The friction force produced during the plastic process induces a non-linear strain gradient on the material. This affects the history of microstructural deformation depending on the size. The miniaturization of the product increases the importance of an accurate evaluation of the effect of friction on the size-dependent plastic process. In this study, we investigated the effect of the relative size of the polycrystalline microstructure on the roll-forming process by establishing a nonlocal finite element square (FE²) model based on the second-order homogenization method with boundary condition applicable to arbitrary contact situations including friction. Using the established model, we performed FE² simulations of the roll-forming process of the polycrystalline metal sheets with different grain sizes. Results obtained clarified that the distribution of the accumulated strain was characterized by friction and the relative microstructure size. An additional shear strain occurred at the surface layer of the sheet metal during the roll-forming process. A larger additional shear strain was observed in a surface layer for the sheet metal with smaller grains, whereas a high strain gradient occurred at the polycrystalline microstructure for the sheet with a larger microstructure.

A glass fiber reinforced plastics (GFRP) adhesive joint, designed to improve the rigidity of aluminum structures, was developed. The fatigue strength and failure detectability of the joint were investigated to support the practical application of lightweight multi-material structures, which are a promising means of energy saving in various modes of transportation. As a result of applying a second-generation acrylic adhesive, the SN curve of a structure stiffened with GFRP did not change largely regardless of adhesive thickness. The roughness of the bonding surface, however, affected the fatigue life significantly. By comparing the acquired fatigue lives with the SN curve for adhesive joints specified in the EN standard, it was found that the designed joint had a high safety margin in terms of fatigue strength only when the effective shear stress defined in the standard was used for the vertical axis value of the SN diagram. In addition, to confirm delamination of the bonding area, a method of detecting the delamination using a strain gauge and a digital camera was used during fatigue tests. Although the output value of the strain gauge decreased gradually at the beginning of the tests, the trend changed clearly after the delamination reached the center of the strain gauge. Furthermore, the debonded area was detected during delamination propagation using a camera and quantitatively evaluated through binarization processing. Thus, the goal of realizing stiffening with a high safety margin and easy inspection with GFRP adhesion was achieved in this study.

Mechanical nonreciprocity is outside the scope of conventional symmetry in mechanical and physical systems, enabling a range of engineering applications such as asymmetric transmission, energy conversion, impact protection, and biological manipulation in unprecedented ways. This study numerically investigates the asymmetric inhomogeneous deformation of a nonreciprocal gel under cylindrical indentation. A nonreciprocal gel exhibits significant mechanical nonreciprocity owing to the buckling of the nanosheets embedded within the hydrogel matrix and is distinguished by three key characteristics: anisotropy, tension-compression asymmetry (TCA), and nonlinearity. To study the effects of these three factors systematically and quantitatively, finite element analysis of the cylindrical indentation is conducted using non-TCA (nTCA), TCA, and TCA-nonlinear (TCAn) models. The results reveal that the nTCA model shows almost no asymmetric deformation, whereas the TCA and TCAn models exhibit significant asymmetric deformation and show good quantitative agreement with the experimental results. The interaction between the TCA factor and the interfacial slip between the cylinder and the gel is essential for asymmetric inhomogeneous deformation. The frictionless condition is also validated in comparison with the previously reported experimental data. The slight difference in results between the TCA and TCAn models is attributed to strain relaxation in the unbuckled direction due to interfacial slip.

One of the effective pathways toward a sustainable society is to rapidly phase out coal. To replace coal and coal coke used in both power and industrial sectors, torrefied solid biofuels are expected to serve as an alternative fuel. This study focuses on the wet torrefaction (WT) process as a biomass reforming treatment, and investigates universal evaluation method of energy properties for wet torrefied woody biomass (WT-WB). The solid mass yield (SMY) of WT cedar and castanopsis decreases with increasing WT residence time and WT temperature, which results from the mass reduction due to WT reaction such as pyrolysis and hydrolysis. The higher heating value (HHV) property of WT cedar and castanopsis can be expressed as a single experimental correlation as a function of SMY by using the enhancement factor of HHV (EFH). From experimental results on three constituent polymers in WT cedar and castanopsis, it is found that the solid mass yields of cellulose and hemicellulose decrease and the solid mass yield of lignin remains nearly constant or increases as WT reaction proceeds. The mass gain of lignin is correlated with the total mass reduction of cellulose and hemicellulose. Assuming that the solid mass yield properties of three constituent polymers derived from the average data for both softwood cedar and hardwood castanopsis can be applied to any WT-WB, experimental correlations of solid mass yield for cellulose, hemicellulose and lignin are obtained using multiple regression analysis. The evaluation method of energy properties of WT-WB for given WT conditions is constructed by combining both experimental correlations of solid mass yields of three constituent polymers and EFH. From the comparison between experimental data and estimated SMY and HHV, it can be stated that the method constructed here is useful for evaluating energy properties of WT-WB.

