In this paper, nonlinear boundary value problems are analyzed by using the over-range collocation method (ORCM). By introducing some collocation points, which are located at outside of domain of the analyzed body, unsatisfactory issue of the positivity conditions of boundary points in collocation methods can be avoided. Quite accurate numerical results of the nonlinear partial differential equations have been obtained. Because the ORCM does not demand any specific type of partial differential equations, it shows promise of wide engineering applications of the ORCM.

View full abstract

We present some improvement methods for Smoothed Particle Hydrodynamics (SPH) method to obtain accurate solutions for surface wave propagation problems. To verify the accuracy of this method, a water tank test of wave propagation on an artificial reef is performed. As a result, simulated wave height histories agreed well with measured ones even in the case of wave breaking situations in which deformation of wave takes place strongly.

View full abstract

This paper discusses a multiscale stochastic stress analysis of a heterogeneous material via the stochastic homogenization analysis. In particular, a nonuniform microscopic random variation is taken into account in this paper. The influence of the nonuniformity on the random variation of the microscopic stresses is investigated. Monte-Carlo simulation or the first order perturbation-based stochastic homogenization method is employed for the analysis. The Coefficient of Variance (CV) of the microscopic stress is analyzed for several cases of nonuniform microscopic random variations. With the numerical results, influence of the nonuniform microscopic random variation on the CVs of the microscopic stresses is discussed. Also, applicability of the perturbation method to the analysis is investigated in comparison with the Monte-Carlo simulation

View full abstract

In this paper, we considered a polymer-based piezoelectric stack actuator based on MEMS nano-positioning technology. A no-wiring structure was used for simplicity of fabrication and to avoid irregularity deformation under wiring. We selected the key parameters for actuation design, and a simulation was executed through FEM analysis of the electric field and deformation. From this simulation, we decided upon the appropriate parameters of the actuator.

View full abstract

A thermophotovoltaic (TPV) system is able to convert directly thermal energy, generated by a high temperature heat source, into electricity through thermophotovoltaic cells. Although the energy flux has three steps, designing a TPV system with high efficiency is a challenging task. This particular device has been studied for house heating applications in order to reach better performances and higher efficiency values, compared to traditional boilers. The main issue is to achieve high and uniform temperature values on the emitter surface. In the first step of this project a novel swirl gas burner is being developed and optimized in order to fulfill these objectives. Experimental tests have been performed on a first prototype considering different values of input power, thus fuel flow rate and air mass flow rate, changing some geometrical characteristics of the burner. Collected results have then be used to create response surface functions, to be used in a multi-objective optimization considering efficiency, maximum and mean temperature of the emitter.

View full abstract

A data exploration method based on POD was investigated to get clues to improve actual-product designs from a small set of rough-design candidates. The current method is based on the Oyama method which was used to acquire design knowledge of Pareto-optimal solutions obtained by evolutionary approach with large set of simulation models. The expanded method adopted (1) 2 steps POD analysis, and (2) voxel representation for different topology models. These approaches enabled to evaluate rough-design candidates which had non-parametric structures and those were difficult to handle in the original method. Especially, for the case of handling rough-design candidates, there was no explicit rule for model number ordering at the “first” POD analysis. We found that reordering the model numbers based on the first POD result was important, and rich information was obtained from the expansion coefficient vectors that made it possible to classify the feature regions and to detect feature regions with complicated flow structures at the "second" POD analysis. The effectiveness of the expanded method was verified in a cooling design problem for an optical disc drive. We found that the expanded POD-based data exploration method is a promising method for detecting key factors for actual product designs.

View full abstract

Design informatics, which is the efficient design methodology, has three points of view. The first is the efficient exploration in design space using evolutionary-based optimization methods. The second is the structuring and visualizing of design space using data mining techniques. The third is the application to practical problems. In the present study, the influence of the difference among the seven pure and hybrid optimization methods for design information has been investigated in order to explain the selection manner of optimization methods for data mining. The practical problem of a single-stage hybrid rocket is picked up as the present design object. A functional analysis of variance and a self-organizing map are employed as data mining techniques in order to acquire the global information in design space. As a result, mining result depends on not the number of generation (i.e. convergence) but the optimization methods (i.e. diversity). Consequently, the optimization method with diversity performance is the beneficial selection in order to obtain the global design information in design space.

