Architectural structure of living systems is well-designed in a functional point of view. It is an interesting issue how such functional structure is self-organized in terms of algorithm and rule of design. In this report, we studied a method of biomimetic design for topological optimization, based on the equations of motion for cellular movement in an amoeboid organism of Physarum plasmodium. The key concept is use-and-growth rule: a part of system grows to be reinforced when it is used more often. As this rule is observed in a wide range of physiological processes, we compared similarity and difference in the model equations that can be described the rules in general. First we examined a possible application of the equation to modeling and remodeling of human leg bone. Second we designed a shape and structure of cantilever by using the biomimetic equations for use-and-growth rule. Discussion was made on possibility of biomimetic design according to the use-and-growth rule.

View full abstract

During the 2011 Tohoku earthquake, crustal deformations and tsunami time histories were observed by the GPS devices. In this investigation, inversion of the observation data for fault slip models was carried out. Here, multi time window model was employed for the source time function. Trial inversion analyses with several kinds of combination of rupture velocities and number of time windows indicates the decreasing misfit with the increasing time window numbers and slow rupture velocity. However, the duration time of fault movement under this condition is quite longer than the observed temporal response of crustal deformation. Based on the investigation, it could be said that the inversion models with the rupture velocities 2.5-3.0km/s and 5 time windows are realistic options.

View full abstract

Wild bamboo is a plant which has many peculiar stiff nodes and tissue structures. It has a hollow cylindrical shape, which gives it strength over its height, although there is a risk of losing resistance to bending. The purpose of this study is to examine how the nodes and vascular bundles contribute to bamboo's bending resistance by forming a nearly optimal shape, obtained over millions of years of evolution. From a standpoint of structural mechanics, bamboos can be modelled as cylindrical shells with ring stiffeners. In addition, cylindrical bamboos are regarded as a functionally graded structure due to its distribution of vascular bandles in the cross section of cylinders. From the measurement data analysis and theoretical formulations, we found that wild bamboos effectively control their node spacings as well as other geometric parameters corresponding to the bending moment diagram along the vertical direction. Moreover, we demonstrate theoretically that the vascular bandle distribution enhances the flexural rigidity of bamboos. These new findings give us the possibility of a new design concept of lightweight and high-strength functionally graded structures.

View full abstract

The paper discusses the approach to predict future settlement in soft grounds. The inverse analysis method has been proposed, in which the parameters to simulate the long-term consolidation behavior well including the secondary consolidation, can be identified. Firstly, the consolidation tests were conducted with the separate type interconnected consolidation apparatus in which four consolidation cells were connected. The strains and pore water pressures of the four layers were measured. To predict the secondary consolidation, the log t approach was introduced, and for the primary consolidation, the stochastic nonlinear model was employed to consider the nonlinearity and the spatial variability of parameters, and necessary parameters were identified from measured data. By using proposed method, measured settlements and pore water pressures could be predicted well.

View full abstract

In bridges, vibrations caused by the traffic loading are responsible for advancing the fatigue damage of the bridge. To continually grasp the vehicle weight of the traffic load that generates vibrations are very important in the maintenance of bridge. Currently, as a system for indirectly measuring the axle load of the vehicle traffic on the bridges, there is a BWIM (Bridge Weigh-In-Motion) to implement the load measurement of the traveling vehicle from the analysis of the strain response. This study proposed a method to identify the load effect and the vehicle weight to the beams of the moving load by using the control theory.

View full abstract

Bending strain distribution on wide flanges does not become uniform because of the shear lag. For solving the the shear lag problem, an analytical approach with assumptions of the shear lag displacement on certain particular cross section was proposed, and was subsequently generalized by the harmonic analysis. Authors have proposed semi-analytical approach to solve the shear lag problem with the help of the homogenized beam theory. This approach uses FEM for a representative volume element (RVE) which consists of the solid elements to obtain the pattern of the shear lag displacement. In this paper, we propose to utilize two dimensional distribution of the shear lag displacement in flange cross section obtained by the numerical result of the RVE in order to improve accuracy of the semi-analytical approach. The numerical examples demonstrate that the error of in-plane shear strain on the flange due to the shear lag can be reduced by the proposed method.

View full abstract

In Japan, where mountains districts occupy about 70% of the land, a lot of shallow landslides caused by heavy rains have occurred each year. It is important to identify field soil hydraulic parameters, in order to simulate rainwater infiltration processes adequately. Recently, due to development of sensing technology, inverse analysis for identifying the soil hydraulic parameters by using field measurements, has received a lot of attention. In this study, analytical accuracies of three algorithms based on particle filter, which was one of the inverse analysis method, was compared, in order to grasp an appropriate algorithm based on particle filter, that could identify soil hydraulic parameters effectively and efficiently. As a result, SIR_omit, reducing the number of simulations with the lapse of time, could identify field soil hydraulic parameters adequately. Thus, SIR_omit was most available as inverse analysis method based on field measurements.

