A method of numerical plate testing for composite plates with in-plane periodic heterogeneity is proposed. A thick plate model with transverse shear deformation is employed at macro-scale, while three-dimensional solids are assumed at micro-scale. The homogenization process to evaluate the macroscopic plate stiffnesses is called “numerical plate testing” (NPT) in this study, which is in fact a series of microscopic analysis on a unit cell. The specific functional forms of microscopic displacements are originally presented so that the relationship between the macroscopic resultant stresses/moments and strains/curvatures to be consistent with the microscopic equilibrated state. In order to perform NPT by using general-purpose FEM software, we introduce control nodes to facilitate the multiple-point constraints for in-plane periodicity. Numerical examples are presented to verify that the proposed method of NPT reproduces the plate stiffnesses in classical plate and laminate theories. We also perform a series of NMT (homogenization analysis), macroscopic and localization analyses for an in-plane heterogeneous composite plate to demonstrate the performance of the proposed method.

View full abstract

This paper proposes a topology optimization method for steady state incompressible viscous flow problems, based on the finite volume method incorporating level set boundary expressions. The optimization problem is formulated to minimize the power dissipation under a volume constraint. The optimization algorithm is developed based on this formulation, using the adjoint variable method for the sensitivity analysis. The update scheme for design variables uses a reaction-diffusion equation derived from the concept of the topological derivative. Here, the finite volume method is applied to solve the governing, adjoint, and reaction-diffusion equations because it is more suitable than the finite element method for solving relatively large-scale problems that include higher Reynolds numbers. Several numerical examples are provided to confirm the utility of the proposed method.

View full abstract

This paper proposes a fast computation technique for particle (vortex element)-grid interpolations in the Vortex-in-Cell method (hereafter referred to as the VIC method). This technique is based on the incremental computation algorithm for the Gaussian, expressed as exp(–x²) in the recurrence formula, proposed by Turkowski to generate coefficient tables for the Gaussian blur filter used in image processing. The technique can reduce computational costs by replacing an exponential function call, exp(), with two floating-point multiplications per grid point, thus avoiding some expensive operations. The technique is applied to the vortex element-grid interpolation in the two-dimensional VIC method to simulate a plane mixing layer, and error evaluations and performance measurements are performed to confirm its suitability. The error evaluations show that the errors accumulated in computational directions do not affect the large-scale structures of the flow, although the errors do affect the time evolutions of the flow. The performance measurements show that the average execution time of the vortex element-grid interpolation is reduced to one-third and that the average total execution time is also significantly reduced, thus verifying the technique’s suitability for the VIC method.

View full abstract

The finite element method is commonly used to simulate the behavior of sheet forming processes, in order to realize high precision machining. In the conventional shell elements, the plane stress condition which ignores the transverse normal stress is assumed. Thus, the conventional shell elements are not sufficient to simulate the complex behaviors, such as, the deformation of the sheet and the contact force at the sheet-die interface. In this paper, we present a formulation for considering the thickness change and the stress distribution by the surface traction in the shell element. We introduce a displacement variation along the transverse direction to MITC shell element, which is widely used for avoiding the transverse shear locking. Then, we can evaluate the equilibrium equation for the transverse direction by using the introduced variation. Further, we verify the proposed approach to compare the results of the proposed shell with that of the continuum elements.

View full abstract

This paper describes a novel computational algorithm of a hybrid method composed of the thin-layered-element and the finite-element methods to obtain the vibrational responses of foundations embedded in soil. The modified-displacement method, which can efficiently solve a modified system of linear equations, is incorporated into the conventional hybrid method in order to avoid inverting a free-field flexibility matrix. Since the floating-point operations of the inverting operation, which is needed to acquire the soil stiffness matrix explicitly, dominates the whole calculating work of the conventional method, the proposed method may extremely reduce computational time. It is concluded that the floating-point operation counts of the proposed method is less than that of the conventional method, and the ratio of these two numbers is proportional to the cubes of the ratio of the number of degrees of freedom of embedded foundations and substructures to that of the whole system. The computational accuracy and efficiency of the proposed method are demonstrated through numerical examples.