Dynamic analysis of catenary–pantograph systems is an important engineering issue because it contributes to predicting current collection performance and understanding accident phenomena. This field has been active since around the 1960s, and many models have been developed and experimentally validated. More recently, installation defects and their effects on current collection performance have also been studied. However, experimental validation of models under defective conditions remains limited. In particular, installation defects in the steady arm, an important component of the system, have not been sufficiently investigated. Therefore, in this study, full-scale experiments and numerical analysis were combined to investigate the effects of a steady arm installation defect (SAID) on Shinkansen system dynamics. First, excitation tests were conducted by applying impulsive external forces to an actual catenary. The experimental results revealed that SAIDs exhibit nonsmooth nonlinear behavior. The validity and limitations of the linear catenary model were clarified through comparison with the experimental results. Full-scale running tests further showed that SAID-induced nonlinearity significantly affects contact wire uplift in the high-speed range. Subsequently, a bilinear stiffness model was introduced to represent a SAID, and its parameters were identified by minimizing the residual sum of squares between the excitation test results and simulation results. The developed model was also shown to reproduce transient responses observed in running tests with a SAID. These findings clarify the range of validity and the limitations of the linear model under conditions with SAIDs and provide a methodology for identifying the nonlinear characteristics associated with such defects. The findings of this study are expected to contribute to efficient maintenance based on digital technologies, the importance of which is increasing in response to the aging of railway infrastructure and the shrinking labor force.

Structural singularities in catenary–pantograph systems, such as connectors and splicers, induce transient responses that accelerate local degradation and maintenance. While recent analytical work has clarified the fundamental transient-wave mechanism at structural singularities, the local mechanical response around a singularity in a practically relevant setting remains insufficiently understood. We therefore conduct complementary theoretical and full-scale experimental investigations of these transients for a pantograph traversing a local structural singularity. We formulate a tractable analytical model, namely a simply supported, tensioned beam coupled to a local mass–spring singularity and subjected to a moving load. Closed-form expressions clarify the wave interactions at the singularity and the parameter dependence of contact wire strain at the singularity. Full-scale experiments using actual equipment validate the proposed model over a wide range of operating conditions, with various singularity locations, and train speeds. The proposed model faithfully captures the qualitative behavior within the time interval of engineering interest and reproduces the maximum and minimum strains with quantitative accuracy up to 200 km/h. Mechanistically, the minimum strain is shown to originate from waves radiated when the load starts to move, leading to a monotonic increase with both singularity mass and train speed, whereas the maximum strain exhibits a non-monotonic dependence arising from interference between local singularity motion and incoming waves. These results provide useful insights for fitting design and construction guidelines.

Frequency-averaged energy transmissibility and energy dissipation are formulated for a 2-DOF vibration system with both a coupling spring and a coupling damper, and their usefulness for vibration control design is discussed. In the previous paper, the authors presented the derivation of energy transmissibility for a 2-DOF vibration system with only a coupling spring and discussed the interpretation and control of vibration phenomena using this concept. This quantity is closely related to the coupling loss factor in SEA (Statistical Energy Analysis)-based formulations, but the term is used here to emphasize its interpretation as a design-oriented energy transfer index in a low-degree-of-freedom vibration system. In this paper, a 2-DOF vibration system with both a coupling spring and a coupling damper is considered. Based on frequency-averaged input power and vibration energy under broadband excitation, simplified expressions for energy transmissibility and energy dissipation are derived and organized to clarify the role of each parameter. The simplified expressions allow the power transferred between the two subsystems to be clearly divided into terms proportional to energies of different magnitudes, and the effect of the coupling damper on the frequency-averaged power balance is clarified for each term. A numerical analysis of the effects of each parameter revealed that the coupling damper affects not only the magnitude of the energy transmissibility but also, particularly, the effective energy dissipation of each subsystem. Furthermore, as an example of vibration control based on energy transmissibility and energy dissipation, the reduction of engine shake in a car is demonstrated and energy-based design indices for a non-conservative coupling system are discussed.