View full abstract

Internal erosion is often observed in soil structures and it is a geotechnically important phenomenon. Modelling of internal erosin is still in development on the soil deformation due to internal erosion and the evolution law for internal erosion. In the present study, we have formulated a multiphase analysis method for the coupled phenomena including soil skeleton deformation, water flow, internal erosion and soil paticle advection. We have also adopted initiation conditions and the evolution rate equation for internal erosion as constitutive equations. From the numerical calculation for an experiment of suffusion, we have discussed the suffusion mainly focusing on the movement of eroded soil mass.

View full abstract

The numerous methods of a crack detection have been developed for the purpose of assessing the integrity of the structure. Among them, pattern recognition techniques for digital image are to evaluate cracks morphologically based on image processing, which are effective in the case that the crack on the surface is visually apparent. However, fault detection might be unavoidable in principal of the system for the crack which is difficult to identify visually due to surface condition of the observation structures.
An attempt is made in this study to develop an invisible crack detection system based on non-contact displacement measurements. The maximum principal strain and the principal direction, which are calculated from the displacement vectors, are the prime parameters for crack identification. Based on the maximum principal strain and principal direction, the proposed system detects crack in each discretized elements on the digital image. The validity of the proposed SHM system is demonstrated through laboratory experiment.

View full abstract

This paper presents the long-term prediction of the ground behavior, including secondary consolidation via the particle filter, using model test results. There have been some difficulties in evaluating the behavior accurately, such as two-dimensional displacements, because an unsuitable value for Poisson's ratio and the constitutive model are generally employed for consolidation problems. This paper identifies Poisson's ratio in addition to the compression index, the permeability, the initial volumetric strain, and the secondary compression index, and adopts the modified Cam-clay model, considering anisotropy, to overcome the problem, i.e. the accurate evaluation of the two-dimensional displacement. The prediction is performed using the identified parameters, and numerical examples of accurate predictions of the ground behavior are shown.

View full abstract

This paper presents a formulation of unified models for beam and shell in curvilinear coordinate system, based on meta-modeling. Traditional structural mechanics is restricted to simple geometries and governing equations are available for a handful number of simple geometries. They are mainly derived based on stress resultants equilibrium of free body diagrams. This gives difficulty in analyzing structural elements with complex geometries analytically and numerically; involving tedious and error prone process. These limita- tions can be eliminated by formulating beam and shell theories for arbitrary geometry defined in curvilinear coordinate system, based on continuum mechanics. Motivated by this improvement, main objective of this work is to develop beam and shell models for arbitrary geometry, using curvilinear coordinate system based on meta-modeling. Meta-modeling guarantees the consistency of the derived beam and shell mod- els with continuum mechanics and tensorial formulation in curvilinear coordinates produces models valid for arbitrary geometries. Governing equations for any specific geometry can be easily obtained simply by substituting the metric tensor of the coordinate system for the problem. This work is mainly based on first order approximations of field variables involved and standard variation process of Hellinger-Reissner func- tional. Some verification tests are done with simple geometries found in literature and it can be clearly seen that they are well matched with literature. Some analytical advantages of derived models are: availabil- ity of governing equations for arbitrary geometries; possible rigorous treatment of material non-linearity; etc. while some numerical advantages are: reduction of per-node number of degrees of freedom; faster convergence of iterative solvers; reduction of number of elements required; increase in accuracy; etc.

View full abstract

This paper presented a stochastic approach to identify time-related capture zones of solute in pumping wells in geostatistically generated hydraulic conductivity fields based on field data. Random walk particle tracking was applied to assess the travel time and subsequent spatial probability distributions of introduced ensemble lattice cells having the probability that particles located initially within cells reach the pumping wells during a certain time of the pumping. Proposed methodologies provided a few proper time-related outcomes corresponding to the boundary conditions, pumping rates and pumping locations. Based on the probability of each lattice cell, entropy was introduced for computing a quantitative measure expressing the uncertainty relevant to the spatial probability distributions of the solute capture zones, resulting in the increase of the entropy with the extension of the solute capture zone. Moreover, it was demonstrated that the magnitude of the time-related capture zones of groundwater, which was identified using backward particle tracking from a pumping well, was smaller than that of the time-related capture zones of solute due to the effect of solute dispersion phenomena.

View full abstract

We developed an equivalent continuum form for brick structures based on continuumnization proposed by Hori et al.¹⁾ This allows one to analytically study the characteristic properties of masonry brick walls, and apply numerical techniques used in continuum mechanics to simulate brick structures. Further, the continuum form opens up the possibility of developing simplified models like shells or beams, which will be convenient in designing brick structures. Based on the continuum form, we study how the wave char- acteristics, like wave speeds, etc., change according to brick arrangement and material properties. Further, we develop a PDS-FEM model²⁾ for simulating brick structures and verify the developed model comparing numerically obtained wave speeds with that of analytical predictions.

