The group-representation theory guarantees that the (tangent) stiffness matrix of symmetric structures can be put into a block-diagonal form by means of a suitable (local) geometric transformation. This transformation decomposes the linear equilibrium equation of symmetric structures into a number of independent equations, and hence is advantageous for parallel analysis. The block-diagonalization method, which so far has mainly been applied for translational displacements, is extended here to rotational ones. The interrelationship between the symmetries of rotational and translational displacements is investigated by means of group theory to arrive at the transformation matrix of rotational ones.

View full abstract

This study investigates ultimate stability strength behaviors of cylindrical steel shell panels under various types of uniaxial compression using the non-linear finite element procedure for the ultimate strength determination of shell-type-plate. The effects of residual stress and initial geometric imperfection are all included in the analysis. The numerical calculation method is verified and compared with previous theoretical works. Based on the investigation results, a practical formulation for the ultimate strength is also proposed.

View full abstract

This paper presents a unified prediction method to the ultimate strength for rectangular stiffened plates under biaxial in-plane forces in the elasto-plastic range. For a given strain ratio in two biaxial directions, the elasto-plastic buckling strength is obtained from the elasto-plastic material behavior and the residual stresses in two directions. Then the elasto-plastic ultimate strength is predicted in terms of the imperfection sensitivity curve in the neighborhood of the particular point, where the elasto-plastic post-buckling curve intersects the pathological failure mechanism curve. The ultimate strength is found to be in good agreement with the previous numerical results.

View full abstract

This paper proposes new seismic hazard maps and seismic macrozonation for the Philippines based on earthquake occurrence data. The completeness of two earthquake catalogues is examined and the time periods in which the data are complete are identified for each magnitude level. For the maximum use of the available data, the data of large earthquakes are taken from the long time period while the data of small earthquakes are taken from the shorter time period in which they are complete. The seismic design provisions of the Philippines are compared with those of Japan and are found to be substantially lower. To check the significance of this, the seismic hazard of Japan is also analyzed and compared with the Philippines. Based on the distribution of the 100-year return period peak ground acceleration, a new seismic zonation map of the Philippines is developed.

View full abstract

Fatigue crack growth behavior under single-peak overloads is investigated for structural steel JIS SM520B. Fatigue test is carried out on seven center prenotched specimens. Three parameters, i. e. overload ratio, overload application stage and stress ratio are introduced to investigate the retardation effect. It is verified that fatigue crack growth rate is retarded after overloading and then gradually recovers to the normal. It is also observed that the retardation effect is intensified when the overload ratio and the stress range between overload stress and the following minimum stress increase. From the test results the crack length increment in which retardation occurs is compared with Irwin's and Dugdale's plastic zone sizes.

View full abstract

A finite element method for analyzing transient incompressible creeping flows is presented. Marker particles are used to represent free surfaces and to visualize flow patterns. A six-node triangular isoparametric (second-order) element is used for the analysis. Marker positions are determined by the use of triangular area co-ordinate system. A simple solution algorithm is obtained and a higher order accuracy, which is difficult with the finite difference method, can be expected with the present method. To verify the scheme, a simple punch indentation problem is solved.

View full abstract

Using a Fourier transform and coordinate transform, the space-time cross-spectral density function, which is often used to represent spatially varying seismic motion, is derived from a general frequency-wavenumber power spectrum model. Although the existence of a dominant wave component is not assumed, the resultant cross-spectrum model becomes the same as the one proposed by Loh. Hence the Loh's model is found to be applicable to express general stochastic wave fields.

View full abstract

A basic analysis is carried out on the identification, prediction and control on a single degree of freedom system. First, the responses of the system excited by an active control device installed on the system are effectively utilized to identify the dynamic properties of the system which is modeled by a multi variate ARMA model. Then, general modes of a instantaneous optimal prediction control rule are formulated in terms of the identified components of the coefficient matrix ofthe ARMA model and the weights included in the control objective function. Based onthe formulation, a neural network is derived whose links have physically meaningful weights.

View full abstract

This paper presents an application of the spline prism method which is a three-dimensional semi-numerical method to analyse bending of thick rectangular plates subjected to surface loadings. The effects of thickness ratios, b/a, aspect ratios, L/a and bonudary conditions on the behaviors of stress distributions and displacements along the thickness of thick rectangular plates have been analysed.
It is seen that good convergence and accuracy are obtained. The stresses along the thickness show nonlinear and unsymmetrical distributions by the increment of the plate thickness and by the surface loading. High stress concentration is also come out near the clamped edges.

View full abstract

A boundary element method to nonlinear analysis of free-surface motion in a two-dimensional container is developed based on the Lagrangian particle description. By using the weighted residual formulation to the relation between the velocity of the fluid particle and the gradient of the potential on the surface, a unifying formulation is obtained for the relation between the kinematic condition and the boundary values. A time integration scheme is constructed in accordance with the Newmark method. The proposed method with linear element provides second-order convergence in both space and time directions. Through numerical experiments, influence of the treatment of nonlinear equations on the accuracy is discussed.

