With the advent of structural optimization, the researches for shape optimization have also been under development with growing interest. In this research, in order to minimize the stress concentration, one optimal shape design method originated from the concept of biological adaptation process was proposed. The characteristic of this new method is in no need of structural sensitivity analysis. The algorithm and availability of this method was explained and verified in this paper. In addition, the excellence of computation efficiency of this method was also illustrated.

View full abstract

A practical simulation method for analysing direct derailment limit of running vehicle for high speed train on oscillating track under an earthquake using a 3 dimensional one vehicle model was proposed. In this method, the Kalker's simple theory and a newly introduced 2 point contact algorithm were applied to estimation of wheel/rail contact force. These application carried a possibility of direct derailment estimation which differed from previous indirect derailment estimation. It was clarified that the jumping derailment occurred at lower lateral frequency of track oscillation but the mounting derailment did at higher frequency and that the derailment limit of track amplitude showed larger value than that of indirect previous method.

View full abstract

Identification algorithms were proposed that use the H_∞ filter to identify linear structural systems. Characteristics of the H_∞ filter for structural system identification were studied in detail by checking digital simulation results obtained by using the H_∞ and Kalman filters. Application of the proposed identification algorithms to SDOF and MDOF structural systems shows that the identified parameters obtained with the H_∞ filter converge faster and closer to the exact values than do those obtained with the Kalman filter. The H_∞ filter is more robust than the Kalman filter for identifying linear structural systems.

View full abstract

In the FEM formulation of a plane beam element, the interpolation is improved toward consistency with the beam-column theory. The geometrical effects of the flexural and axial displacements upon the longitudinal strain are explicitly analyzed in up to their second-order terms. The flexural displacement is framed, as usual, in the cubic-polynomoal distributions, the magnitudes of which are linear with the nodal parameters. But, in the interpolation of the axial displacement, the secondary effects of the flexural displacement are taken into account, which are expressed in a quadratic form of the deflection parameters.

View full abstract

Long term earthquake observations at different tunnel sites within a variety of alluvial soil deposits have clearly demonstrated that a circular tunnel is liable to deform in such a way that its two diagonal diameters crossing each other expand and contract alternately. Based on this knowledge, the soil-tunnel interaction for this particular vibration mode is investigated. The soil surrounding a tunnel is assumed to be a homogeneous elastic medium, and two extreme boundary conditions on the soil-tunnel interface are considered; the condition of fixed tangential strain and the shear stress-free condition. The former case suggests a firm bond between the tunnel lining and the soil, whereas the latter case is associated with an artificial slippery soiltunnel interface.

View full abstract

A series of shear loading tests were carried out on 14 panel specimens that had various layouts of diaphragms, shear connectors or full-webs, taking a practical example of the critical portion of members on a shear design in the Kobe Minatojime immersed tunnel into consideration. In particular, the panel specimens included 7 of them that had not only diaphragms but also a full-web. Behavior of the panels did not coincide with beam theory based on continuum mechanics but agreed with a truss analogy.

View full abstract

Efficient hierarchical elements are presented for the p-version of the finite element method. The shape functions of the elements are expressed by simple polynomial series based on the p-approximation concept and the polynomials and those derivatives are given in the form of recurrence formulas, respectively. Furthermore, integral formulas for the stiffness matrices of regular elements are given in the explicit forms.
In this paper, quadrilateral, triangular and hexahedral elements are formulated for plane stress, plate bending and three dimensional stress analysis, and singular elements could be derived easily from those elements by the use of hierarchcal mapping technique and show superior convergence for concentrated loading.

View full abstract

This paper studies the homogenization theory for fluid-saturated elastic porous solids that have been of great interest to many geotechnical engineers. Starting from the field equations separately defined for the fluid and solid in a microscopic scale, the method of two-scale asymptotic expansions enables us to characterize the macroscopic mechanical behaviors that are identical to those of the Biot's consolidation theory and to identify the homogenized material properties from the micromechanical analyses. The method provides not only the homogenized governing equations for overall structure but also, by the localization process, the detailed mechanical responses within the microstructure such as the viscous flow in the fluid phase and the stress distribution in the solid phase. Thus, we can examine the mechanical aspects of the formulation and the derivied equations by clarifying the relationship between the classical and the homogenization theories from engineering view points.

