A formulation of geometrical nonlinear problems of shells with particular emphasis on the treatment of finite rotations is presented. The finite rotations are deleted in the formulation by removing them and hence only infinitesimally small rotations appear in the formulation. The deformed shell surface is given with the help of interpolation functions by the position vectors and by normal and tangential vectors of the surface at the nodes. A few numerical examples on shallow spherical shells are presented to compare the numerical results with the experimental results.

View full abstract

The combined upper and lower bound analysis is used to predict the shear strength of reinforced concrete beams without web reinforcement. Theoretical considerations reveal that the ultimate shear strength of a beam with given material and sectional properties may be expressed as a single continuous function of its shear span to effective depth ratio. Numerical computations for the ultimate shear strength of beams with steel reinforcement tatios of 0.028, 0.0188 and 0.008 and shear spar to effective depth ratios of 1 to 8 are carried out. In addition, the contribution pf compression zone, aggregate interlock and dowel action to the shear capacity of the beam are determined and the failure modes of the beams are examined. The results are compared with available experimental data and the possibility of the use of the combined upper and lower bound analysis for other structural members is indicated.

View full abstract

The three dimensional and two dimensional fundamental solutions for the transient incompressible viscous flow are determined by solving the singular differential equations. which are obtained in the induction process of the integral equation, by means of the Fourier transform. The fundamental solutions determined in this paper have been proved to be practical in applying to a two dimensional problem by the boundary element method.

View full abstract

The Locked Coil Rope cables of cable-stayed bridges have reportedly undergone significant relaxation and the deflections of the girders have also increased noticeably year after year from a number of field measurements. In this paper, an attempt is made to account for such behavior analytically assuming that the cables follow the linear visco-elasticity law; whereas the girders and towers remain linearly elastic. Described herein are the determination of the visco-elastic constants, and numerical analyses performed on totally fifteen cable-stayed bridges by finite element method and the numerical Laplace inverse transformation; The results were compared with the measured data on two bridges.

View full abstract

A new axisymmetric joint element model for semi-analytical finite element analysis of axisymmetric bodies subjected to non-axisymmetric loading is proposed. The new joint element model for axisymmetric interface has three degrees of freedom: normal tangential and circumferential directions and describes debonding (normal direction), and slip (tangential and circumferential direction) on the interface. The new joint element model was incorporated into a finite element program and test-case analyses were performed. The results indicate that the new joint element model is useful in the semi-analytical finite element analysis for deeper understanding of stress distributions on the axisymmetric interface.

View full abstract

A general Lagrangian formulation of structural finite elements is presented, in the context of nonlinear continuum mechanics. The element characteristics are ‘degenerated’ from 3D field equations, using kinematic characteristics of the structural member. Consistent linearization is performed to establish a Newton-Raphson solution scheme. Numerical examples are tested and the results clearly indicate the effectiveness of the present formulation.

View full abstract

This paper deals with the in-plane ultimate strength of steel frames with the thin-walled box sections subjected to the vertical loads. Firstly, a numerical method for analyzing the “critical strength” of frames with the initial imperfections in elasto-plastic regions is developed on the basis of second order analysis by approximating the influences of local buckling of thin-walled column elements on the overall buckling collapse. Secondly, an experimental study is performed on the test specimens of five portal frames. Finally, the effective column lengths of frames are discussed and an approximate method to evaluate the critical strength of frames are proposed through experimental and theoretical studies.

View full abstract

Model studies are carried out to investigate the scattering behavior of elastic waves in a half space. The scatterers examined are a semi-circular surface irregularity and a circular cavity embedded in a half space. The detected waves are simulated by the boundary integral equation (BIE) method. The results show that the observed waves are slightly different from the simulated waves due to the resonance of transducer. In spectral analyses, therefore, we employ the linear system theory to eliminate the effect of transducer. The obtained frequency spectra reproduce an essential feature of the analytical results computed by the BIE method, except for their absolute values.

View full abstract

The combined finite element-transfer matrix method is applied to the linear and nonlinear problems of thin-walled members. The transfer matrix is derived from the tangent stiffness matrix used in the FEM. To deal with complex structures the transfer matrix for the substructure, into which thin-walled members is devided, is introduced. In this method, procedures used in the FEM based on load increment are employed except for the estimation of approximate displacements for each specified increment load. Approximate displacement are estimated by the transfer matrix procedure. Some numerical examples presented in this paper show that for long thin-walled members, this method can be successfully applied to the linear and nonlinear problems by reducing the size of the matrix relative to less than that obtained by the FEM.

View full abstract

Based on the control of input energy distributed to each story in multi-mass systems, the relation between input energy and structural characteristics is examined fundamentally to establish the rational earthquake resistant design. As the results of numerical analyses, the plastic response of shear multi-mass systems can be treated as the linearized systems in apparent, in the case of equal stiffness distribution between upper and lower story. The theoretical formulation based on the modal analysis is developed to the elasto-plastic systems and the method of input energy control is discussed. The optimum distribution of input energy and stiffness in the earthquake resistant design are also presented.