View full abstract

Change in fatigue strength of unpainted welded joints of weathering steel and structural steel due to long term atmospheric exposure is one of the essential factor in using the weathering steel for bridges. Fatigue tests of 10-year weathered 19 stiffener specimens and 9 gusset specimens were carried out. Additional 11 non-weathered gusset specimens were tested for comparison. The fatigue test results were compared with previous fatigue test data. It was found that no reduction in fatigue life was observed even after 10-year of weathering, comparing with the non-weathered, 2-year and 4-year weathered specimen. The runout levels of fatigue strength were also increased, because more data became runout even at higher stress range levels. Measurement of weld toe profiles indicated that the gain in fatigue strength is due to the increase in weld toe radius.

View full abstract

A new stability design method for frames is presented. Contrary to the effective buckling length concept, the proposed method is based on the elastic second-order analysis. In this method, the introduction of equivalent initial deflections corresponding to the imperfections is of utmost importance. Therefore, a new equivalent deflection formula using curvature is suggested. Moreover, a methodology is proposed to explain the systematic application of this formula to a frame with stepped cross sections. The validity of the metod is demonstrated by some examples.

View full abstract

This paper shows an aproach for analysing color hamony of a bridge scenery image obtained by CG technique. The CG image of bridge scenery in full-color format RGB color model is translated into Munsell color name (HV/C) format, then, Moon-Spenser's color harmony theory is applied to the translated CG image. Thus quantitative color harmonic index M of bridge scenery is calculated. The result of color harmonic analysis is compared with the result of monitering test, and they show good agreement.

View full abstract

Sandwich steel plate, composed of two structural steel plates with synthetic resin in the middle, can be applied to bridge members to reduce the structural borne (noise) problem. In this study fatigue tests were carried out on five welded joints of sandwich steel plate of 10mm thickness. Six specimens of each type were fatigue tested and fracture mechanics analysis was carried out based on the observations of fatigue tests. It was found that almost the same fatigue strength was obtained for the sandwich steel and the normal steel, as long as weld quality of two types of steels was the same.

View full abstract

Appropriate evaluation of effective length is important in the buckling design of steel structures. Recently, eigenvalue analyses of overall structures have frequently been used as a method for the calculation of effective length. However, depending on the type of structure, the effective length obtained by this method does not produce rational values, and this sometimes presents a problem. Some methods have been proposed for the modification of this analysis. In this study, the effectiveness of this modified procedure is evaluated by means of numerical evaluations. Discussion on the application of eigenvalue analyses to an overall structure is presented, and a rational method for the calculation of effective length in practical designing is proposed.

View full abstract

Fatigue damages around shoe of plate girder bridges were investigated focusing on the two causes, the deterioration of movability of shoe and the gap between sole plate and lower flange of main girder. The results of fatigue test and FEM analysis showed that these two causes raised high stress concentration around shoe leading to fatigue crackings. Five types of improved details in order to prevent fatigue damage around shoe were tested and their effectiveness were discussed.

View full abstract

The objective of this study is to develop a stochastic interpolation method of conditional simulation on a temporal and spatial gaussian field. The major contributions of this paper makes use of simple auto-regressive modeling in the time domain which expresses the very property of spatially and temporally deviating phenomena such as an earthquake wave propagation, and the development of an effective method in terms of computer time.

View full abstract

Records of strong motion are often used as the ground motion inputs in earthquake analysis and in the design of structures. The design of underground structures, however, requires more than just the time history at any particular point on the earth's surface. It also requires the space-time variation of the ground motion. We have therefore developed a method for generating a simulated space-time variation that includes strong ground motions recorded at observation points. This simulated variation is characterized by an apparent velocity and by a coherence function. Here we presented a case study using motion recorded during the 1940 Imperial Valley earthquake.

View full abstract

The elucidation of the issue of waves propagating through the ground is very important in the area of seismic engineering. This requires the further development of a model experimental method effective for the clarification of a phenomenon of wave propagation. In the past, the use of low-elastic material to make a ground model made it possible to reduce the velocity of propagating waves, resulting in an easier execution of experiments. This paper discusses that in addition to an analogic method of recording used in the past, we introduced a digital method of recording to analyze quickly and accurately the phenomenon of waves propagating through a model ground. An example is given to prove the validity of the proposed method through investigation of the phenomenon of waves propagating through the ground in case there is a trench in an elastic layer lying over a rigid substratum.

View full abstract

This paper which concerns the dynamic analysis of ground pile foundations places interest not only on evaluating the pile head impedance function and the effective pile head input motion but also on investigating the pile behavior during soil-structure interaction. The thin layer formulation is applied for soil analysis. The transfer matrix procedure is used for pile analysis. The flexibility approach that takes the interaction force as unknown is adopted for establishing the coupled behavior under the perfect bond condition. For group piles, the rigorous 3-dimensional analysis and the ring pile approximation are made for comparison. Soil reaction and internal forces of pile are investigated and the contributions from the inertia interaction and the kinematic interaction are clarified. The validity of the Chang's formula is checked from the design point of view of piles.

