In this study, damage analysis was performed using 2D micro-scale FEM models of sintered Ag nanoparticles and Ag microparticles to investigate the effect of the internal microstructure of Ag particles on the fatigue crack propagation characteristics. The 2D micro-scale FEM model was able to reproduce the behavior of crack propagation through pores, and the difference in crack propagation rate between Ag nanoparticle and Ag microparticle sintered bodies was reproduced. However, the 2D microscale FEM model could not reproduce the periodic structure in the model thickness direction, resulting in extremely large porosity in the crack propagation path of the Ag nanoparticle sintered body, and consequently the power exponent of the fatigue crack growth law could not be accurately reproduced.

The relaxation behavior of UV adhesives during curing was measured by dynamic viscoelasticity tests during continuous UV irradiation. The results showed that the UV adhesives reached the gelation point at about 0.1 degrees of cure, changed to a gel state from 0.1 to 0.8 degrees of cure, and from a gel state to a solid state after 0.8 degrees of cure, with a large relaxation process in the initial gel state. It was found that there is a large relaxation process in the initial gel state. This behavior was investigated using a general purpose FEM code which implements an originally developed cure shrinkage stress calculation and it was predicted that the viscoelastic master curve moves to the high modulus and long time side with a decreasing gradient in the relaxation zone as the material progresses from liquid to gel to solid.

The rate constants of the strain-induced particle growth equation were obtained from low-cycle fatigue tests. A creep constitutive equation incorporating the strain-induced particle growth equation was then implemented in a general-purpose FEM solver, and thermal fatigue analysis of solder joints in BGA semiconductor packages was performed. Thermal diffusion and strain-induced particle growth reduced the creep strength and increased inelastic strain energy range ΔW_in, the driving force for thermal fatigue failure. However, under the thermal fatigue conditions of this study, the contribution of strain-induced growth to particle growth was small, and thermal diffusion particle growth dominated the increase in ΔW_in of BGA solder joints during the ongoing thermal fatigue cycle.

The dominant factor governing the thermal fatigue driving force ΔW_in square BGA semiconductor package solder joints is the interaction between PCB thickness and Si die size, for which a response surface equation was derived. For structures with thicker PCBs and larger Si, the single factors of mold thickness and package substrate thickness were highly significant among other structural factors. The response surface also depended on the temperature range and the high temperature hold time. In the case of the rectangular package, the PCB thickness - Si die size interaction was also dominant, and its response surface had the same shape as for the square package and ΔW_in depended on the package aspect ratio. By reflecting the high temperature hold time, temperature range and single structure factor, and package aspect ratio in the response surface equation of the square shape of PCB thickness - Si die size, the ΔW_in of the BGA semiconductor package solder joint can be easily predicted.

In the field of aerodynamic design for turbomachinery, automatic optimization techniques have been widely used to maximize performance. However, there are many challenges to achieving rapid development because these techniques often require extensive computational time. To address this issue, this study introduced a novel optimization algorithm known as Factorization Machines with Quantum Annealing (FMQA) for aerodynamic design using fluid analysis. The implementation of FMQA enabled the development of turbomachinery configurations with enhanced aerodynamic performance in a shorter time compared to conventional optimization algorithms.

Partial differential equations (PDEs) play an important role in engineering since various physical phenomena can be well described by PDEs. The key problems in solving PDEs are memory usage and computational time when the size of the system of interest grows. Given the potential of quantum computing, we propose a quantum algorithm for solving PDEs based on the Hamiltonian simulation, which is well known as the possible application of quantum computing for physical simulation. First, we review the method of linear combination of Hamiltonian simulation and propose its implementation via a tensor network technique. Second, we discuss how to deal with the spatially varying physical constants via a logic minimizer technique. Finally, we provide numerical experiments to demonstrate our proposed method.