View full abstract

A computationally efficient optimum design method is proposed. Unlike conventional optimization methods, all the optimized properties are obtained without conducting iterations. In this method, redundant forces of a statically indeterminate structure are computed from the condition of geometrical compatibilities, the members are fabricated so as to introduce prestresses which nullify the incompatibilities. Such an introduction of prestressing is compatible with conventional erecting process of cable-stayed bridges.
A series of analyses have verified that this method is practical and makes it easier to determine economical structural types of cable-stayed bridges.

View full abstract

An advanced method for a qualitative description of bifurcation buckling behavior of dome structures is proposed. While conventional finite displacement analysis technique is employed to trace the bifurcation behavior, a group theoretic method is adopted to describe the hierarchal structure of bifurcation paths.
The bifurcation behavior of polygonal-shaped truss domes is proved to be analogous to the subgroup structures of dihedral groups, which are extensively employed to describe the symmetry of polygons in mathematics. The categorization and description of the bifurcation behavior of a hexagonal-shaped truss dome structure insure the applicabilit and validity of the analogy in describing the bifurcation behavior. Such an analogy is able to monitor the interrelationship between the geometrical properties of the polygonal-shaped domes and their bifurcation behavior.

View full abstract

In modern civil engineering structures which tend to be light and slender, the problem of aerodynamic stability in wind has gained increasing importance. Long-span suspension bridges are typical among such structures. If streamlined box girder is used as stiffening frame, its small dead load can be advantageous in view of economy, compared with the use of stiffening truss. However, it is needless to say that various kinds of wind-induced vibrations shall be also examined in this case.
In the present paper, the effects of dead weight on the aerodynamic stability of suspension bridges stiffened by streamlined box girder are investigated, and it was found that the increase of dead weight could improve the resistance of the structure against wind, without noticeable reduction of natural frequencies. Furthermore, based on sectional model wind tunnel tests, the effects of suppressing wind-induced vibrations by dead weight are examined. Finally, for model suspension bridges preliminary designed, the effects of dead weight are concretely examined to obtain information on the application to design for aerodynamic stability.

View full abstract

This paper reports the in-plane “critical strength” representing the buckling or ultimate strength of thin-walled steel frames subjected to the combined actions of vertical and horizontal loads. To investigate the critical strength of steel frames, an experimental study is performed by using seven portal frame models with thin-walled box cross-section. The critical strengths of portal frames are also inquired on the basis of elasto-plastic and second order analysis including the approximate method for evaluating the local buckling of thin-walled column sections. Through these results, the lower bound of critical strength of frames is clarified in detail. Finally, two approximate methods for estimating critical strength of thin-walled steel frames are proposed herein.

View full abstract

In this report, discussions on the numerical calculation of the non-linear equilibrium equation are stated, Especially, the problems arising in the tracing calculation of the path that exists the conjugate path in the vicinity of the branch point and that changes the direction due to yielding of the material are made clear and a more generalized incremental method elaborated for their disposal is proposed.
Consequently, by adopting a spherical search surface in which the arc-length of the path is used as restrictive parameter, the equilibrium points can be obtained automatically with good intervals for the system in which the path makes an abrupt change, and furthermore, it is made clear that the elimination method proposal by the authors is highly effective for the systems that have small initial imperfection.

View full abstract

A finite displacemtment theory is developed for an arbitrary plane curved Timoshenko beam, in which an elastic constitutive relation is defined not by tensor components of stress and strain but by other physical components. This selection of components makes the governing equation simple and easy to handle. The nonlinear stiffness equation, with the use of nodal positions and the appropriate selection of local coordinates, is formulated for an elastic plane straight beam element. Three numerical examples of plane beam problems, involving geometrical nonlinearity, are employed as illustrative examples.

View full abstract

End plates have been used for two-dimensional aerostatic and aerodynamic wind tunnel tests. Although it is often discussed that there exists the optimum size of end plates to create a two-dimensional air flow on testing body in wind tunnel tests, the mechanism of end plate effects is not yet made clear in detail. This paper discusses the mechanism of end plate effects from various view points by measuring drag forces, base pressure distributions, wind velocity distributions around the testing body and flow visualization by hydrogen bubble techniques for the flat plate normal to flow. It is concluded that the diameter required for the circular end plates has the relation with the wake vortex formation region and it is larger than 8 times of the testing body depth normal to flow.

View full abstract

The effects of stress ratios on the fatigue crack growth in a high tensile residual stress field are examined. Specimens, made of 600MPa class steel, are center-cracked type which contain longitudinal weld beads along the center of the plate. Fatigue crack growth rates and crack opening levels are measured at several stress ratios (-1 to 0.91). Under these test conditions, cracks fully open regardless of the value of stress ratio, and fatigue crack growth rates and threshold stress intensity factor ranges do not depend on the stress ratios.