View full abstract

A multi-material Eulerian finite element formulation for contact with slip is proposed in this paper. Eulerian finite element methods, which are effective for large deformation problems and free-moving boundary problems, cannot simulate contact problems with slip using the mixture theory because it assumes fully bonded interfaces between materials. Considering the above backgrounds, we propose a new Eulerian formulation based on the phantom-node method to model discontinuous velocity fields across material interfaces for contact with slip. Material interfaces are defined by the 3D PLIC method and surface force vectors are introduced using a penalty method. The present approach is validated through simulations of contact problems.

View full abstract

The Material Point Method (MPM) uses two kinds of space discretization. One is physical space, in which physical quantities are evaluated, and the other is calculation space, in which discretized domain movements are solved. In the course of the MPM process, physical quantities are mapped from material points to a numerical grid using interpolation functions. On the contrary, material points recover physical quantities from their motions, which are mapped from the numerical grid to material points using interpolation functions. In its original form, MPM treats each material point as a point. Later, an enhanced method called the Generalized Interpolation Material Point (GIMP) method was proposed in which the effect of the size of the controlling domain is considered. More recently, the Convected Particle Domain Interpolation (CPDI) method has been proposed, in which a material point is considered to be an arbitrarily shaped controlling domain. Thus MPM and its derivatives are gradually gaining improved and enhanced numerical characteristics.
The CPDI method, which is able to consider arbitrarily shaped domain integration, has been formulated as an interpolation function for a two-dimensional controlling domain using a direct integration technique. However, it is difficult to extend direct integration to three-dimensional problems because of a drastic increase in the number of terms, leading to computational complexity. In this paper, the authors employ a numerical integration technique for domain integration over the material points in the CPDI method. Called the Arbitrary Particle Domain Interpolation (APDI) method, it has been applied to three-dimensional analysis. The APDI method allows use of the same numerical procedure even if material points have controlling domains of different shapes. Through a demonstrative analysis of two- and three-dimensional numerical problems related to geomaterials, the applicability and effectiveness of the APDI method are clarified.

View full abstract

When Spectral Stochastic Finite Element Method (ie. SSFEM) is applied to elasto-plastic analysis, the key points are how to evaluate stochastically varying stress which is used as the evaluation point of tangent stiffness matrix and how to evaluate tangent stiffness matrix based on such a stress. Although Non-Linear Spectral Stochastic Finite Element Method (ie. NL-SSFEM), which is developed based on upper and lower bounding media theory under the assumptions of quasi-static and small strain condition, is one of the ways to use SSFEM in non-linear frame, once we have tried to extend this method to the large deformation problem, the treatment becomes very complicated and the same assumptions such as quasi-static and small strain condition are needed. Therefore, we tried to build NL-SSFEM from a different viewpoint by using partially the approach called NISP (Non-Intrusive Spectral Projection scheme) to estimate tangent stiffness matrix, then we have considered the advantages and disadvantages arising from the use of this approach.

View full abstract

In this paper, a computational method for thermal interactions between compressible fluids and solids was applied to the natural convection around two horizontal circular cylinders that have constant temperature in vertical array. This method enables us to simultaneously calculate natural convection and heat conduction around and in arbitrarily shaped solids considering compressibility of fluids. As a result of computations, predicted isotherms around the cylinders were in good agreement with experimental results. In addition, numerical experiments for heat-generating cylinders, which have heat generation per unit volume, were conducted and reasonable temperature distributions were obtained considering effects of heat conduction in the cylinders.

View full abstract

This paper describes a preconditioner of the conjugate gradient method which utilizes the sparse Cholesky factorization to solve particular dynamic soil-structure interaction problems. This preconditioner is designed for finite element models in which some building models consisting of beam or shell elements and a large soil model consisting of solid elements are combined together. The idea of the preconditioner is based on the two observations: 1) a sole building model can efficiently be solved by the sparse Cholesky factorization; and 2) the building models worsen the iteration behavior of the conjugate gradient method. Within an algebraic study, it is shown that the preconditioner reduces the coefficients with respect to the degrees of freedom of the building models into the identity and eliminates the terms related to the building models from the preconditioned coefficient matrix. This algebraic property is expected to facilitate improving the iteration behavior of the conjugate gradient method applied to the combined model. Numerical studies are also performed to demonstrate that this property holds in the actual computation. The other beneficial characteristics of the preconditioner are also qualitatively discussed through the results of residual force and moment.

