A thin membrane (thickness = ～10 µm) made of a lanthanum highly-doped cerium oxide (La_0.5Ce_0.5O_2-δ, LDC) was fabricated on a mixed porous layer consisting of Ni and the LDC. Comparing the results of permeation experiments for hydrogen and helium gases confirmed that the asymmetric membrane acted as a mixed protonic and electronic conducting hydrogen separation membrane and that the thin LDC membrane was an almost dense layer. The most striking result of the permeation experiments was that the hydrogen permeation flux increased with increase in the crystal grain boundary length per unit area of the surface and increased in proportion to the square root of hydrogen partial pressure, showing that the flux is controlled by a surface reaction between the adsorbed hydrogen and proton at the crystal grain boundary.

View full abstract

In this paper, AutoMT that is a library for tensor operations in 2-D/3-D spaces is presented. AutoMT stands for “Auto (Automation)”-“M (Matrix)”-“T (Tensor)” and is a kind of DSL (Domain Specific Language). DSL is a programming language specialized for a specific domain, such as graphics, communication, linear algebra, financial, chemical, etc. AutoMT is a library-based DSL/framework which is aimed at enhancing efficiencies in both the computations and the developments of computer programs involving many tensor operations. The paper describes the design, usage, and implementation of AutoMT, first. Then, the results of performance benchmark and an application example in large strain elastic-plastic analysis are presented. AutoMT accelerates the tensor operations that are often involved in the element-wise operations of finite element procedures. Usages of processor registers and SIMD (Single Instruction Multiple Data) mechanism are optimized to enhance the computational efficiencies in tensor operations. The performance of AutoMT is demonstrated by the results of performance benchmark test. It is shown that AutoMT at least doubles the performance. Finally, an example problem of finite strain elastic-plastic analysis is presented along with a small portion of example source program for elastic-plastic analysis. It is seen from the example program that it becomes much simpler than that without using AutoMT. Therefore, AutoMT can accelerate computer program developments. Finally, the problem of diffused necking is presented for a demonstration purpose.

View full abstract

Liquid molecules form ordered structures in the vicinity of the solid-liquid interface. Mass transport in the ordered structure exhibits decidedly different characteristics from the ordinary diffusion property governed by Fick's law. In order to analyze mass transport at this nanoscopic scale, we have monitored the PMF (potential of mean force) obtained from the number density of liquid molecules, which corresponds to the free energy change with respect to the molecular migration. Using this PMF profile, the kinetic process model to describe the molecular mobility inside these ordered structures was developed in this study. In order to verify whether the suggested kinetic model works correctly, molecular dynamics (MD) simulations of the liquid-solid interface system were performed. In a simple system consisting of a LJ liquid and flat solid walls, which are modeled by the continuum media interacting with liquid molecules, molecular adsorption and desorption phenomena were observed. In each process in which molecules migrate to the neighboring adsorption layer, the height of the free energy barrier was calculated using the PMF distribution. It was found that the PMF has a dominant influence on the molecular transport perpendicular to the wall and that the developed kinetic model can accurately predict the molecular mobility over the free energy barrier due to the adsorption and desorption process.

View full abstract

Microscopic stress was calculated with 0.017 mm resolution in a macroscopic model with approximately 100 mm size using the finite element mesh superposition method. To bridge the large gap in resolution, an intermediate finite element model was newly used. This multiscale computational procedure was applied to the biomechanical problem to analyze the microscopic stress in the trabecular bone around acetabular cup implant in total hip arthroplasty, which occurs by direct contact of the implant with trabecular bone. In the microstructural modeling of highly porous media such as the trabecular bone, special attention was paied to the boundary of microstructure model for both homogenization procedure and mesh superposition procedure. Three demonstrative numerical results revealed that higher stress occured at microscale due to macroscopic stress concentration, which is hardly estimated by only bone volume fraction.

