The free-molecular-type kinetic scheme is a new lattice Boltzmann model which can simulate supersonic flows described by the compressible Euler and Navier-Stokes equations. We simulate two-dimensional supersonic flows over a wedge and a circular cylinder using this scheme. Numerical results agree well with the corresponding theories and experiments and show that this scheme can capture shock waves sharply without using any spatial fitting method.

View full abstract

In most studies of numerical simulation about Kawasaki Disease, vessel wall assumed to be rigid, although the real vessel wall has hyperelasticity. One of the reasons for that is difficulty to calculate the motion of fluid and hyperelastic wall (structure) simultaneously since fluid calculation is Eulerian while structure is Lagrangian. To solve this problem, the method to introduce the weight function to decide whether the calculation point is fluid or structure is effective. In this study, as a first step for revealing the inner phenomena of a coronary artery, we model it as a hyperelastic circular tube and calculate the inner flow with the above method to consider the interaction between the blood flow and structure, and investigate the necessity of assuming the hyperelasiticy on the vessel wall. In conclusion, the effect of the hyperelasticity is confirmed and it should be assumed to calculate the flow in a coronary artery.

View full abstract

Numerical experiments suggest interesting properties in the several fields of fluid dynamics, plasma physics and population dynamics. Among such properties, there is a striking manifestation of support splitting and merging phenomena in the behaviour of non-stationary seepage. The model equation in one dimensional space is written in the form of the initial-boundary value problem with the effect of a non-linear filtration. In this paper, such phenomena are realized by use of finite difference schemes, and are justified from numerical and analytical points of view.

View full abstract

The purpose of this study is to investigate rheologically complex fluid flow in a Couette-Taylor flow reactor (CTFR) by a numerical calculation. A shear-thinning fluid which can be expressed by Carreau model was selected. The shear-thinning property causes spatial distribution of fluid viscosity in the CTFR. This work, therefore, defined an effective Reynolds number (Re_eff) based on an effective viscosity obtained by a numerical simulation. The critical Re_eff at which Taylor vortices appear corresponded to the critical Re obtained by a linear stability analysis in Newtonian fluids. This means that Re_eff is applicable to predict the occurrence of Taylor vortices in a shear–thinning fluid.

View full abstract

We present numerical results for vibration responses of two circular cylinders which move to in-line and cross-flow directions along a smooth flow, and which are arranged at a spacing larger than the critical spacing of the centre-to-centre of the cylinders. When the Scruton number is low, it is well- known that two kinds of in-line vibrations, named first- and second-excited vibrations, occur in the range of low reduced velocities. On the other hand, when the reduced velocity exceeds approximately 5, it has been reported from many experimental data that the vortex-induced vibration (VIV) was observed in the wide range of the reduced velocity. In order to capture such vibration responses, numerical computations of the Navier-Stokes equations are carried out by means of the Petrov-Galerkin type finite element method and the arbitrary Lagrangian-Eulerian method. In addition, the numerical solution stabilization of the Navier-Stokes equations is performed by the third-order upwind scheme. From our numerical results, we discuss the response mechanisms of the cylinders in detail.

View full abstract

This paper investigates local flow geometry characteristics of inflow vortices in isotropic homogenous turbulence. Eigenvalues and eigenvectors of the velocity gradient tensor are applied to specify the Galilei invariant local flow geometry, and the real part of complex eigenvalues is used to classify the vortex as inflow or outflow. However, the real part can classify only average inflow or outflow, whereas another property, which we term “sourcity,” specifies uniformity of the radial flow direction. We analyze the characteristics of inflow vortices using sourcity with inflow from all or partial directions.
The analysis shows that the rate of vortices with inflow from all directions is approximately 16% of all (average) inflow vortices, and that most inflow vortices include partial outflow. A vortex with inflow from all directions has greater symmetry. Although sourcity mathematically decreases for high-intensity swirling, it increases if swirling develops with vortical flow symmetry, especially for weak vortices.