View full abstract

This paper presents an FE-based elastic wave transmission analysis method for a composite layer in which defects are imbedded periodically. Scattering waves originated from the defects are represented by a dynamic load acting on a contour surrounding the defect. This formulation enables to reduce the original problem to the dynamic reaction of the periodic layer without defects. By virtue of Floquet transform, the solution is derived from FE analyses of a unit region of the periodic layer. Numerical results showed that the introduction of a defect array contributes to the enhancement of capability for insulation, while the stop band will be vanished.

View full abstract

This paper presents a simulation method based on the wave-acoustic theory and its application to the auralization of construction noise. The CIP method using adaptive mesh refinement (AMR) is employed for solving the governing equation accurately. In order to obtain the accurate numerical solution by CIP method for the propagation of an impulse wave considering the reflected wave, a pseudo impulse developed by Lubich is employed. The auralization of construction noise is performed by the convolution using the results and the sound source data. The present method is applied to three-dimensional wave propagation problems in order to investigate the validity of the method. The present method is shown to be an useful tool to investigate the construction noise.

View full abstract

This paper deals with nonlinear analysis of elastic wave scattering by cracks with contact boundary conditions in three-dimension (3-D). A generation mechanism of sub-harmonic waves due to contact acoustic nonlinearity has not been understood clearly. However, 2-D simulations of sub-harmonic generation by cracks with specific shape and arrangement have been conducted (Maruyama et al. 2015). Therefore, 3-D numerical simulations for similar models to the case of 2-D are performed. Through obtained numerical results, cause of the sub-harmonic generation phenomenon is discussed. The implicit Runge-Kutta based convolution quadrature time-domain boundary element method is used for precise and stable simulation.

View full abstract

Present seismic design of soil embankment considers only shear failure, e.g., circular slip. However, open cracks likely caused by tensile failure were observed during the past earthquake. We perform the numerical analysis to reproduce failure tensile crack propagation in soil embankment. In previous study, elasto-plastic analysis for soil embankment considering shear and tensile failure was performed. The elasto- plastic model used in that study can reproduce strain-softening of dense sand by both shear and tensile deformation. We apply eXtended finite element method (X-FEM) to the dynamic behavior of soil embankment with considering crack propagation. The model enables to include typical crack behaviors such as strain concentration, crack opening and contact of crack planes. We introduce criteria of crack occurrence, crack position and angle, and then show a numerical example of dynamic analysis of soil embankment. The numerical result shows typical failure patterns of soil embankment.

View full abstract

A linearized inverse scattering technique with the aid of the convolution quadrature time-domain boundary element method (CQBEM) has been developed for the reconstruction of a defect shape in an anisotropic plate. The convolution quadrature based BEM is more practical numerical method better than the conventional one, because the CQBEM can produce better accuracy and take stably large time steps. A Green's function is applied to the CQBEM and the inverse scattering approach to save their computational time and memory. The inverse scattering with the Born inversion is applied to the scattered wave forms which include the direct back scatter, the back scatter due to re ection at the bottom of an anisotropic plate, and the both of them. It is found that the proposed inverse scattering technique with CQBEM can reconstruct the image of a cavity properly in an anisotropic plate under the limited use of scattered wave form data.

View full abstract

This paper presents the comparison between numerical and experimental results of RC beams with different shear reinforcements failing in flexural or shear mode to demonstrate the validity of 3D fracture simulation with a damage model. The modified von-Mises damage model based on fracture mechanics for concrete is applied to modeling of crack propagation in concrete whereas the von-Mises plasticity model is applied to reinforcements. To simulate meso-scale fracture behavior of RC beam, the material behaviors of steel and concrete are separately discretized with the finite element mesh in which the detailed geometry of reinforcement is reflected. We first show the formulation of the constitutive models for steel and concrete. Particularly the modeling of crack formation due to the degradation of stiffness using the damage model is accounted for. Then, numerical and experimental results are presented for 4-point bend tests of RC beams with different shear reinforcements. The comparisons show that the proposed method is capable of simulating the fracture behavior of RC beams with different failure modes without changing material parameters, and additionally the simulated results are comparable with the experimental ones.

View full abstract

This paper presents a numerical simulation of a road traffic noise based on fast multipole BEM in time domain(TD-FMBEM) and its application to the auralization using VR system. Because this problem is the large scale problem in time domain, the stabilized method is applied to TD-FMBEM. We solved the traffic noise field having a sound barrier and obtained the stabilized sound pressure using the stabilized TD-FMBEM. We also realize the visualization and auralization of the numerically obtained sound data using VR system.

View full abstract

The influence of spatial variation of elastic modulus of ballast material on vibration response of a ballasted railway track is investigated using the spectral stochastic FEM(SSFEM) based on a coupling of a beam-spring-mass system and a 2-D elastodynamic body. In the present SSFEM the dynamic response of ballast layer is considered using numerically generated Green's function. The fictitious reflection of the elastic wave travelling in the ballast layer is reduced with a viscous boundary. Through the SSFEM simulation on the vibration of the railway track with spatial 10% variation of elastic modulus of ballast, the variations of the sleeper-ballast force, the railpad force and the wheel-rail contact force are under 1% of their maximum forces.