What impact might AI have on society if it surpasses human intelligence in the future? Recently, there has been growing discussion about the potential for AI to accelerate research and development in science and technology, suggesting that science itself is becoming a target for automation. Computational mechanics and Computer-Aided Engineering (CAE) are fields that have long embraced automation, and their insights could be widely applicable to other disciplines. Moreover, the integration of quantum computing into CAE holds the promise of further enhancing efficiency in product development and contributing to the broader automation of science. This paper explores the current state and future prospects of the automation of science and the role of quantum computing in CAE. We will discuss how these advancements might influence not only industrial processes but also scientific discovery, potentially leading to breakthroughs that were previously unimaginable. The knowledge gained from quantum computing in this domain could have far-reaching implications across various fields, offering new pathways to accelerate innovation and tackle complex challenges.

This study applies an efficient random encoding method whose uncertainty are controlled through Gaussian distribution to two-dimensional truss structure analysis based on quantum annealing (QA). As the basic operation model of QA, Quadratic Unconstrained Binary Optimization (QUBO) model adopts binary qubits as variables, which means that generally an additional encoding method should be adopted to extend their expression range to any real numbers. The encoding method applied in this study selects a certain number of digits within the range of real numbers, and then controls whether they work in the total summation of real number expression through the 0/1 quantum state. A Gaussian distribution are adopted in it to control the selection of digits to represent real numbers, ensuring its stability. Other three commonly used encoding methods (binary, one-hot and on-off encoding) are employed as comparison under different truss structures and load conditions. The results show that the Gaussian on-off encoding outperforms present encoding methods in structure analysis in terms of accuracy and stability.

We have developed origami hats that realize stylish design. It can be folded to make it easy to carry, but without compromising the comfort of wearing. In addition, we will develop safety helmets by adding energy-absorbing structures to these hats, and also will develop bicycle helmets. As energy-absorbing structures, a honeycomb structure can be considered, but we show that the space-filling structure that we have developed independently gives ideal characteristics. We will examine the energy absorber by not only experiment but also by simulation, for practical using.

The purpose of this research is to develop a sound-reducing shade that is lightweight, foldable, and reduces the transmission of sound to the outside with relatively cheap materials. Sound pressure reduction in the high frequency range can be achieved by using sound-absorbing material in some areas, and tubular structures are used for the shade's walls, with the aim of using an origami structure to deal mainly with the frequency range below around 500 Hz. The effect of reducing sound pressure was confirmed by FEM simulation, and the reasons for this were considered by analyzing the sound pressure levels in fine details.

A two-dimensional fan painting may be a perfect circle, but if a bone is inserted into the painting and it is made into a fan, it will become an oval; the fan will be distorted. The way in which it distorts also depends on the length of the bone. By utilizing this distortion, it is possible to express a story with the fan, for example by changing the emotions of the characters or the behavior of the characters depending on the viewing angle. However, as mentioned above, each side distorts differently, making composition difficult. The authors have clarified the mathematics behind how the distortion occurs, and show that it is now possible to recreate the original charm of the fan, such as a collaboration between the fan and tanka, where the upper verse and the scenery can be seen when viewed from the left, and the lower verse and the scenery when viewed from the right.

We have developed FQHNN (Fuzzy Quantification Theory Embedded Holographic Neural Network) which has been already applied to various problems successfully because of its causality and versatility. It is confirmed the excellence of the FQHNN in the time series against LSTM (Long Short- Term Memory). We try to develop concentration confirmation system with facial expression analysis based on FQHNN in the time series. At last we discuss whether the system can be applied to judge the concentration of the driver of auto-driving car in real time

In finite element analysis (FEA) that takes into account nonlinearities such as large deformation and elastic-plasticity, a solution is often not obtained depending on the input parameters. Therefore, we proposed the parallel search method for analyzable input parameters in nonlinear FEA using Bayesian optimization. However, an issue remained that the search would be adversely affected as the number of parallel processes increased. Therefore, in this study, we proposed a parallel search method that reflects the progress of FEA being executed in each processors. The proposed method was applied to a simple rectangular body torsion analysis and achieved a 24% computational time reduction compared to conventional Bayesian optimization in 120 parallel executions.