View full abstract

The performance of the selective smoothed finite element methods (selective S-FEMs) using 4-node tetrahedral (T4) elements with deviatoric/hydrostatic split is evaluated. The selective SFEMs discussed in this study are called FS/NS-FEM-T4 and ES/NS-FEM-T4 for short, which are known as locking-free formulations with no increase in the degrees of freedom. This study reveals that both the selective S-FEMs have three major issues: limitation of material constitutive model to handle; pressure oscillation in case of nearly incompressible material; and corner locking, through some examples of analysis. At the same time, the examples imply that the performance of the selective S-FEMs is comparable to an advanced hybrid T4 finite element formulation in the majority of practical problems.

View full abstract

We conducted a theoretical analysis of the peak factor, defined as the ratio of the response maximum to the response standard deviation of a single-degree-of-freedom system. We constructed a peak factor formula that uses the natural circular frequency, damping factor, and duration of ground motion, and thereby estimated the peak factor of each mode of a multi-degree-of-freedom system. Using these mode-dependent peak factors, we then modified the existing complex complete quadratic combination method to estimate the maximum acceleration of a non-classically damped model. The response of a mid-story isolated building model under white noise ground motion was analyzed, and it was confirmed that the reformed formula globally provides good estimates in most cases.

View full abstract

Equivalent non-Gaussian excitation method is proposed to obtain the moments up to the fourth order of the response of systems under non-Gaussian random excitation. The excitation is prescribed by the probability density and power spectrum. Moment equations for the response can be derived from the stochastic differential equations for the excitation and the system. However, the moment equations are not closed due to the nonlinearity of the diffusion coefficient in the equation for the excitation. In the proposed method, the diffusion coefficient is replaced with the equivalent diffusion coefficient approximately to obtain a closed set of the moment equations. The square of the equivalent diffusion coefficient is expressed by the second-order polynomial. In order to demonstrate the validity of the method, a linear system to non-Gaussian excitation with generalized Gaussian distribution is analyzed. The results show the method is applicable to non-Gaussian excitation with the widely different kurtosis and bandwidth.

View full abstract

In order to analyze mathematical structures of a transient orbit converging to an equilibrium, it is of great use to obtain time evolutions of vectors in tangent spaces along the orbit. We propose a new algorithm to efficiently pull back eigenvectors of linearized system at the equilibrium by using a modified method to calculate covariant Lyapunov vectors. These pulled-back vectors are termed pullback vectors in this paper. We also apply our algorithm to a transient orbit of a simple three-dimensional ordinary differential equation to give the pullback vectors. The pullback vectors are used to illustrate appropriate perturbations to give an orbit whose direction becomes parallel to the corresponding eigenvector of the linearized system at the equilibrium.

View full abstract

Following the recent acknowledgement of a synchronization and stochastic resonance coupling phenomenon, which is known as stochastic synchronization, there have been no further studies conducted in relation to the ocean circulation. Therefore, this study investigates the responses of the oceanic double gyre to external wind forcing with red noise using a 1.5 layer quasi-geostrophic model, and then considers the possibility of stochastic synchronization in the ocean circulation. Results show that by adding red noise to external forcing, synchronization can occur within a parameter range in which it does not occur without noise. These results suggest that potential signals in the system are amplified and appear as synchronization in relation to the added noise, and that stochastic resonance therefore occurs.

View full abstract

We investigate daily share prices of the Nikkei 225 stock index to estimate jump times of the stock index. Since such daily share prices are observed at discrete times, it is difficult to find real jump times. In this paper we consider how to separate jump times from the observed times.

View full abstract

Advection-dispersion phenomena in 1-D open channels are described with a diffusion process model governing Lagrangian particle dynamics. The objective of this paper is to propose a new diffusion process model for analysing advection-dispersion phenomena in open channels with aquatic vegetation, which serves as a physically more appropriate alternative to the conventional models. The concept of regime-switching diffusion process is utilized in formulating the problem so that the physical effects of aquatic vegetation in the advection-dispersion phenomena are described in a consistent manner. The linearity of the system of extended Kolmogorov’s backward equations associated with the model leads to the governing equations of the spatially-distributed statistics, which are key indicators in analytically assessing purification ability of water bodies. Numerical analysis on the deposition probability of suspended sediment particles is performed as an application example of the model. Qualitative and quantitative differences between the proposed and conventional models are also investigated.