View full abstract

In this study, based on complex network science, we propose a new method to determine the location of the distributed generators such that the efficiency and the fault tolerance of the power grid can be improved. In the case of using only network-topology data, the numerical experiments show that the arrangement of the distributed generators which increases the fault tolerance produces some location deviation. It is found that this result can be explained by the lowest capacity allocated to a link, and increasing the lowest value of the capacity enables more efficient arrangement of the distributed generators without reducing the fault tolerance. In the case of using the network data which have some properties of a real power grid, we find a new explanation variable to analyze the power loss of the power grid evaluated by an exact formula used in the field of electrical engineering.

View full abstract

As a Lagrangian meshfree method, the MPS (Moving Particle Semi-implicit) method has been shown useful in engineering applications widely. Especially, the MPS method is used for the free surface and multi-phase flow analysis. In this research, the stirred fluid flow with rotating the cam-shaft is calculated by the MPS method. And the experiments are conducted and the snapshots of the oil distribution and the torque are compared with calculation results. As a result, calculation results of the behavior of oil and torque were in good agreement with the experiments in each revolving speed. Uncertainty of the calculation results were estimated by the GCI (Grid Convergence Index) method.

View full abstract

In this paper, a finite-volume method on a moving unstructured computational grid for simulation of incompressible flows is presented and developed. In order to strictly assure both physical and geometric conservation laws, the unstructured moving-grid finite-volume method is constructed based on four-dimensional control volume in which space and time are unified. In the method, the velocity and the pressure are connected through a fractional step approach in four-dimensional control volume. We show the detailed formulation of the method and that the method works effectively for numerical simulations of flows including moving boundary. Furthermore, the method is applied for coupled simulation of fluid and motion dynamics. The motion equation with six degrees of freedom is solved in a coupled manner together with the fluid Navier-Stokes equations. Hukidama, which is a japanese toy that has been around since long ago, is lifted up in a jet stream while swinging. The motion of Hukidama is demonstrated and it is shown that the method works effectively for coupled simulation of fluid and motion dynamics.

View full abstract

This paper describes a performance evaluation of a newly-developed ubiquitous learning support system for isogeometric analysis using Android-based mobile devices. Isogeometric analysis seems harder to understand than the conventional FEA because of its complex basis functions, NURBS, which were originally developed for representing free-form surfaces in CAD systems. The proposed system helps users understand isogeometric analysis and do analyses with it anytime and anywhere. The system, using Android-based mobile devices, consists of three subsystems : a NURBS graph display system, a 3D visualization system and a parallel isogeometric analysis system. In this paper, the proposed system are described and the drawing and computational performances of the latter two subsystems are tested in detail through sample analyses.

View full abstract

The advantages of full-field strain measurement are non-contact and wide area. However, its stability and higher resolution are necessary for a practical purpose in engineering application. Accordingly, we should reduce all of errors from shutter mechanism, mechanical vibration, flicker noise caused by light condition and so on, otherwise it is difficult to ensure the accuracy of strain less than 0.001. In this paper, we propose the definition of the resolution of strain measurement and how to obtain the higher resolution by Integrated Dot Centroid Tracking Method. The resolution is defined by standard deviation of the centroid position and can be described by three parameters, which are number of images, dot diameter and distance between each dot. We measure the standard deviation can be computed by the proposed formula. For the validation, we obtained the five standard deviations using averages of nine dots obtained from 5,000 data and infer the resolution using the interpolated standard deviations. We discuss the accuracy of the proposed formula by simple tensile test and the validity of the resolution.

View full abstract

Extended voxel finite element method (X-VFEM) for the analysis of heterogeneous solids using voxel mesh is developed in this paper. The method is based on the enrichments with the level set function for material interfaces employed in the extended finite element method. To model and simulate very complex heterogeneous materials, we formulate the multiple enrichment for elements including number of material interfaces. We first formulate the X-VFEM and explain the procedure of multiple enrichment by means of level set functions and voxel data for representing complex material interfaces. In section 3, several verification examples are presented. The validity of multiple enrichment is demonstrated in 1-D tensile problem of a bar. The accuracy of stress along material interface is assessed with reference to the solutions by the standard FEM and the voxel FEM in 3-D tensile problem of a 2-phase material. In section 4, a numerical example is presented to demonstrate the effectiveness of the X-VFEM for extremely complex heterogeneous materials.