View full abstract

The mechanical behaviors of solid-fluid mixtures are characterized by using the homogenization method which is based on the method of two-scale asymptotic expansions. The asymptotic behavior reflects the features of periodic microstructures where the constitutive relations for solid and fluid phases are separately defined and is recognized as that for linear viscoelasticity. The result is due to the present choice of the so-called effective parameters and clarifies the roles of individual phases in the mechanical behaviors of an overall structure. In addition to the new interpretation about the initial conditions for viscoelasticity problems, the present developments and several numerical simulations will provide insight into variety of engineering problems regarding to solid-fluid coupled systems.

View full abstract

Many physical systems give rise to large systems of simultaneous ordinary differential equations, the magnitudes of the eigenvalues of which vary greatly. The problem associated with such systems referred as stiff present the most serious numerical stability and accuracy problems, especially in the nonlinear systems.
A family of implicit two step time integration algorithms are presented in this paper. They are unconditionally stable and highly advanced for stiff systems where it is necessary to require accuracy for both large and small eigenvalues. Numerical examples relate to Navier-Stokes equations are shown to illustrate the features of the method. An ultimate basis for comparison is accuracy.

View full abstract

A stabilized finite element method based on the Q1/P0 element is presented for the analysis of incompressible viscous fluid flows. A presssure stabilization matrix (PSM) for the Q1/P0 element is introduced to avoid the occurrence of the checkerboard pressure mode. Also, SUPG formulation is employed to improve the numerical stability and accuracy. For the temporal discretization, a modified semi-implicit scheme (MSI) is applied. The pressure poisson equation system is solved by the Element-by-Element conjugate gradient method. The present method is shown to be an accurate and efficient tool for the analysis of incompressible viscous fluid flows.

View full abstract

In this paper, two and three dimensional numerical flow analyzes around a circular cylinder over wide range of the Reynolds numbers are performed. Characteristics about flow around a circular cylinder are grasped, and effectiveness of numerical flow analysis is investigated to compare calculation results with experimental results. Investigations about axial domain and layer division in case of three dimensional analysis are performed. Flow field is assumed to be incompressible viscous flow that is described by the incompressible Navier-Stokes equations. A finite element method based on implicit time integration and equal-order interpolation is applied to the discretization method.

View full abstract

In this paper, we propose an appropriate approach to appling large-eddy simulation to the flow around structures, in which high Re flow is turbulent and complicated. Because of the insufficient computational power to the engineering sense, it is difficult to get fine mesh to the all simulating domain and high shear flow with coarse grid causes numerical failure. We employ domain decomposition technique to set grids efficiently and, incorporate 3rd-order upwind scheme to avoid numerical failure with controlling excessive numerical dissipation by expanding filter width as well as reducing the coefficent of numerical viscosity term. We applied this methodology to the complicated flow past a square cylinder at Re=22, 000.

View full abstract

Wind tunnel experiments using circular tower elastic models were conducted to investigate the mechanism of the vortex-induced vibration at a high wind speed. Special attention was paid to the generation of the tip-associated vortices. In order to clarify the role of this vortex in conjunction with that vibration, the models with different free-end shapes were used. It was shown that the tip-associated vortices were the cause of that vibration and this consideration was consistent with that obtained by a previous experiment using a circular tower rocking model. Also, the reason why the tip-associated vortices had a frequency lower than the Karman vortex shedding frequency was discussed based on the result of a flow visualization.

View full abstract

This paper presents test results on lateral-torsional buckling of welded monosymmetric I-beams with various flange plate sizes. Ten different sections including doubly symmetric ones are tested under a concentrated load for a simply supported span. Four span lengths ranging from 1.5m to 3.75m are chosen for each of the sections. The effects of overall and flange local buckling on the rotation capacity and the ultimate strength are investigated for the beams with noncompact section. Test results are also compared with design stregnth curves which are evaluated by the Q-factor method and other approaches.

View full abstract

The three surface model for cyclic metal plasticity is presented primarily in view of the application to the analysis of thin-walled steel bridge piers under cyclic loading. This model is devised to include less material parameters and internal variables such that the model can be easily implemented in the FEM analysis. The present model is also characterized by the point that the two parameters independent of material types are calibrated by the cyclic loading tests for steel piers. As a result, the FEM analysis based on the three surface model can well predict the histeretic behavior of steel piers.