This paper proposes a method to extend the partial-derivative neural network (PDNN) to a multi-degree-of-freedom approach. In a previous study, the authors developed a single-PDNN (s-PDNN) for a single-degree-of-freedom nonlinear vibrating system and verified its validity numerically. s-PDNN is characterized as a gray-boxed neural network that identifies nonlinear forces by extracting linear parameters through partial differentiation of external forces using physical quantities such as acceleration, velocity, and displacement— which constitute the equations of motion—within the local network operating behind the nonlinear sub-network. In the present study, a multi-PDNN (m-PDNN) was constructed by extending this concept to multi-degree-of-freedom nonlinear vibrating systems and its practicality was verified using experimental data. In m-PDNN, the external forces and responses acting on the target multi-degree-of-freedom nonlinear vibrating system are measured, an s-PDNN is constructed for each degree of freedom, and linear parameters and nonlinear forces are identified from the balance between external and internal forces at each degree of freedom. In the experiments, single- and two-degree-of-freedom systems were constructed, and the practicality of the proposed method was confirmed by identifying linear parameters and nonlinear forces from the measured data. The results confirmed that even under actual measurement noise, it is possible to identify linear parameters and nonlinear forces with a certain degree of accuracy for both single- and two-degree-of-freedom systems.

In this study, we propose a frequency-domain identification method using a neural network (NN) for representative nonlinear vibration systems conceivable in mechanical vibration, specifically single-degree-of-freedom (1-DoF) nonlinear systems with nonlinear spring characteristics symmetric about the origin. The method identifies the linear parameters and nonlinear forces of a given system by using the frequency response obtained from experiments as input to a NN model. Then, the accuracy of the identification is evaluated using values calculated from the equation of motion of the imaginary part. In addition, the NN identification uses a learning schedule with an exponential function to prevent the model from becoming trapped in a local minimum, and the parameters in the function are adjusted to improve the accuracy of the identification. To evaluate the performance of the proposed approach, we first considered a 1-DoF Duffing system as a numerical example. When the method was applied to the response of the system by numerical analysis, the results showed that the identification was performed correctly, and its accuracy was evaluated. In addition, we also performed an identification using data with noise added to the numerically obtained responses to confirm the applicability of the proposed method in a realistic environment. The results showed that the proposed method performed well even under these conditions. Finally, we constructed an experimental model of a single-degree-of-freedom system with a piecewise linear spring characteristic due to clearance, and identification was performed using the experimental response. The identification and its accuracy evaluation were performed successfully, which demonstrates the applicability of the proposed method to real structures. Thus, our findings indicate that the proposed approach is numerically valid and that it can be applied to real structures.

Accurate severity assessment during emergency medical service (EMS) calls is essential for timely dispatch and efficient resource allocation. This study investigates whether callers’ voice acoustics can provide a low-latency, acoustics-only decision-support signal that operates in parallel with protocol-guided triage in operational call centers. Severity labels were defined using post-transport clinical severity at the first medical examination after transport (Japanese shōbyō teido) as recorded in EMS outcome records. This label should be interpreted as a clinically meaningful downstream outcome surrogate rather than a direct observation of urgency at call time. Accordingly, because this outcome-based label may not fully align with call-time dispatch priority, the proposed score is positioned as a protocol-parallel auxiliary alert cue rather than a replacement for protocol triage or a direct model of dispatcher-perceived urgency. Using anonymized Japanese EMS call recordings from the Tokyo Fire Department, we analyzed a balanced subset of 204 calls (102 serious, 102 minor) after excluding moderate cases. We extracted call-level acoustic descriptors (F0 statistics, MFCCs, Mel-band energies, spectral measures, and voice-quality features) and conducted Mann–Whitney U tests with false discovery rate (FDR) correction, revealing systematic differences, particularly in pitch variability and spectral energy patterns. Robustness was evaluated using a waveform-level Wiener-filtering baseline and systematic data augmentation (pitch shifting, gain perturbation, and additive noise) under stratified cross-validation. Logistic regression, RBF-kernel support vector machine, and random forest models were assessed using area under the ROC curve (AUC) and recall-focused operating points selected to satisfy a target recall for serious cases (Recall_serious) of at least 0.90. Within the present internal cross-validation setting, full augmentation yielded the best discrimination, whereas Wiener filtering provided limited and model-dependent gains. The best configuration within the present dataset (random forest with full augmentation) achieved AUC = 0.93 and Recall_serious = 0.933. These results suggest the potential feasibility of integrating an acoustics-only severity score as a complementary alert signal, while highlighting limitations when vocal arousal cues are weak or caller speech is sparse.