View full abstract

The method of manufactured solution (MMS) proposed by Roache et al. is widely used to verify numerical codes in fluid dynamics. In this methods, second derivatives of given solutions or spatial derivatives of stresses derived from given solutions are required to calculate body forces and hence it has not been popular in solid mechanics. The authors proposed the method of nearby problems, which includes the calculation process of body forces similar to that in MMS, to the finite element analysis of solid structures. In this work, this approach is applied to MMS for the finite element analysis of large deformation problems of hyperelasticity.

View full abstract

This paper presents a 2D-3D hybrid method based on unstructured mesh for tsunami simulation. As a coupling method for the 2D and the 3D, the overlapping method using arbitrary grid has superior applicability for the wave transferring from bidirection. The 2D shallow water equation and the 3D Navier-Stokes equation are used as the governing equations. Several numerical examples are used to show the validity and the effectiveness of the presented method.

View full abstract

In this study, the local scour behind the weir due to overflows was calculated using three-dimensional multiphase numerical model (MICS). In this method, the shape of a gravel particle is represented by multiple tetrahedron elements, and collisions between gravel particles are treated by the contacts of the multiple collision detection spheres placed inside of the gravel particle. Fluid forces acting on the gravel particle are calculated from the pressure-gradient and viscous terms in the momentum equations for multiphase fields. As a result of the numerical experiments, movements of 14,400 gravel particles due to the overflow were predicted and local scour arose behind the weir. In addition, the scour depth and area become larger when the flow rate is large.

View full abstract

In a previous paper, we have proposed a novel reduction method for the treatment of the continua. Here I provide several lines of evidence showing that the reduction method can be regarded as a generalized force method. In the reduction method, the simultaneous equation is constructed with both equilibrium and compatibility conditions using the stiffness matrices of each individual finite element. Therefore, the reduction method includes both deformation method and force method. When the compatibility condition rows are satisfied at first, then equilibrium condition rows should be solved separately. However, it is important to notice that the remaining rows can be regarded as deformation method matrix because the unknowns of simultaneous equation are represented by a parameter vector contracting the plural parameter vectors of the individual element. If a scheme for the statically determinate main system is successfully established by eliminating extra parameters, then only the force method matrix remains. In this context, the reduction method should be regarded as a generalized force method, which I call GFM, rather than classical two-step-type force method, which at first looks for the statically determinate main system and then solves the force method matrix. Although we previously expected that the iteration method would be adaptively improved for solving the simultaneous equation of GFM, I found in this study that neither the explicit methods such as the improved SOR method nor the implicit methods such as the Krylov's space method cannot be useful. Hence, I propose an alternative scheme based on Guassian elimination method, modified with skyline method that memorize nonzero elements only and sweep-out method that eliminates extra parameters by using absolute max. pivot.

View full abstract

Concentration of chloride ion in reinforced concrete involving internal cracks is quantitatively analyzed by using a numerical method which is capable of simulating 3D crack propagations and the mass transfer. The FEM with a damage model based on fracture mechanics for concrete in terms of fracture energy is applied to simulate 3D crack propagations. The transfer of chloride ions is described by the unsteady diffusion equation. The diffusion coefficients in cracked concrete are expressed as a function of the damage variable obtained from the crack propagation analysis. The formulation of crack propagation analysis with the damage model and unsteady diffusion analysis in consideration of damage in concrete are presented in Section 2. Section 3 shows a verification analysis of internal cracking in concrete to demonstrate that the crack width can be evaluated independently of mesh-size. Finally, the comparisons between numerical and experimental results show that the present analysis is capable of evaluating the chloride ion concentration in concrete involving internal cracks with satisfactory accuracy.

View full abstract

In order to increase accuracy of predicting the dynamic behavior of unsaturated soil, the method was proposed in which the change of void during cyclic loading was considered in the constitutive model of soil skeleton and the soil water characteristic curve. Moreover, the scanning curve which aims to the main wetting curve was also prepared.
These methods were employed in the governing equations of cyclic triaxial test in undrained water and air conditions. For the validation of the methods, the simulations of two kinds of the cyclic unsaturated triaxial tests were performed. It was revealed that the simulations in consideration of these methods could successfully describe the decrease of suction and skeleton stress and the amount of deviatoric stress observed in experiments.

View full abstract

We proposed a direct numerical simulation model of seepage flow and granular soils by combining the lattice Boltzmann method and the discrete element method. Multiple Relaxation Time (MRT) model was introduced in order to obtain stable solutions of the fluid flow under high Reynolds number condition. Partially Saturated (PS) model, which retain a local operation at each fluid node and keep from intensive increasing the computational costs for the calculation of collision term, was also introduced as a solid-fluid coupled model. We show the effectiveness of the PS-MRT lattice Boltzmann model through several validation tests.