POD (proper orthogonal decomposition) is a method for obtaining a basis for representing physical phenomena and has been used in reduced order models for fast analysis. To apply POD to large-scale problems, distributed memory parallel computing, in which parallel processes are allocated to each partitioned subdomain, is effective. On the other hand, Local POD, which obtains a basis for each decomposed subdomain, has been proposed to improve the computational efficiency of POD. Against this background, we propose a method to independently set subdomains where the basis is acquired and subdomains where parallel processes are allocated. In this study, the selection of the appropriate number of bases for each subdomain and the extension to load balancing will be implemented.

S-version Finite Element Method (SFEM) can superimpose several different meshes which is expected to reduce the computational cost and mesh generation cost compared to the standard FEM. However, as the number of mesh divisions becomes larger, the computational time becomes huge even using SFEM. In particular, the time required to solve simultaneous linear equations is enormous. In this study, we employ the iterative method as the solver for simultaneous linear equations, assuming the use of parallel computing, and aim to reduce computation time by using the preconditioning of iterative method. We apply deflated conjugate gradient method with eigen-modes to SFEM and evaluate the convergence of the number of iterations and computational time.

In the field of structural analysis using the finite element method (FEM), the demand for detailed analysis is increasing, and techniques for large scale analysis have become an important factor. Distributed memory parallel computing is the key method that mitigates the effects of memory usage by distributing tasks and memories to multiple computation nodes. However, the performance of computers is limited, and the degree of freedom that can be handled depends on each environment. Thus, it is important to develop the method that can save the memory usage. Parametric studies are used in manufacturing to optimize designs, but it is impractical to cover every parameter combination. Therefore, projection based reduced order models (ROM) come into focus. However, the cost of projection operations is the bottleneck for ROM analysis. To solve this issue, empirical cubature method, which is one of the hyper-reduction methods, save the memory usage by the efficient integration performed by sparse element-wise weighting. This element-wise weights are obtained by sparse non-negative least squares method (sparse NNLS). In this study, we suggest the sparse NNLS algorithm corresponding the distributed memory parallel computing and evaluate the impact on the memory usage, the computation time, and the accuracy.

Physics-informed neural network (PINN) is a machine learning method for approximating the initial-boundary value problems of partial differential equations. This research focuses on parallel computation of the PINN using the iterative domain decomposition method (DDM). The DDM is known as a parallel numerical method for the finite element method (FEM) and an iterative DDM solves an interface system, which is introduced by a static condensation, using iterative methods. This research applies the iterative DDM to the PINN for the two-dimensional Poisson’s equation (differential form) and evaluates performances between FEM and PINN.

In the dual Lagrange method used in a mortar finite element method, a block matrix with respect to the slave DOFs of the constraint matrix becomes a diagonal matrix due to the biorthogonal shape functions. This property makes the elimination of the slave DOFs easily, and the reduced linear system can be solved by the conjugate gradient method. In the present study, this approach is compared with the conjugate projected gradient method and the BDD-MPC method.

Smoothed Aggregation Algebraic Multigrid (SA-AMG) methods are known as fast methods for solving systems of simultaneous equations. When applying SA-AMG methods to systems of equations in structural analysis, it is known to converge in fewer iterations by using rigid body motion modes called near-kernel vectors. On the other hand, by transforming unknowns for equations in structural analysis, constraints such as mesh connectivity and rigid body interfaces can be imposed. In this study, it is demonstrated that by transforming near-kernel vectors in structural analysis with constraints from unknown transformations, convergence can be achieved in the same number of iterations as when no constraints are present.

The use of GPUs in computational science is growing, and some libraries for solving sparse matrix linear equations by direct methods are capable of parallel execution using multiple GPUs. These libraries achieve distributed parallel execution by internally reordering matrices to construct independent patterns. On the other hand, in structural analysis, linear equations are constructed in each domain based on mesh decomposition. Since there is overlap between the regions, the solution cannot be obtained when using the direct method without not only solving each equation but also handling the dependencies. Since there is no library that solves domain decomposition based linear equations in parallel on GPUs using the direct method, an implementation using the CUDA library, for example, can be incorporated into a structural analysis application. The AmgX library provides a solver and preprocessing using the AMG method, which includes a direct solver on coarse grids. Since this direct solver has limited applicability depending on the block size of the matrix, extensions for structural analysis applications were considered.