View full abstract

Seismic behavior of steel bridge piers is influenced mainly by the width-thickness ratio parameter of the flange plate and slenderness ratio parameter of both the column and stiffeners. In this study, a proposal based on the findings from available experimental results is made for designing steel bridge piers with high ductility, and its validity is verified by pseudodynanic tests. A key point of the proposed design method is to increase the deformation capacity by setting upper bounds to both the plate width-thickness ratio parameter and column slenderness ratio parameter, and a lower bound to the relative flexural regidity of the stiffeners. Steel bridge pier models designed in acordance with the proposal were tested using the earthquake accelerograrns recorded during the Hyogo-ken Nanbu earthquake. The results have shown that steel bridge pier specimens tested are of very high ductility capacity.

View full abstract

The seismic characteristics of steel bridge piers with the box cross section are substantially governed by the local buckling of steel stiffened plates and cracks caused by the low-cycle fatigue of welding parts around the base plates and at the corners of cross sections. In view of such facts, the use of low width-thickness ratio of plate panels and filling concrete into the steel cross sections may be useful to prevent the local buckling and a method for strengthening the base of pier columns with triangular ribs can be considered to be effective to prevent the cracks around the base plates. The effects of these methods are investigated in this paper by mean of the pseudo-dynamic tests as well as cyclic loading tests by increasing displacement gradually for 7 test specimens.

View full abstract

Regarding a total of 8 test specimens of steel pipe cantilever column of high strength steel (SM570Q) which parameters are diameter-thickness ratio and axial force ratio, a quasi-static cyclic loading test assuming seismic force was conducted to investigate the load carrying capacity and ductility.
We obtained load carrying capacity and plastic rate, which has a large diameter-thickness ratio parameter and on which experimental data are scarce. These results showed the same tendency as that of the existing experimental results obtained using ordinary steel.
Based on the experiment, as well as on test data accumlated by various institutions to date, we proposed a formula of load carrying capacity and plastic rate of steel pipe columns.

View full abstract

P75 steel rigid-frame pier in Kobe Harbor Highway was done damages proceeding large cracks to the corner under the Hyogoken-Nanbu Earthquake. This paper presents the consequences of observation of the fracture appearance, examination of the material property and calculation of the strain occurred under this earthquake. According to observation of the fracture appearance, the brittle appearance is recognized on the cracks proceeding from the toe of fillet weld. According to examination of the material property, it is found the toughness of the base metal and the deposited metal around the initiation of crack is deteriorated by the prestrain exceeding the yield strain. According to calculation of the strain using finite element analysis, it is presumed that the strain and the strain velocity on the corner are very high. From these results, it is concluded that the brittle fracture is occurred by the deterioration of the toughness of the base metal and the deposited metal, and the high strain velocity.

View full abstract

During 1995 Hyougoken-nanbu earthquake, IKUFA Bridge for railway were severely damaged and centrifugal cast steel bridge piers, cast steel shoes and reinforced concrete rigid frames were fractured. The bridge which was built in 1967, is constructed with 2 steel rigid frames and 2 reinforced concrete rigid frames with a steel girder of 50m span.
In order to make clear the fracture mechanism, dynamic response analysis, static elastic-plastic analysis, and material test were carried out. As a result, the fracture of steel bridge pier was brittle fracture due to high speed strain. The results of Analysis and tests are reported in this paper.

View full abstract

During Great Hanshin-Awazi Earthquake, brittle fracture was caused at beam-to-column connection of P75 steel pier. In this study, concerning the brittle fracture at P75 pier, simulations of brittle fracture at beam-to column connection is carried out by using large scale specimens simulating the P75 beam-to-column joint. In addition, as improve or retrofit details of beam-to column connection against such brittle fracture, some improved specimens which include ribs at corners of connection are also used. As a result, there is a high possibility that brittle fracture is caused at corner of connection where there is a high strain concentration. On the other hand, improve or retrofit details studied in this study are much effective.

