from 3D-LiDAR. If geometric information is obtained from 3D-LiDAR, it can be applied to global localization, object identification, etc., and robots will be able to run autonomously with greater stability. However, 3D-LiDAR cannot always measure point densities high enough to acquire sufficient geometric information. Conventional methods have been used to control point density by interpolating areas of low point density or by varying the rotational scanning speed of 3D-LiDAR. However, these methods have problems such as the shape of the measured object differing from the real object and the loss of real-time measurement. In this paper, therefore, a remaining time is set for each point based on the shape information of the object, and each point is allowed to remain until the remaining time is reached, depending on the robot’s movement. The remaining time is set so that the point density is preferentially increased in areas where shape information is estimated to be abundant and in distant areas where the point density tends to decrease due to the LiDAR mechanism. For dynamic point clouds, velocity information is added, and the point cloud is moved to the estimated position the next time and superimposed on the latest dynamic point cloud while considering the distance between points, thereby improving point density while suppressing shape collapse. We demonstrate the usefulness of this method through experiments on actual equipment.

In automatic harvesting of vegetables such as cabbages which are loaded in containers, autonomous driving of carriers is needed to carry containers. In order to operate multiple harvesters in a wide field such as in Hokkaido, it is necessary to run multiple carriers at the same time. The carrier needs to switch from forward to reverse for connection with the harvester. Moreover, the crop area decreases and the drivable area changes as the harvest progresses. In this paper, we propose a path planning method for multiple carriers that considers these features and shortens the total waiting time of harvesters after they finish harvesting in order to maximize farm work. In this method, the optimal paths are decided by Conflict Tree expansion. A path planning method of single agent considering the switching to reverse from forward is also proposed. In the actual driving, speed control is performed to avoid collision considering time lag from expectations of path planning. Finally, we show experimental results and explain the effectiveness of the proposed method. Sliding mode control is used for autonomous driving of the harvester and carrier. We confirmed that our method allows vehicles to travel simultaneously without collision or being stuck.

An analytical method was considered for estimating vehicle interior noise and vibration and identifying main contributors in case the power source is replaced by combining operational (OTPA) and component transfer path analysis (CTPA) methods. A simple vehicle model equipped with motor and small test bench was prepared for the basic consideration in this study. In the method, OTPA and CTPA were applied to the simple vehicle model in the following three steps. In the first step, OTPA was applied to the vehicle model at the operational condition to obtain the contribution and the transfer functions from the vibration at multiple reference points to the response point. In the second steps, CTPA was applied to another motor on the test bench to obtain the blocked force. Subsequently, the reference point (motor attachment point) vibration on the vehicle model was estimated by using the blocked force and the transfer function in case the motor was replaced on the vehicle model. Finally, the estimated reference point vibration was integrated into OTPA model, and the response point vibration was calculated. Through the simple verification test using the vehicle model and test bench, the response point vibration was observed to be estimated accurately by the method. In addition, the high contributing motor attachment point was also indicated correctly.

The computational simulation of the rolling process of the polycrystalline metal plate is performed using the second-order homogenization finite element (FE) method. The dynamically updated boundary condition (BC), which can describe the relative slip displacement under counter-friction force, is introduced to reproduce the progress of nodal displacement and force BCs during the rolling process. To directly solve the displacement under the BC accompanying friction force, the components of the stiffness matrix constructed by the macroscopic FE structure are edited using the friction coefficient and the contact angle. The elasto-viscoplastic mechanical behavior of the microscopic polycrystalline structure is described by the conventional crystalline plasticity FE method, and the relative scale-depended micro- to macroscopic FE model is established using the second-order homogenization method. The computational simulations of the rolling process with different friction coefficients, rolling radii, compression ratios, and grain sizes were performed using the established FE model. The macroscopic computational results could represent the fundamental mechanical behaviors of the rolling process, e.g., reduction of the plate thickness, increase and decrease in the rolling force owing to changes in the contact area, tensile and compression stresses at the inlet and outlet areas of rolling, X-shaped local deformation band beneath the work roll, increase in the rolling load owing to the friction coefficient, work roll radius, and compression ratio. Furthermore, the microscopic computational results clarified that nonuniform deformation in the polycrystalline structure changes depending on the macroscopic position, and the nonuniformity increases with the compression ratio. The microstructure located at the upper area of the rolled material experiences a two-stage deformation characterized by the contact angle while such a two-stage deformation is not observed in the microstructure located inner area of the rolled material.

Finite element mesh quality, which affects analysis accuracy and computational cost, is usually determined based on the designer's experience. In this study, we propose a GCN (Graph Convolutional network)-based method for evaluating the quality of 3D meshes composed of shell elements. In the proposed method, a three-layer network is constructed using GENConv with graph adjacency matrix and feature matrix as input. The feature matrix is created based on geometric shapes such as coordinates and the mesh qualities such as aspect ratio. A dual graph is constructed by converting finite elements into graph nodes and their adjacent parts into graph edges, and the created features are given to the graph nodes for learning. The proposed method is applied to a dataset of automotive side-member FE models and the effectiveness of the constructed network and the introduced mesh features is confirmed. We also visualize the attribution of mesh features and important factors in the obtained results by calculating the integrated gradients of the network. This allows us to select the important mesh features both positively and negatively and explains the basis for our predictions.

Currently, elderly people of 65 years and older constitute 29.1% of Japan’s population. Frail patients are 8.7% of all elderly people 65 and older. Frailty is a weak state both physically and mentally. The probability of illness increases with age. Pole walking is an exercise designed to improve muscle strength and thereby prevent frailty in elderly people. This study applied musculoskeletal model analysis during pole walking and normal walking to clarify pole walking training effects and balance effects. Seven people with 35 plug-in gait markers attached to the body surface were examined for this study. Normal walking and pole walking were measured using a three-dimensional motion analysis system (Vicon Motion Systems) and two force plates. Position and force data were acquired at velocities of 58, 77, and 96 bpm. Measured data were analyzed using musculoskeletal model analysis software (OpenSim) procedures: scaling, inverse kinematics, residual reduction algorithm, inverse dynamics, and computer muscle control. Results obtained using a musculoskeletal model indicate details of muscle force, lower limb joint moments, and lumbar moments during pole walking. Pole walking is effective for the training of muscles working against gravity in cases of lower walking speed (58 bpm). However, cases with higher (96 bpm) walking speed were associated with better conditions for lumbar muscle training.

Due to growing demand and rising cost for delivery services, companies in several sectors have implemented truck sharing to achieve efficient logistics over the past few years. This study proposes a two-stage stochastic program to design multi-supplier collaborative logistics networks with truck sharing under demand uncertainty. The stochastic model optimizes the number of trucks owned by suppliers as a strategic decision and the freight volume by transportation mode and route as operational decisions. By applying the model to a case study of automotive parts delivery in various demand conditions, the impacts of the level and the realization of demand uncertainty on the optimal number of trucks and the objective are evaluated. The results at a variety of demand deviation and skewness show that truck sharing helps to reduce the number of trucks and improve the utilization rate especially when the demand is expected to increase. Furthermore, we demonstrate the performance of the optimal network with respect to synchronous and asynchronous demand for two products by using a simulation-based approach. Our results show that the optimal distribution decisions highly depend on the relationship between the transportation distance and the price of each transportation mode, and truck sharing leads to an economic benefit even when the demand synchronizes.