In this research, the objective is to utilize AI-generated motion pictograms to enhance the communication of information. Although still pictograms are widely used, they have limitations in conveying dynamic information and complex concepts. On the other hand, morsi-static pictograms can overcome these limitations by providing a more dynamic visual representation. The use of AI to automatically generate motion pictograms from static pictograms is considered; the creation process can be made more efficient and less costly by utilizing AI. This will allow motion pictograms to be used in a wider range of applications. However, the quality and reliability of AI-generated motion pictograms remain an issue, especially in emergency and disaster situations where the quality of the pictograms is critical to safety. In order to overcome this issue, experiments were conducted to compare the effectiveness of AI-generated motion pictograms and human- created motion pictograms. The results suggest that while AI-generated motion pictograms can improve comprehension in certain cases, there is still room for improvement in terms of overall quality. Future research should focus on developing processes to mitigate potential misunderstandings caused by motion representations; addressing the challenges associated with AI-generated motion pictograms can create more effective and usable visual communication tools.

There is a lot of interest in the technology that uses various types of semiconductor packages, from BGA and QFP to power modules, as compact thermal models for simulation. There is also a growing demand for compact thermal models for general electronic devices, not just semiconductors. The technology of compact thermal models allows the internal structure and material properties of devices to be treated as a black box, and so they are increasingly being used in simulations of electrical devices. In recent years, there has also been a high demand for transient simulations, particularly from manufacturers of transport equipment such as automobiles and manufacturers of communication equipment such as mobile phones, and conventional compact thermal models are no longer sufficient to meet these needs. Transient simulations generally cause a dramatic increase in calculation time, so there is also a great deal of expectation for compact thermal models in terms of reducing calculation time. This paper shows that by defining appropriate boundary conditions on the surface of reduced order model of the type embedded in CFD software, one of compact thermal models, the temperature history of the chip can be reproduced for fine power changes such as pulses.

Thermal CVD is a promising process for epitaxial growth of silicon carbide (SiC), which is widely used in various types of electric devices these days. For improving CVD processes, CFD have a key role because to understand the kinetics, flow and heat is significantly important. In the previous report, we examined the applicability of Simcenter STAR-CCM+ for a 2D SiC-CVD reactor model and performed a simplified design exploration with Simcenter HEEDS and Simcenter ROM. In this paper, we expand the CFD model to a full 3D geometry and compare the difference between 2D and 3D results. Furthermore, a simplified numerical experiment is conducted to evaluate the effect of wafer rotation by applying velocity on the surface.

Plasma simulation has become an indispensable tool to elucidate the plasma behavior in semiconductor manufacturing processes, but in industry it is important to obtain results that can serve as a guideline for product development within a realistic time frame. In this study, we present a simulation study of Ar/N₂ low pressure inductively coupled plasma using a fluid model that is possessed of a relatively small computational load. Based on the solution of the fluid model, we construct a surrogate model with a very light calculation load. The effectiveness of the surrogate model is verified by comparing the electron density, electron temperature, potential, and density distribution of ions and chemical species predicted by deep learning method with the calculation results from fluid model.

The objective of the study was to investigate the possibility that vacancy-oxygen complexes (VO_X), which are presumed to be generated in the rapid thermal process (RTP) wafers, exhibit a gettering effect on serious metal impurities (Fe, Ni, and Cu) in semiconductor manufacturing process. The binding energy (E_b) was obtained using first-principles calculation. The calculated E_b of metastable VO₄ and metals are 1.14 eV for Fe, 1.00 eV for Ni, and 1.47 eV for Cu, which are larger than the E_b of Fe-B (0.65 eV).

Si−Ge alloys have been studied in various purpose over the past several years. However, there are few studies focused on atomic arrangement of Si−Ge alloys. It is said that Si atoms and Ge atoms in Si−Ge alloys aggregate and form Si−Si and Ge−Ge bonds, but specific atomic arrangement is unknown. In this study, we evaluated formation energies and expected values of coordination number in many different structures by using density-functional-theory (DFT) first-principles calculations and genetic algorithm (GA), and analyze the trend of atomic arrangement of Si−Ge alloys. From the evaluation of coordination number according to the Boltzmann distribution, we found that structures which have Si−Si and Ge−Ge bonds are more energetically stable than the zincblende-type structure which only has Si−Ge bonds.