View full abstract

Grammatical Evolution (GE), which is one of evolutionary computations, can find the function or the executable program or program fragment that will achieve a good fitness value for the given objective function to be minimized. This paper describes the use of the Stochastic Schemata Exploiter (SSE) for improving the convergence property of the original GE. The convergence property of the original GE and the improved GE algorithms is compared in the symbolic regression problem. The results show that the Grammatical Evolution using Stochastic Schemata Exploiter (GE-SSE) has the faster convergence speed than the original GE.

View full abstract

During plastic deformation, effects of microscopic voids on macroscopic mechanical behavior are important, because ductile fracture of metal materials occurs through nucleation, growth and coalescence of voids. Macroscopic anisotropy is also caused by distributions of voids and it depends on strain history, this is because growth behavior of voids is influenced by strain direction. In order to predict deformation or fracture behavior considering these effects of voids, numerical simulation using a constitutive equation which represents the influences of voids by internal state variables is required. In this paper, an anisotropic evolution equation of damage tensor considering three dimensional growth of voids is proposed. The theory is based on anisotropic Gurson's yield function which is established by substituting net stress tensor proposed by Murakami and Ohno into Gurson's yield function. Thermodynamical consideration and internal state variable theory are also used to derive the evolution equation. The matrix material is assumed to be elastic perfectly plastic solids in order to neglect work hardening and focus attention on the growth behavior of the voids. The accuracy of the presented theory is validated by using finite element analysis to unit cell model postulating regularly arrangement of voids. It is compared with numerical integration of the presented constitutive equation under proportional strain history with various strain ratios, and it is shown to be in good agreement with the theory.

View full abstract

Membrane electrode assembly (MEA) serves as a central component in proton exchange membrane fuel cells. Reliability of the MEA is critical to ensure a proper functioning of the fuel cell. The objective of the present study is to develop a numerical model to predict the onset of MEA crack formation. Tensile tests have been conducted at different humidities and temperatures to determine the mechanical properties of MEA. The results of these experiments were combined to a finite element model to establish a failure criterion of MEA, both for direct and fatigue crack. The proposed criterion, based on dissipated energy, was shown to successfully predict the failure of the MEA in various environmental conditions.

View full abstract

Even though there have been many efforts to develop carbon nanotube (CNT)-based electronic devices and sensors, drastic variation of the electronic functions depending on fabrication process has been reported. The authors have also validated the possibility of a highly sensitive strain sensor using various resins in which multi-walled CNTs (MWNTs) were dispersed uniformly, however, the strain sensitivity of the developed sensors fluctuated significantly. Therefore, it is indispensable for clarifying the dominant factors which change the electronic state of a deformed CNT for assuring the stable performance of the devices. In this study, the relationship between the deformation characteristic of a CNT under strain and its electronic properties was analyzed by using molecular dynamics analysis and the first principle calculation based on density functional theory (DFT). Orbital hybridization were found to occur when the local curvature exceeded about 0.3Å^-1, inducing the decrease in the band gap.

View full abstract

In this study, the enriched finite element method (enriched FEM) is applied to evaluate the intensity of singularity for three-material joints with power-logarithmic singularities. Using this method, the intensity of singularities can be directly evaluated and very refined meshes around the singular point are unnecessary. Analyses for eigenvalue and eigenvector are conducted to calculate the order of stress singularities and the asymptotic displacement fields on the enriched elements. Accuracy of the results for different mesh types (4-node and 8-node elements) and different sizes of the enriched region are compared with the value obtained from extrapolation for stress distribution based on FEM with very fine mesh. Finally, the models with various lengths and thicknesses are investigated to study an influence of geometry on the singular stress field.

View full abstract

Computer simulation of electromigration (EM) in microelectronics devices has been reviewed and a multi-physics numerical simulation method has been proposed and developed so that the electric current, temperature, stress can be solved simultaneously and the vacancy concentration can be predicted in a seamless framework. The design considerations for resisting EM is also discussed in this work and a shunt structure for solder joint pad is proposed and its potential for the reduction of EM risk is demonstrated.

View full abstract

There are two major methods, experimental and numerical approaches, for dealing with the strain and stress analysis essential to reliability assessment of electronic packaging. Both approaches are mutually complementary in view point of their assessment capability. In this paper, experimental and numerical simulation results for analyzing some reliability issues of electronic packages were reviewed, and the role of both approaches were discussed.