View full abstract

Records of strong motion are often used as the ground motion inputs in earthquake analysis and in the design of structures. The design of underground structures, however, requires more than just the time history at any particular point on the earth's surface. It also requires the space-time variation of the ground motion. We have therefore developed a multivariable autoregressive method for generating a simulated space-time variation that includes strong ground motions recorded at observation points. This simulated variation is characterized by an apparent velocity and by a coherence function. Here we presented a case study uisng motion recorded during the 1940 Imperial Valley earthquake.

View full abstract

The prediction of the fundamental period and damping ratio of bridge pier-foundation system is of importance in seismic design of bridges. An analytical approach, based on linear simple superstructure of mass and height supported by a rigid embedded foundation with mass and mass moment of inertia having two degree of freedom with swaying and rocking motions, has been described. Although the model used in this study is simple, a concept of the equivalent complex spring and the equivalent input motions of soil-foundation and pier systems has been presented. This concept makes it possible to assess the dynamic soil-structure interaction effect on seismic responses of bridges and to obtain closed form formulas for the fundamental period and damping ratio of bridge pier-foundation system.

View full abstract

The prediction of the fundamental period and damping ratio of bridge pier-foundation system is of importance in seismic design of bridges. An analytical approach, based on linear simple superstructure of mass and height supported by a rigid embedded foundation with mass and mass moment of inertia having two degree of freedom with swaying and rocking motions, has been described. Although the model used in this study is simple, a concept of the equivalent complex spring and the equivalent input motions of soil-foundation and pier system has been presented. This concept makes it possible to assess the dynamic soil-structure interaction effect on seismic responses of bridges and to obtain a simple procedure of predicting the fundamental period and damping ratio of continuous girder bridge pier-foundation system, in conjunction with the virtual displacement method. The validity of the proposed method is examined by the dynamic analysis of multi-degree-of-freedom system.

View full abstract

Based on a convolution type Green function derived from distributed force application to the 2-dimensional linear elastic body, the transient wave field in the dynamic soil-structure interaction is thoroughly investigated by the time-domain BEM, when the soil is modeled either by a halfspace or by a stratum. The Rayleigh wave generation for the former case and the eigenmode wave propagation for the latter case are clarified. Depending on the period ratio between Ricker wave type loading and the natural period of stratum, the resonance/wave impediment phenomenon of the stratum soil in contrast to the halfspace, is focused. Further, a flat buried block system (called WIB) below the foundation proved to be an effective wave impeding measure, provided that it is installed at a proper depth over a sufficient width.

View full abstract

Based on recent studies suggesting that Rayleigh waves are observable in microtremor ground motions, an improved procedure is proposed to estimate a natural period of a site using amplitude ratios between horizontal and vertical motions of the mictrotremors. The procedure is to make the best use of not only peaks but also troughs in a spectral amplitude ratio. Using microtremor data measured at all junior high schools in 23 wards of Tokyo, the proposed procedure is applied to a period-mapping of the area and compared with an existing map to a satisfactory level of accuracy, demonstrating its practicality.

View full abstract

A model ground of air dried sand compacted in a shear testing apparatus and two composite models each of which had the model pit of different rigidity buried in same model ground as above one were excited horizontally with sinusoidal motions on a shaking table. From measured acceleration and dynamic earth pressure it has been found that the vertical vibration is produced in the ground due to the dilatancy or volume change of soil and yet it produces the dynamic earth pressure acting on all outside walls of the model pit. The vertical vibration is more strongly influenced on than the horizontal one by the dynamic soil-structure interaction. In addition, we show that the vertical vibration produced by horizontal exciting is able to be predicted by finite element procedure with elasto-plastic dynamic response analysis.

View full abstract

The Higashi-Kobe Bridge is a three-span continuous cable-stayed steel bridge adopting a Warren truss as its main girder. The main span is 485 meters, which is one of the longest spans of this type in the world. An experimental study was carried to ascertain the rigidity and stress concentration around the gusset plate of truss panel points designed to the existing provisions. Through the experiments, the folloing results were obtained; (1) the rigidity of panel points satisfies the design requirement, (2) the fatigue strength corresponding to a hot spot stress in the gusset plate is not sufficient under dynamic vehicle loading. Based on the experimental results, we improved a design of structural. details in truss panel points in order to reduce the stress concentration.

View full abstract

The performance of gradient methods such as Gauss-Newton method. Davidon-Fletcher-Powell method and modified Successive Linear Programming is compared for the solution of identification of dynamic soil properties. The identification problem is formulated on the basis of multiple reflection and refraction theory. Through numerical analysis, we showed that modified Successive Linear Programming is as good as Gauss-Newton method on the speed of convergence of the solutions, and is superior to others on the property of gloval convergence.

View full abstract