Regarding theoretical studies of Ge−Sn alloys which are expected to be applied to optoelectronic devices, there are few previous studies directly dealing with the crystal structures of high Sn concentration because they have an enormous number of possible atomic configurations. And thus, there are few academic knowledge on the Ge−Sn alloys with high Sn concentration. In this study, we linked a genetic algorithm with a machine learning model that predicts the total energy evaluated by first-principles calculations to search for stable structures of the Ge−Sn alloys with various Ge/Sn concentrations (0.000 < x < 1.000). We identified optimal crystal structures of the Ge−Sn alloys that are more energetically stable than the other Ge−Sn compounds registered in the existing crystal structure database.

In this paper, hypersonic flow around flight test model is simulated using scFLOW, which is a comprehensive computational fluid dynamics (CFD) simulation package designed to handle complex geometries with unstructured grids consisting of arbitrary polyhedral cells. The heat flux is simulated and evaluated using laminar and several Reynolds Averaged Navier-Stokes (RANS) models. The validity of this comprehensive simulation package is shown in comparison with previous experimental data.

With the final goal of investigating the flow within a landfill containing waste and recycled materials, in this study a landfill layer was prepared using gravels as a simulated sample, and the rainwater infiltration behavior was evaluated through calculations. The simulated landfill layer packed with gravel was subjected to X-ray CT analysis to obtain precise 3D pore structure as digital data, from which a central cross section was extracted and the vertical 2D unsaturated flow was solved using two-phase flow analysis with phase field method. This series of evaluation procedures made it possible to investigate the infiltration behavior of large-porous media using computational engineering, resulting in the estimation of the drainage capacity and the water channel flows in such media.

In this study, POD (Proper orthogonal decomposition) using VFBasis (VINAS Co., Ltd.) is applied to the computed flow field in an idealized urban canopy model with many identical blocks placed uniformly to analyze the turbulent structure formed inside and outside the urban canopy. Assuming the universality of the spatial structure, the POD analysis was performed by superimposing data from domains of identical shape at different locations in the group of blocks to highlight the structure near the blocks. The turbulent structure with circulating flow near the blocks was extracted as POD modes.

Oscillating flow near the first branch of the human bronchus was experimentally and numerically investigated. The investigation was conducted with three different Dean numbers. Dean number was varied by changing curvature of pharynx. Resin bronchial model was employed for the experiments. Velocity distribution was measured by PIV. 3D unsteady numerical simulation was conducted under the same conditions as in the experiment. The velocity and pressure distributions were investigated by the simulation. In the inspiration flow, the results showed the velocity distribution around the first branch is becoming more uniform as the Dean number increases. It is speculated that the incoming flow to the bronchial bifurcation was accompanied by the Dean vortices. These Dean vortices may make velocity distribution uniform. It is considered that the more Dean number increases, the more the velocity distribution becomes uniform.

Airbags are restraint devices essential for automobile crash safety, and improving the accuracy of predicting their deployment behavior is an important issue. Comparisons between simulations and experiments are generally made by comparing signals obtained from attached sensors or by comparing videos, but accurate shape comparisons are difficult for airbags that deploy at high speed while changing their shape in a complex manner. In this report, we propose a method that combines a technology to create a three-dimensional shape from images obtained by multiple high-speed cameras and a technology to compare complex shapes in a dimensionally reduced space in order to solve the above problem. The results showed the possibility of quantitatively assessing the comparison of test and analysis results.