View full abstract

A new formulation of elastic body is proposed based on the concept of multiple shear mechanism referring to a strain space multiple shear mechanism model for geomaterials such as sands. In the proposed model, the relationship between the virtual simple shear strain and the virtual simple shear stress in the multiple shear mechanism is assumed to be linear or nonlinear considering the effect of strain softening. In addition, the formulation is written by both the material and spatial description based on the finite strain theory in order to consider the geometrical nonlinearity due to large deformation and rotation. Comparison between experiments for rubber materials under simple or pure shear condition and its simulation results demonstrates the capability of the proposed model even though only two model parameters (i.e., shear modulus of rigidity and shear strength in the virtual simple shear mechanism) are required to be specified.

View full abstract

Image analysis with ultra-high-speed camera and dynamic numerical analysis are applied to the unstable crack growth of wing crack under the uniaxial compression. During the unstable growth, growing wing crack terminates and restarts their growth in some cases. The termination and restart behavior of the growing crack is discussed through the experiment and numerical analysis in this research.
First, images of rapid propagation of wing cracks under uniaxial compression are captured by the ultra-high-speed camera with the frame rate of 500k frames per second. Stress field around the moving crack tip is visualized by means of photo-elastic technique, and images of the resultant fringe patter are also captured by the camera.
Next, two dimensional dynamic numerical analysis is performed. PDS-FEM (Particle Discretization Scheme), which allows the discontinuity of the displacement in the continuous body, is combined with the central difference time integration scheme to simulate the unstable growth of the wing crack dynamically. The accuracy of the simulation is discussed through the comparison with the images which are captured by the high speed camera.

View full abstract

Elastoplastic constitutive theory with tangential plasticity, which is able to describe the inelastic strain rate due to the stress rate tangential to the plastic potential surface, has been proposed for prediction accuracy enhancement of the inelastic material behavior under non-proportional loading. This tangential plastic theory, however, is not able to apply to the prediction of cyclic softening behavior of metal subjected to the non-proportional cyclic stress histories with the stress amplitude less than the yield stress such as high cyclic fatigue problem. In this article, the proposed formulation of tangential plasticity theory is enhaced to more accurately describe the cyclic softening behavior of metal under non-proportional cyclic loading which has stress amplitude less than the yield stress. The numerical simulations with this enhanced tangential plasticity theory for the material behavior prediction under the non-proportional circular shaped stress path is conducted. The response characteristics of the enhanced tangential plasticity theory is revealed in detail according to comparing the simulation results with ones by the traditional plasticity theory and the conventional tangential plasticity theory.

View full abstract

Fatigue life extension of materials and structures have great impact on any engineering field, and then have been extensively researched up to the present.
Among several life extension techniques for metals and its structures, the laser peening (LP) with nano-second laser and/or femto-second laser have been highlighted, since the both of the laser peening techniques can generate deep compressive stress and material hardening fields. However, the difference between these processes and the optimum condition on the fatigue life extension are not well understood since their are conducted under extremely high strain-rate conditions.
In this work, the effects of laser peening processes incorporating both nano-second laser and femto-second laser on the predicted mechanical properties are studied by using a FEM models created with ABAQUS Dynamic schime. A Johnson-Cook material model considering temperature and strain-rate effects is adopted in the analysis, and a mechanism of residual stress, material haedening and deformation generation in the models and difference of the predicted results between both LP processes is discussed in detail.

View full abstract

In this paper, to discriminate the material damping and scattering damping of elastic waves in concrete material, Coda waves obtained by actual specimen and numerical simulation was evaluated. Furthermore, to evaluate the complexity of wave structure, Hurst index was obtained by the group delay velocity, assuming that the group delay velocity is one of Fractional Brownian Motion. Material damping was evaluated by P-waves and semi-variogram was obtained to represent the complexity of special randomness. Finally it was found that phase randomness has some correlation with special randomness and this fact may lead to enhancement of quality evaluation of concrete.

View full abstract

A series of cylinder drag experiments were conducted using the density-matching Poly-Styrene Beads particle-fluids mixture to study the flow characteristics of liquefied sands. PIV technique was used to visualize the velocity fields around the moving cylinder. SPH (Smoothed Particle Hydrodynamics) simulations were also conducted to compare with the experimental results. Experiments results show that the solid fraction of 0.555 is quite important in this particle fluid mixture because the drag force exerted on the cylinder increases sharply with the solid fraction if it is greater than this value. This value coincides with the random loose packing density of mono-disperse spheres. PIV analysis shows that the velocity fields are quite localized around the cylinder, and the localized zone is expanded more in the moving direction than in the perpendicular direction. On the other hand, it turned out that the SPH simulations with a simple viscous fluid model cannot reproduce the similar velocity field, which indicates this particle-fluid mixture cannot be regarded as viscous fluid.