View full abstract

Flame propagation over methane hydrate in an axisymmetric geometry with a non-uniform initial surface temperature profile is investigated experimentally. The methane hydrate in a container with 100 mm diameter and 10 mm depth is ignited at the center of the hydrate surface by a pilot flame. The ignition temperature, T_s,_ig, which is measured by a thermocouple 1 mm beneath the hydrate surface at the center of the hydrate, is varied from 173K to 253K. When T_s,_ig, = 173 K, which is lower than the nominal equilibrium temperature of methane hydrate (193 K), the flame propagates axisymmetrically at a low velocity of around 10 mm/s. As the flame propagates, the heat from the flame front dissociates the hydrate. As T_s,_ig exceeds the nominal equilibrium temperature (T_s,_ig = 193 and 213 K), the flame propagates at a high velocity of around 1000 mm/s. The hydrate surface dissociates and a methane-air mixture within the flammability limit is formed before ignition, and the flame propagates in the mixture after ignition. At T_s,_ig = 233 K, the flame propagates fast at around 1000 mm/s until the flame front reaches the radial distance r about 40 mm and the propagation velocity drops to around 10 mm/s and propagates to 50 mm, where it reaches the container wall. The methane hydrate dissociates and methane is ejected into the air up to r = 40 mm. However, in the region between r = 40 and 50 mm, the surface temperature is around 253 K where the hydrate is under the self-preservation conditions. At T_s,_ig of 253 K, the initial temperature is in the self-preservation region, and then the flame propagation velocity is as low as that at T_s,_ig = 173 K.

View full abstract

The stochastic prediction of apparent mechanical property of sprayed porous ZrO₂ film considering the scattering of Young's modulus of ZrO₂ is the purpose of this study. To take the morphology at nanoscale into consideration, FIB (focused ion beam)-SEM image-based modeling technique was employed. Pores were categorized into dot-like, large slit-like, small slit-like and small miscellaneous pores. Their representative dimensions were statistically measured using FIB-SEM and 2D SEM observation in order to generate larger sized microstructure model for homogenization analysis than the observable region by FIB-SEM. The dot-like pores were homogenized first, and hierarchical model was generated. The first-order perturbation based stochastic homogenization method for hierarchical model of porous material was proposed. The stochastic homogenized Young's modulus in the thickness direction, which is the spraying direction, was calculated and discussed.

View full abstract

Microstructural design for improving the strength-toughness balance was studied in low-alloy steel. High-strength medium-carbon steels with an ultrafine elongated grain (UFEG) structure with an average transverse grain size of 0.3 µm and an ultrafine equiaxed grain (UFG) structure with a grain size of 0.7 µm were fabricated by multi-pass caliber rolling at a warm working temperature and subsequent annealing. For comparison, conventionally quenched and tempered steel with a martensitic structure and a 480 MPa-class low-carbon steel with a ferrite (grain size 20 µm)-pearlite structure were also prepared. A quasi-static three-point bending test was conducted at a temperature range from 200°C to -196°C. The steels, except for the UFEG steel, exhibited a typical energy transition curve, in which the fracture energy decreases with decreasing temperature. In the UFEG steel, the fracture energy increased with the occurrence of delaminating cracks as the temperature decreased from 200°C, reached a maximum at ambient temperature, and then decreased. In other words, the steel showed inverse temperature dependence of the toughness. As a result, the strength-toughness balance of the UFEG steel was excellent compared with that of all other steels. For stronger, tougher steel, it is important to design a heterogeneous microstructure rather than a homogeneous microstructure.

View full abstract

Plenty of numerical simulations were performed to analyze the Molten Core-Concrete Interaction (MCCI) phenomena since 1980s. However, uncertainties remain among thermal hydraulic codes. Thus, in order to avoid the effect of uncertainties due to the different empirical formulas and interfacial models, a new CFD code based on Moving Particle Semi-implicit (MPS) method was proposed to simulate the MCCI event in the literature. Validation of the heat transfer and phase-change models is still necessary to utilize MPS method for the MCCI modeling. In this study, a small-scale visualization experiment was conducted using melted hot u-alloy and solid transparent gel wax to acquire quantitative data for validating heat transfer and phase change models of MPS method. Comparison of the simulation and experimental results demonstrates good agreement on ablation behavior and deformation profile of the molten metal moving downwards by the phase change of wax. Additionally temperature evolution in time showed consistency with the thermocouple measurements. Obtained results suggest that MPS code have capacity to simulate the phase change and heat transfer mechanisms of moving molten metal pool which are important mechanisms for the MCCI event.