View full abstract

Stochastic delay differential equations (SDDEs) are used for models of phenomena, the future states of the systems depend on both the present states and their past states. For SDDEs, several approximate solutions have been considered. In this paper, we investigate the Euler-Maruyama approximate solutions for SDDEs and estimate the mean square error of approximate solutions.

View full abstract

We investigated the hetero-coagulation rates between oppositely charged particles subjected to a turbulent flow. The turbulent hetero-coagulation was induced by mixing and stirring the aqueous suspensions of positively and negatively charged latex particles. The sign and magnitude of the surface charge densities of the particles were conrmed by the analysis of measured electrophoretic mobilities. The turbulent coagulation experiments were carried out as a function of particle diameter, KCl concentration, and turbulent intensity. The rate of coagulation was obtained by measuring temporal change of suspension turbidity. The experimental results show that the hetero-coagulation rate constant increases with decreasing KCl concentration and with increasing particle size and turbulent intensity. These trends qualitatively agree with the theoretical prediction based on the aggregation kinetics in a turbulent ow combined with a trajectory analysis including the attractive electrostatic interactions. The facilitated coagulation rates in lower KCl concentrations are considered to be due to the increase of the magnitude of attractive electrostatic force accompanied by the development of electrical double layer.

View full abstract

The early stage of iron nanopowder fabrication using an argon thermal plasma jet is numerically demonstrated especially focusing on simultaneous growth and transport of iron nanomist around the plasma jet. A simple model is developed to describe the nonequilibrium processes of the nanomist’s collective growth through homogeneous nucleation, heterogeneous condensation and coagulation among the composing nanodroplets, as well as transport by convection, diffusion and thermophoresis. An original solver is also developed to express a turbulent-like plasma flow with multi-scale eddies and to capture discontinuous profiles with steep gradients in a nanomist distribution. A thermal plasma jet entraining cold non-ionized gas, a natural fluid-dynamic feature, is successfully simulated. The collective growth and transport of iron nanomist which is converted from iron vapor transported with the plasma flow are also clarified. A large number of small nanodroplets are generated especially around the plasma fringe. The regions where nanodroplets have large sizes almost coincide with those where nanodroplets exhibit low number densities because the nanomist also grows by coagulation.

View full abstract

The present paper introduces a numerical analysis method for the shape identification problem to prescribe von Mises stress distributions on sub-domains in a thermoelastic field from the point of view of strength design. The square error integrals between the actual von Mises stress distributions and the prescribed distributions on the sub-domains are used as an objective functional. The shape gradient of the shape identification problem is derived theoretically using the adjoint variable method, the Lagrange multiplier method, and the formulae of the material derivative. Reshaping for the shape identification is accomplished using a traction method that was proposed as a solution to shape-optimization problems. The validity of the proposed method is confirmed based on the results of a 2D numerical analysis.

View full abstract

Upward convective heat fluxes in the Venusian low-stability layer (~55 km) become larger as wave-forcing and heating amplitudes are increased in 5.5-day wave and cloud feedback heating (CFH) experiments. In contrast, the upward heat flux is weak and insensitive to the wave-forcing amplitude in 8-day wave experiments, because the forced wave predominantly breaks below the low-stability layer. The planetary-scale wave breaking induces downward heat flux at 45-50 km. In addition, convective penetration produces downward heat fluxes near the top and bottom of the low-stability layer when the convection is fully developed. Above 60 km, vertically propagating gravity waves emitted from the low-stability layer have negative momentum fluxes. The maximum downward eddy momentum flux is proportional to the upward heat flux in the low-stability layer. Fine structures of atmospheric static stability vary between wave propagation, convective penetration, and planetary-scale wave breaking.

View full abstract