View full abstract

This paper reports a seismic assessment of ground-chimney interaction system based on the seismic response analyses of an actural chimey structure with in a 2D plane strain condition. It finds that, 1) seismic stability of aboveground structure cannot evaluate only from the degree of input acceleration, but the relationship between natural frequency of structure and dominant frequency of input seismic motion is important. That is, if natural frequency of structure and dominant frequency of input seismic motion is close, resonance of the structure causes large oscillation. Moreover, if ground liquefies, the decline in the bearing force of the ground and the consequent unevenness in the settlement of the structure causes collapse. 2) If the irregularity of geological formation was take into accout for evaluation, residual displacement of the structure increases. 3) If presence of an adjacent structure was take into account for evaluation, residual displacement of the structure increase towards each other direction which indicate that the seismic stability assessment of isolated structure with in a crowded place may calculate smaller displacement.

View full abstract

Abstract - This study aimed to elucidate mechanical behavior of bundle clusters which are composed of magnetic particles using DEM (Discrete Element Method). A commonly used particle-wise magnetic interaction in a uniform magnetic field was introduced in the 3D DEM code. First, a single straight chain of the particles subjected to extension or shear was simulated, and the results were verified with the theoretical formula for the particle chain of infinite length derived by Furst and Gast. Then, the simulation result of the single chain was validated with the experimental results of 1.5 mm steel balls chain. Finally, a series of shear tests of bundle clusters of a magnetic powder formed between magnetic plates were performed, and the observed shear strength was compared with the experimental results by Oshima. It was found that (1) the mono-disperse powder exhibits 25% higher shear strength than that of normal size distribution, (2) the shear strength is correlated with the average angle of the bundle clusters, and (3) the major axis of the contact normal distribution within the bundle clusters also coincides with the bundle cluster orientation.

View full abstract

Cushioning materials, such as sand cushion and granular mats placed on rock sheds, can effectively disperse and reduce rock fall energy before rocks collide with protection works. Although the construction costs for sedimentary layers are low, these layers are attracting attention as construction devices that can substantially improve the margin of safety ratio. To elucidate the relationship between the rigidity of the sand cushion bottom boundary and the impact force generated by the collision of a rock on sand cushion, this study examines the stress propagation behavior in the sand cushion with different bottom boundary conditions by 2D-DEM simulation. The results are summarized as follows: (1) According to the obtained experimental result with different rigidity of bottom boundary of sand cushion, the maximum impact force is smaller in the low rigidity condition than in the high rigidity condition, and the duration of the impact force is longer in the low rigidity condition than in the other condition. If the rock shed is simplified modeled by elastic beam in DEM analysis, the obtained experimental result can be qualitatively reproduced by using DEM. (2) The stress waves generated by the collision of a rock on sand cushion propagate forward the bottom of sand cushion, and the stress waves are reflected on the bottom boundary. After the reflect stress waves pass through a part of sand cushion, the high rigidity condition and the low rigidity condition value, which are the mean principle stress waveform generated in the part of sand cushion, become separated from each other. (3) The mean principle stress of reflected stress waves are smaller in the low rigidity condition than in the high rigidity condition. Because in the case of low rigidity condition, the lower part of the sand cushion become low-density due to the loading point displacement is large.

View full abstract

The percolation of fine grains through granular bed is an important phenomenon in various engineering fields. This study focuses on the effect of inter-granular adhesion on the percolation ratio. A series of experiments with spherical glass beads and irregularly-shaped particles was performed, and the results were compared with Discrete Element simulations considering van der Waals (vdW) attraction model. It turned out that the simulation can reproduce the experimental results quantitavely. It was also found that the effect of inter-granular adhesion is not negligible even in the percolation of relatively large grain as 400-700μm.

View full abstract

In order to study suitable calculation conditions for estimating impact force of sediment flow using DEM, a simulation of a large scale sediment flow experiment was performed. By changing element size and element shape, the effects of thease factors were investigated in the simulation. Based on the obtained results, importance of element size and flow chracteristics of the spherical ekement and non-sherical eklement are analyzed. In addition, a flow of the estimation process is proposed, and some points of attention are summarized from the view points of engineering.

View full abstract

Effects of mass density ratio on clustering of small particles in a homogeneous isotropic turbulence have been investigated by direct simulations of the Navier-Stokes equations. An emphasis is placed on particles with small excess density, which are representative to those encountered in marine and aquatic environments. It is found that particles with specific gravity of 1:005 exhibit a weak preferential concentration, only when the fluid and particle velocities are initially not the same. The non-uniform distribution of particles is a transient phenomenon and disappears in a period of several tens of the Kolmogorov time scale. With increasing the specific gravity, the particle distribution becomes non-uniform, irrespective of the initial conditions. It is shown that the extent of preferential concentration does not depend on the particle specific gravity any more when it reaches to around 100.