In recent years, to improve fuel efficiency, electric efficiency, and driving range of automobiles, a reduction in aerodynamic drag is being sought in tires for automobiles. Despite the fact that tires with a tread pattern undergo load deformation while rotating on a moving vehicle, there are few instances of aerodynamic simulations that take this factor into consideration. In this study, we focus on the tire rotation method in PowerFLOW which treats the treaded tire surface as an immersed boundary. Therefore, volumetric force is applied to the interface between the tire and air because the computational grid is not match tire surface. The treaded tire rotation and deformation are expressed by the morphing method in the PowerFLOW’s function without fluid structure interaction calculation. To evaluate the accuracy of this method, wind tunnel tests and simulations for different types of tires with an automobile were conducted. They were good agreement delta CD and wake total pressure distribution tendency.

It is important to understand the mechanical behavior of the surrounding tissues when prosthetic devices such as dentures are placed in order to predict the onset of pain and the prognosis of the disease. 3D dental scanners and CAD are being considered, but creating an anatomical model is still difficult due to the uniqueness and complexity of each patient's oral morphology. Therefore, there are limited examples of numerical analyses that agree with clinical findings. In previous studies, the authors have proposed a modeling procedure to create finite element models using 3D dental scan images. In this study, we apply the proposed procedure to evaluate the behavior in basal seat mucosa under mandibular complete denture.

With the rapid advancement and widespread adoption of Large Language Models (LLMs), there is an increasing anticipation for automating the construction of models and the interpretation of results within the realm of Physics-based simulations. In this study, we propose a framework that integrates LLMs with commercially available general-purpose CAE software to automatically perform various tasks, including the investigation and selection of computational methods, model development, and the summarization and discussion of results, while also identifying key issues that need to be addressed.

CAE democratization has made advanced technology available to many people. In addition to this situation, the evolution of AI has been remarkable, and in 2022, a new Generative AI, “ChatGPT,” will be born. In this paper, we have tried structural analysis by prompt engineering to explore the feasibility of CAE using generative AI. From the trial results, we confirmed that CoT Prompting is unable to perform structural analysis like CAE. However, we felt that its feasibility depends on the development of generative AI. On the other hand, it was found that an ordinary user of ChatGPT, who does not have knowledge of computational mechanics or CAE computational skills, can learn and perform Prompting engineering knowledge and skills to obtain hallucination, that falsifies the CAE structural analysis results. This result is a serious problem, and it was found that there is an immediate need to ensure CAE governance that guarantees quality assurance of analysis results.

A modified formulation of the edge center-based strain smoothing element with selective reduced integration using 4-node tetrahedral meshes (EC-SSE-SRI-T4) is proposed. The modified EC-SSE-SRI-T4 introduces Node-to-Node (N2N) strain smoothing for volumetric components so that the pressure checkerboarding in nearly incompressible solids is suppressed somewhat strongly with relatively lower computational cost. The demonstrative analysis shows that the modified EC-SSE-SRI-T4 gives better pressure distribution than the conventional one, but takes about twice longer computational time.

The development of tetrahedral elements that are automatically generated and free from shear and volume locking is highly desirable. This study aimed to develop a high-precision deformation and stress analysis element using the Finite Element Method (FEM) for soft materials such as rubber and gels. The proposed method features a large deformation analysis by combining strain smoothing elements (SSE) with hybrid elements. To verify the effectiveness of the method, it is applied to several benchmark problems. The results confirm that the proposed method avoids volume locking and enables high-precision deformation analysis. Additionally, it was observed that the method reduces stress disturbances in nearly incompressible materials and large deformation analysis, which are challenging issues in the S-FEM.

The resource saving of tires is an important attempt to supply sustainable products while minimizing environmental impact. In this context, the durability improvement of rubbers will be able to contribute the resource saving of the tires. In consideration of the developments of new rubbers with high durability via materials informatics, fatigue crack growth (FCG) tests of eleven styrene-butadiene rubbers (SBRs) were conducted by using a pure-shear specimen at a test frequency of 5 Hz and a strain ratio of 0.1 in air at room temperature. Unfilled, carbon black (CB)-filled, or silica (SC)-filled vulcanizates were prepared by varying filler contents ranging from 5 to 25 vol.%. Although the FCG data of all the vulcanizates could be fitted by power-law equations, the data of the vulcanizates with the filler contents of 15 vol.% or higher were deviated from the power-law equations at higher tearing energy ranges. In both the CB- and SC-filled vulcanizates, although the threshold tearing energy ranges increased and the power-law exponents decreased with an increase in the contents, its values became constant with the filler contents of 15 ~ 20 vol.% or higher. Regarding those vulcanizates, the SC-filled vulcanizates had approximately three times larger threshold tearing energy ranges than the CB-filled ones.