View full abstract

A method for generating a solution-adaptive mesh is proposed. It uses the adaptive mesh refinement (AMR) method and polyhedral mesh, which, as an unstructured mesh, is better suited for complicated bodies than triangular mesh. Furthermore, polyhedral mesh with hundreds of edges is more effective for generating adaptive mesh. Application of the AMR method to a test problem demonstrated that the adaptive mesh had a resolution about 30 times that of the initial mesh. Application of the proposed mesh generation method to compressible flows with a shock wave showed that using an artificial value for making the division judgment instead of the derivative value is feasible and highly effective.

View full abstract

In this paper, the numerical simulations of incompressible flow with heat transfer are presented by using the seamless immersed boundary method on the Cartesian grid. In the seamless immersed boundary method, the forcing term is added not only on the grid points near the boundary but also on the grid points inside the boundary (solid region) in order to satisfy the velocity boundary condition. Then, the seamless physical quantities, e.g., the pressure, can be obtained, so that the characteristic quantities on the boundary can be estimated precisely. The present seamless immersed boundary method is applied to the energy equation. Then, the temperature satisfying the boundary condition can be easily obtained as well as velocity on the Cartesian grid. The present method is applied to flows around an object with the moving boundary and heat transfer.

View full abstract

In this paper, two moving mesh methods are introduced. One is a Trans-mesh method and the other is a Moving Computational Domain method. In the Trans-mesh method, the bodies can move freely in a main mesh that covers the entire flow field. On the other hand, in the Moving Computational Domain method, the whole of the computational domain including bodies inside moves in the physical space without the limit of region size. These methods are constructed based on the four-dimensional control volume in space-time unified domain such that the method assures to be divergence-free in the space-time unified domain and thus satisfies both the physical and geometrical conservation laws simultaneously. The methods are applied to a falling sphere by gravity in an infinite long straight pipe and an infinite long bending pipe. The results indicate that these methods are promising in simulating the interaction of incompressible fluid-rigid body with collisions.

View full abstract

The computational reconstruction of oil and gas reservoir morphology is a challenging task and must consider several natural phenomena. In this work, a sedimentation model, coupled with the Arbitrary Lagrangian Eulerian (ALE) method for tracking bottom changes, is proposed. The scheme considers an incompressible fluid flow of diluted density currents transported by the advection diffusion equation. The stabilized SUPG/PSPG/LSIC finite element formulation is used for the Navier-Stokes equations and SUPG and discontinuity capturing formulation for the sediment transport. Turbulence is treated by a Smagorinsky model. Results show that the proposed scheme is promising and can effectively represent the phenomena of interest.

View full abstract

A preliminary numerical simulation of the microscopic two-phase fluid motion on a solid surface was conducted using an interface-tracking method based on the phase-field model (PFM). Two variations of the lattice Boltzmann method (LBM) based on fictitious particle kinematics are proposed for solving diffuse-interface advection equations which were revised to improve volume-of-fluid conservation in the PFM simulations. The major findings are as follows: (1) the interface-tracking method accurately predicted the capillary force effect on dynamic two-phase fluid systems with a high density ratio between parallel plates; (2) the initial shape and volume of the two-phase fluid were retained adequately in linear translation with the use of the LBMs. These results proved that the PFM-based method and the LBM-based advection schemes can be used for simulating two-phase fluid motions in various macro- and microfluidics problems for devices, machineries and higher-throughput microdevice fabrication processes.

View full abstract

This paper presents a numerical model for an open axial-flow hydro turbine using moving particle simulation (MPS) method and discusses the torque generation via the model. An Archimedes-screw hydro turbine, which is a type of open axial-flow hydro turbine, has advantages such as ease of installation, durability, and debris tolerance. However, it has been difficult to analyze dynamically due to a complicated interaction between the open channel flow and the turbine. This paper conducted fundamental experiments such as water flow of a dam moved a flat plate and confirmed the validity of the model. The model reproduced well the rotational movement of the Archimedes-screw hydro turbine in an open channel when no-load conditions. Calculated result of rotational speed and torque distribution showed that the turbine gained the rotational torque by mainly the conversion of a water pressure on the blades due to the water level differences. At the upstream end of the screw, torque loss occurred through a wide range of flow conditions.

View full abstract