View full abstract

The macroscopic drag-force model for urban/forest/vegetation canopy flows has a bottleneck in providing an appropriate set of model drag coefficients C_D to the canopy layer, even if its geometrical structure is quite simple. To resolve the issue, the authors have recently explored an approach, in which the preliminary analysis, a direct simulation of canopy flow, is carried out in advance to the macroscopic model prediction (main analysis) to obtain a proper C_D profile for the target problem. To introduce the dependence of the model coefficient on the spatial resolution employed in the main analysis explicitly, this paper introduces two modifications to the approach:(i) A low-pass spatial filtering whose size is comparable to the grid resolution of the main analysis is performed to the results of the preliminary analysis, and (ii) not just C_D but λC_D, where λ denotes the canopy density, is estimated together from the smeared out data. Applicability of the modified approach is demonstrated in an open-channel flow with patched vegetation zones.

View full abstract

We developed predictable model for driftwood motions around obstacles by coupling a three-diemensional CFD model and a Lagrangian type driftwood model. In the present model, three-dimensional flow is solved with Reynolds averaged Navier Stokes equations and the driftwood motion is solved with a Lagrangian approach by modeling a driftwood as a series of spheres. The collision of driftwood is considered with DEM approach. The interaction between driftwood and water is considered by two cases; one-way approach and two-way approach. Only the two-way approach takes into account the drag force from the driftwood to water flow. The computations were performed under the condition of existing laboratory experiment for a dam reservoir. The present model could reproduce general features of driftwood motions around obstacles well. It is important to consider the collisions of driftwood and also two-way approach to estimate precisely the driftwood captureing processes around obstacles.

View full abstract

Toe protection works for coastal dikes using concrete blocks have been proposed as one of the countermeasures to mitigate severe damage of the dikes foundation ground due to tsunami overflows. However, both mechanical stability of blocks and applicability of those engineering works are not fully clarified because it is generally difficult to investigate the hydrodynamics in such supercritical flows.
For this study, we examined the hydrodynamic forces exerted on a block of toe protection works on the foundation, using numerical simulations of tsunami flows over a model dike, which were conducted for four levels of tsunami height. Additionally, we performed laboratory experiments under the same hydraulic conditions as numerical simulations. Then, we investigated the validity of our numerical simulations, comparing outputs such as water depth, velocity, pressure, and hydrodynamic forces.
Results showed that the outputs of the numerical simulations are quantitatively consistent with those of the experiments, except for the lift force. This might be attributable to the installation condition of the target block against surrounding blocks of the protection work: a flat, regular, and even situation was not fully controlled for the block installation in experiments.

View full abstract

In this study, a Lagrangian formulation of the Navier-Stokes equations, based on the weakly compressible smoothed particle hydrodynamics (WC-SPH) method, was applied to simulate the seepage failure around sheet-pile. In this simulation, the advantages of SPH will be exploited to simulate the soil-water interaction. Water is considered as a viscous fluid with weak compressibility and soil is assumed to be an elastic-plastic material. The elastic-perfectly plastic model based on Mohr-Coulomb's failure criterion is implemented in SPH formulations to model the soil movement. Interaction between soil and water is taken into account by means of seepage force and pore water pressure. Numerical Simulation of the 2-D classical seepage failure problem of horizontal ground with an embedded sheet pile has been done. The numerical results were verified by comparison with model test results, the results have shown that the proposed model could be considered a powerful tool to simulate extremely large deformation and failure of soil.

View full abstract

Secondary flow effects are considered in two dimensional depth averaged model to simulate the open channel flow in a sharp bend channel, and the simulated flow structures are compared with different kind of secondary flow models: Nays2DH solver without considering the secondary flow effect in flow equations; Onda et al model using nonlinear secondary flow equation with additional nonlinear transport equation for secondary flow strength; a new nonlinear secondary flow model which considers the secondary flow effect on the vertical distribution of streamline velocity and the inertia effect of the secondary flow advection. These three model results are compared to experimental results. Large difference of flow structures between Nays2DH and the experimental result means that 2D (two-dimensional) depth averaged model without secondary flow effect is inapplicable for sharp bend flow prediction. Obvious weaker deviation of the mainstream from inner bank of Onda et al model than that of the experiment shows that Onda et al model is insufficient for strong nonlinear phenomenon in a sharp bend. Mainstream deviation and other unreasonable structures in linear model from the experiment are similar proofs that a refined nonlinear secondary flow modelling is necessary in simulation of sharp bend flow.

View full abstract