A novel method for selecting filled styrene-butadiene rubbers (SBRs) with high durability based on fatigue crack-growth (FCG) properties was investigated. Pure-shear specimens fabricated from eleven SBRs filled with carbon black (CB) and silica (SC) with varying filler contents were prepared and then, their FCG tests were conducted in air at room temperature. Fatigue lives for simple-tension specimens subjected to cyclic loading were predicted from the experimental FCG-test results and rubber materials with high durability were selected based on the predicted fatigue curves. Considering both strain- and stress-controlled conditions, the predicted fatigue curves were constructed with strain-energy density ranges. According to those curves, SBRs filled with the SC content of 20% and the CB content of 15% were selected as rubber materials with higher durability.

This paper deals with a parallel finite element analysis by Hierarchical Domain Decomposition Method (HDDM) of an electromagnetic field problem. New preconditioners for the HDDM are studied to converge stably, and to reduce numbers of iterations and computational times.

It is well known that the full-wave electromagnetic field analysis becomes difficult to be solved with an iterative algorithm as the size of the problem becomes larger. The domain decomposition method is one of the effective approaches for improving the convergence of an iterative solver and many studies have been reported. In many cases, however, those studies are based on the unstructured mesh models and the ones which consist of the orthogonal grids are few in number. The orthogonal grid models can be constructed systematically for the complex target object and partitioned into sub-domains flexibly, which is suitable characteristics for the domain decomposition method. In this paper we report the convergence characteristics of the domain decomposition method in frequency domain applied to the full-wave electromagnetic analysis problems.

Falls are a common problem among the elderly, often leading to severe physical injuries. It becomes more crucial to predict and prevent falls among the elderly. For this purpose, we aim to extract potential abnormal gait patterns and pre-fall indicators from multimodal data, which includes kinematic data, and gait data extracted from Inertial Measurement Units (IMUs). In this study, we first independently assess 31 elder people's Gait Risk Level using the Timed Up and Go test. Then we collected their gait data during Normal and Fast walking by IMUs. By calculating the gait trajectories from the IMU data and extracting the Average Stride Length and Minimum Foot Clearance (MFC), we used a classification model to classify the high, medium, and low risk based on Stride Length and MFC for both Normal and Fast-walking groups. The classification accuracy was 90% in the Normal speed group, increasing to 95% in the Fast-walking group. The result shows that the differences between the risk groups become more pronounced as walking speed increases. Particularly, there were significant differences in average stride length between the low risk group and the medium/high risk groups during both Normal and Fast walking group. The average values for both feet in the low risk group were closer and significantly greater than those in the medium and high risk groups. This result validates the effectiveness of our techniques and methods, providing a robust scientific basis for early diagnosis and prevention of fall risk in the elderly.

In this study, a finite element computational method for higher-order strain gradient viscoplasticity theory, which takes spatial gradients of the equivalent plastic strain as additional nodal degrees of freedom, is proposed utilizing an implicit time integration scheme. It is demonstrated that the present method much improves calculation stability compared to previously proposed methods. The uniqueness of the solutions of the theory is confirmed by solving strain localization problem assuming a material exhibiting strain softening and a material modeled by a pseudo-corner theory that assumes strain hardening but exhibits reduction of stiffness. The present study demonstrates effectiveness of the computational higher-order strain gradient viscoplasticity for solving problems involving plastic flow localization.