View full abstract

The present paper deals with the effects of liquefaction-induced lateral flow on in-ground structures, focusing on the load characteristics of ground lateral flow. For this purpose, a shake table test, an upward seepage flow test and numerical analyses are conducted. The results from the tests and analyses demonstrate that a drag force from ground lateral flow is the essential effect if complete liquefaction occurs around in-ground structures. It is also shown that the drag force has an insignificant effect on in-ground structures when the size of liquefied area is considered to be relatively small comparing to the spatial extent of an in-ground structure.

View full abstract

An important factor in the analysis of the behavior of buried pipelines is the ground restraint force, which is generally considered to be proportional to the amount of one directional movement. However, large cyclic deformations during strong earthquakes drew attention to the fact that the effect of displacement velocity and cyclic loading have to be taken into account. A series of dynamic loading tests and dynamic direct shear tests has been conducted to examine the cyclic and velocity effect using a shaking table to produce seismic loadings.
The test showed the followings. Axial ground restraint force was generally independent of displacement velocity and cycles for the polyethylene coated steel pipes of 300mm and 150mm in diameter in the test.

View full abstract

The deformation of sand subjected to reverse fault displacements was investigated by experiments and numerical analysis. Experimental results could be simulated well by elasto-plastic finite element analysis with joint elements which model shear rupture zone. Calculating bedrock fault movement by dynamic model, earthquake fault rupture propagation in sandy deposit was analyzed numerically. The results suggest that earthquake fault rupture may propagate through deposit and reach the ground surface, if the bedrock fault vertical displacement exceeds several percents of the depth of the deposit.

View full abstract

This paper is concerned with the system identification method of vibrating structures by the virtual instrument software. The proposed identification method have two stages. The first stage is the signals processing by bandpass filter. In this stage, the signals of the multi-degree of freedom system reduce to the signals of the single-degree of freedom system. The second stage is the curve fitting by the nonlinear least square method. The dynamic characteristics of structure are estimated by the impulse response function of the single-degree of freedom system. For the bandpass filter process, the digital filter method and Fourier transform method are examined. As numerical simulations and experiments, it is shown that the proposed identification method enable to estimate highly degree of accuracy the dynamic characteristics for structures.

View full abstract

In this paper, the highway bridge vibration control under a moving vehicle is realized by the Linear Quadratic Gaussian (LQG) control theory. For control mechanism, the hybrid control system is used. The bridge deck slab is modeled by the finite strip method. The first order vibration up to the third order vibration are considered. The moving vehicle and the tuned mass damper (TMD) is severally idealized as a one degree of freedom system. The road roughness is presented by the stationary random process.
The Kalman filter can be designed for bridge-vehicle-TMD-road system and the optimal feedback gain are obtained by the optimal regulator theory. As numerical simulation, it is shown that the hybrid control used by the LQG control theory is effective for the bridge traffic vibration control.

View full abstract

This paper presents both experimental and analytical approaches for the resistance of the local deformation of the filled materials in check dam structures under horizontal impact loading. First, the impact test of filled sand model is carried out by using the horizontal impact test machine. The compaction effect and the effect of collision depth of filled sand are examined on the relationship between resistance and local deformations. The deformation process of filled sand in impact test is also observed by using the high speed video recorder. Second, the modified distinct element method in which the concept of stress transmitting ratio is introduced into the expanded contact range of elements. The simulation results by the proposed method are in good agreement with those of experiments for loose and dense compaction sand under the static and impact loadings, respectively.

View full abstract

Distinct Element Method (DEM) is applied to the simulation of evacuation behavior during disaster. Motion of each element can be compute by solving the equation of motion step by step. We proposed the algorithm in which an element can avoid collision and pass the other elements naturally.
The DEM parameters, such as spring constant and driving force for human body, are determined based on the experiments. We simulate the evacuation behavior for the explosion accident that occurred at the underground shopping center near Shizuoka Station in 1980. Based on the results, a new underground shopping center, which was reconstructed after the explosion, is safer than old one from a view point of disaster-prevention.

View full abstract

The volcanic disaster of Mt. Fugen are prolonged and enlarged more than four years. Disaster prevention officials of roads and railway recognized the danger of the mountain erupting and considered counterplans to protect their facilities. They also considered emergency and parmanent restoration plans after debris flows attacked their facilities. In this paper, emergency counterplan, damage and both temporary and permanent restoration plans of bridges of roads and railway to volcanic disaster of Mt. Fugen are reported and discussed.

View full abstract