View full abstract

A wall boundary condition represented by polygons was presented by Harada et al.⁽³¹⁾ based on the moving particle semi-implicit (MPS)⁽¹⁾ method to reduce the memory cost and calculation time for the wall particles. However, the inaccuracy of the wall weight function near a non-planar wall boundary causes the unphysical motion of the fluid. Therefore, this paper proposes an improved wall weight function for non-planar wall boundaries. Hydrostatic and dam break simulations with and without a wedge in a water tank are conducted to demonstrate the improvement.

View full abstract

The accuracy of a two-dimensional Rayleigh wave calculation evaluated using the moving particle simulation (MPS) method was investigated. MPS is a mesh-free particle method well suited to the analysis of nonlinear physical phenomena such as demolition and the breaking of waves. The objective of this analysis was to confirm the free boundary condition set in MPS, which has not been thoroughly validated before. Numerical experiments showed that when particles are distributed uniformly, the phase velocity of the Rayleigh waves are in good agreement with the analytical values. Furthermore, when the particles are slightly misaligned (displaced by about 10% from a regular arrangement), the results remain very accurate (the error ratio of the phase velocity is within ±1%). A linear case was analyzed in this work; however, the findings are easily extendable to nonlinear problems.

View full abstract

The present study proposes topology optimization for microstructure of two-phase composite considering hyperelasticity to minimize the end compliance of the macrostructure based on a multi-scale analysis. In general the structural behavior of macrostructure depends on the geometric properties of the microstructure. In other words, optimizing microstructure is an effectual way to improve the macroscopic structural performance applying a multi-scale analysis. However, it needs unrealistic computational costs when structural optimization with nonlinear structural response based on the conventional micro-macro coupling multi-scale analysis is considered. The present study challenges to make it possible to solve the problem by introducing a so-called decoupling multi-scale analysis assuming hyperelasticity.

View full abstract

The coupled finite element analysis is widely required for research and development in mechanical engineering. As the coupled problem, there is not only doubly coupled problem such as fluid-structure interaction problem and electromagnetic-structure coupled problem but also triply coupled problem and multiscale coupled problem. Furthermore, since several types of coupled analysis method such as monolithic method and partitioned method are presented, the various type of coupled analysis code is required. In this paper, we propose the coupled analysis technique using in-house code and commercial analysis code combining by user subroutine of commercial analysis code and MPI. The problems and solutions are discussed for the system constitution and implementation. As the implementation and analysis example, elasto-plastic contact, electric current and thermal conduction triply coupled analysis and multiscale coupled analysis of resistance spot welding are also described.

View full abstract

A thrown boomerang flies along an arch path and returns to the thrower. The flying motion is a complex motion caused by the gyroscopic precession which is the coupled phenomenon of air-flow and boomerang-motion. In this study, the flying motion of the boomerang was demonstrated using the numerical simulation and we showed influences of the figure of the boomerang on the flying motion of the boomerang. In the numerical simulation, we combined Moving-Grid Finite-Volume Method for fluid-flow solver, the Forward-Euler Finite-Difference Method for body-motion solver and the Quaternion for representation of body-rotation to solve the coupled phenomenon of air-flow and boomerang-motion. The moving-grid finite-volume method is constructed based on four-dimensional control volume in which space and time are unified and assures strictly both physical conservation laws and geometric conservation laws. The motion equation with six degrees of freedom was solved in a coupled manner together with the Navier-Stokes equations. Furthermore, the experiment was performed to validate the computational results and the computational results showed good agreement with the experimental results. In addition, it was shown that the angle of wing’s flap and the angle of incidence have great effect on the path of the boomerang.

View full abstract