In this study, the effect of the microstructure of rolled metal on its mechanical properties was investigated with crystal plasticity finite element simulation when the geometrically necessary (GN) dislocation density was considered with the preferred crystal orientation. First, the crystal orientation and the GN dislocation density of the rolled metal were quantitatively estimated from experimental data of electron backscatter diffraction (EBSD) in this study. Crystal plasticity finite element tensile simulations were then performed to reproduce the mechanical anisotropy with the models reflecting (i) randomly selected crystal orientation and no GN dislocation density distribution, (ii) preferred crystal orientation and no GN dislocation density, and (iii) preferred crystal orientation and the GN dislocation density distribution obtained from EBSD data. Results of simulations (i) and (ii) demonstrated that the mechanical anisotropy could not be reproduced solely based on the preferred crystal orientation, indicating that the dislocation density distribution had a significant impact. The effectiveness of the simulation method (iii) is validated by comparing the results with experimental data, confirming that incorporating GN dislocation density is crucial for accurately simulating the mechanical anisotropy of rolled metals.

The formation of kinks and their associated strengthening have garnered significant attention in recent years as a key candidate for a new strengthening mechanism. We have successfully established a model based on the Field Theory of Multiscale Plasticity (FTMP) that can represent the entire process from kink deformation to strengthening in a unified manner. However, a critical gap remained—the inability to accurately capture the initial yield stress increase, typically observed in 45° double compression tests. This study addresses this issue by re-examining the information to be transferred from the pre-compression analysis. It is demonstrated that simply incorporating the drag stress history can resolve this problem, rendering the present approach highly versatile and surpassing other FEM-based methods.

A trade-off relationship exists between the strength and ductility of general metallic materials. One way to achieve high strength and high ductility is to utilize the twinning-induced plasticity (TWIP) effect. The TWIP effect is activated by introducing the twinning deformation in addition to the slip deformation as plastic deformation mode. However, the relationship between the twinning deformation behavior and the TWIP effect has not yet been clarified. Conventional methods simplify the twinning deformation, making it difficult to reproduce the details of the plastic deformation behavior. In this research, we focus on the partial dislocation, crystal defects responsible for the twinning deformation, to represent the twinning deformation’s local occurrence and the twin phase’s lamellar shape. Based on this model, finite element analysis is performed to investigate the effect of the twinning deformation on the mechanical properties. The local shear deformation due to twinning deformation makes the stress field around the twin lamellar and invokes the slip deformation to relax it. This additional slip deformation leads to the dislocation accumulation around the twin lamellar and will contribute to the strengthening.

TWIP (twinning-induced plasticity) steels have attracted attention as a useful structural material that retains sufficient ductility even when strengthened by grain refinement because of the effect of deformation twinning that occurs inside the material. However, the strength of TWIP steels has conventionally been calculated by artificially arranging twin nuclei. In this study, the nucleation conditions of twin proposed by Shimokawa based on the MD method are applied to a polycrystal of TWIP steel. Then, the phase-field and dislocation-based crystal plasticity FEM analyses are performed to investigate the effects of nucleation and growth of twin on the strength and ductility of TWIP steel.

To express effects of intrinsic material length scale, Aifantis (1984) proposed introduction of plastic strain gradient terms to the conventional yield function, and this class of theories has been studied for the past 40 years as strain gradient plasticity theories. The addition of a spatial gradients of plastic strain to the conventional yield function requires extra boundary conditions in principle, but there are several studies that have introduced plastic strain gradient effects with no consideration of the extra boundary conditions. It is necessary to investigate the validity of the solutions without extra boundary conditions. In the present study, a shear band problem is treated using finite element method, and convergence solutions with and without extra boundary conditions were compared. As a result, both solutions with and without consideration of the extra boundary conditions gave almost the same numerical results in the specific problem considered.

In this study, FTMP (Field Theory of Multiscale plasticity) was incorporated into a CP-FEM simulator, and simple tensile analyses were conducted in the [001] direction on a single crystal FCC metal. As a result, isotropic structures resembling dislocation cells were observed. To better represent these cell structures, a recovery model based on FTMP, which assumes a pair annihilation process via incompatibility rate, was introduced. This analysis revealed the presence of traveling waves in the incompatibility distribution normal to the slip plane. These waves were further enhanced by the superposition of fluctuating strain energy rates driven by the incompatibility field, whose interactions ultimately produced equiaxed patterns resembling cell structures.