View full abstract

For free-cutting steels, their mechanical properties, especially local toughness, are functionally degraded. The second phase exists in steel matrix, as a chip breaker, tool edge stabilizer and tool life extender. For these purposes, the second phase plays the role of a solid lubricant. Typical materials as the second phase in free-cutting steels are Pb and MnS. The second phase grains are recognized as stress concentrated points during machining. Such points have the possibility to be new crack initiation sites during fatigue loading. Formerly, the present author found that some Al alloys are degraded by loading combination of fatigue and dynamic tension. On the other hand, for some structural steels, almost no degradation is found by the fatigue-impact loading in tension. However, in 2013 the present author reported that free-cutting steels are deteriorated to some extent by such a loading combination. The deterioration may be caused by the stress concentrated points around the second phase, MnS in SUM24L (JIS G 4804:2008, equivalent to AISI 12L14 steel), which is weakened by pre-fatigue process. In this paper, in order to detect microscopic difference between the specimens fractured in dynamic tension with/without pre-fatigue, geometric features of MnS grains are observed by an optical microscope. In the microscopic observation, typical deformation of surrounding matrix steel is found. Into MnS grain, steel matrix is penetrated. Focusing on the existence of the matrix penetration, the number of MnS grains in free-cutting steel is counted with geometrical features. The experimental results show that the information of the penetrations has a possibility to be an indicator of the dynamic strain rate, comparing with the specimens fractured by quasi-static tension.

View full abstract

The crack growth behaviors of Ni-based superalloys were studied under thermal gradient creep condition (TGC) and compared with isothermal creep condition (ITC). Special attention was paid on getting basic understandings on small crack growth behavior around cooling holes in turbine blades. It was found that the cracks under the TGC showed higher crack growth rate than that under the ITC. Also a significant difference in growth rate was observed between the crack nucleated from cooling hole and the naturally initiated crack under the TGC conditions. These behaviors were analyzed in terms of creep J-integral, J_c; a time-dependent nonlinear fracture mechanics parameter. Some mechanistic background was discussed on the characteristic behaviors of the crack advance under the TGC condition.

View full abstract

We developed a walking assistance apparatus which can be used for neuro-rehabilitation of patients. This apparatus assists only the ankle joints of the equipped person while walking. The apparatus attaches the motor on the waist of the equipped person and the flexible shaft transmits the torque of the motor to the self lock less worm gear on the ankle joint. By using this apparatus, the dorsiflexion and plantarflexion of the ankle joint of the equipped patients could be increased while walking. Therefore, the heel contact could be recovered and stumbling could be prevented. Furthermore, the length of the stride of the equipped patients was increased and the posture was raised according to gait training with the apparatus. We found that this apparatus can improve the gait of the equipped person while walking.

View full abstract

The purpose of this study is to clarify how bubbles reduce the speed of sound in a bubbly liquid in a duct, in which a homogeneous medium can no longer be assumed. Getting inspired by the explanation for the origin of refractive indices in the field of optics, we theoretically examine pressure wave propagation in a square duct filled with a compressible liquid containing only a spherical bubble. Theoretical examination reveals that even a single bubble in a square duct can delay the phase of the input pressure wave, causing an apparent reduction in the speed of sound. Based on this result, we can define the speed of sound in a bubbly liquid under the assumption of a homogeneous medium by considering the phase delay caused by radial oscillations of bubbles aligned in the square duct in the limit of numerous bubbles.

View full abstract