View full abstract

The dynamics of a grouped pile foundation in layered solis are investigated from the distributed parameter system modeling with a due consideration of the pile-soil-pile interaction. The substructure concept is applied to evaluate dynamic soil stiffness. The soil-pile coupling motion is solved for the base input through the eigenvalue decomposition and the transfer matrix method, deriving the pile head impedance function and the associated driving force (effective input). An efficient simplified mehod is also developed based on the “ring-pile” idea. From a numerical example, special remarks are given on the dynamic pile grouping effect.

View full abstract

This paper presents a FEM application, with use of the substructuing concept, to the pseudo-3-dimensional dynamic analysis of soil-pile foundation-structure interaction. The grouped piles are assumed to be in a concentric arrangement and are dealt with, based on the ring-pile modeling, by adopting the n=0 and 1 Fourier harmonics for the response in circumferential direction. Taking a bridge pier on a grouped pile foundation, as an example, the investigation is addressed to the pile head impedance functions and the associated input forces due to pile-soil-pile interaction for the base motion, and the inertial interaction between foundation and structure.

View full abstract

The tensile restraint stress and restraint coefficient of fillet welded joints tackwelded on box girder sections are analyzed by the displacement by spontaneous expansion from weld heat input on diaphragm webs or ribs, and rigidity ratio of the diaphragm web and flange or rib and flange. The results are sumarized as follows: (1) The restraint coefficient (R_F0) is directly proportional to the load distribution factor determined by the rigidity ratio of diaphragm web and flange or rib and flange. (2) The restraint coefficient (R_F0) of a diaphragm web is almost constant regardress of the position, but the restraint coefficient (R_F0) of a rib at the end of a welded line is about 3 times that at the center of a weld line. (3) The tensile restraint stress of a rib at the end of a weld line sometimes exceeds 70kgf/mm².

View full abstract

In this study, the reliability of continuous two-span bridge girder subjected to random vehicular loads is analyzed in the space domain as well as in the time domain. First, by the use of the results of the statistical analysis of randomly space-varying vehicular loads which have been already reported by the authors, the probabilistic characteristics of bending moment and the reliability of bridge girder in the space domain are analytically evaluated. Second, the distribution of the maximum bending moments for the lifetime domain are evaluated. The design values of bending moment are calculated for prescribed values of failure probability of girder bridge.

View full abstract

This paper presents a design method of the web slenderness ratio and the corresponding rigidity of longitudinal stiffeners in the curved plate girders. The elastic non-linear behaviors of the curved web panels with/without longitudinal stiffeners subjected to pure bending are inquired by means of analysis using the up-dated lagrangian and an isoparametric finite element method. Based upon these analysis, a required slenderness ratio of web panels is investigated by setting a limit state concerning the maximum out-of-plane deflections and bending stresses. In order to evaluate the strength of the longitudinal stiffeners in the ultimate limit state of curved web panel, a curved beam-column model is proposed, then a design method for the required relative stiffness of the stiffeners is, moreover, recomended by adopting interaction curve of this beam-column model.

View full abstract

A parametric study based on the plate approach and modified column approach including local buckling is presented for evaluating the ultimate strength of stiffened steel plates in compression, in which the effect of panel number and stiffener's rigidity on the ultimate strength is fully considered.
As a result, two ultimate strength curves are demonstrated according to the number of panels, and the approximate ultimate strength curves presented herein can afford to evaluate the use of stiffeners with rigidity more than required.

View full abstract

Inelastic lateral-torsional buckling strength of steel beams depends markedly on the location and extent of high moment regions causing the reductions in sectional regidities.
Rolled I-sections with two different sizes are tested carefully under a concentrated load at the central point and two equal concentrated loads at the one third points of simply supported beams, respectively. The test results of load-deformation curves and strength variations are compared with both loading conditions. Appropriate design strength formulas for the various load patterns of beams are also discussed by a review of the available experimental data on the ultimate strength of stocky beams.

View full abstract

This paper describes a extended engineering beam theory that has been developed for the analysis of cellular box girder bridges in the linear elastic range. The solution includes the distortion of the cross section and the influence of shear deformation. The numerical analysis is based on the modified transfer matrix method, which enables to analyse long span beams. As a numerical example a four span continuous concrete box girder bridge with a cross section of four cells and of variable web-height was treated.

View full abstract

The application of the finite element method to structural analysis generally requires to solve a large and sparse set of linear equations, but it is well known that the reliability of the solution becomes low as the increase of the dimension of the linear system. The purpose of this investigation is to show how to prevent the development of the numerical error according to the elimination procedure. Through the theoretical investigation we show that the numerical error is governed by the conditioning number of the governing equation and also of matrices appearing during the elimination procedure. Successive experimental study concludes that the appropriate elimination ordering may decrease the error by an order of magnitude. Some other important informations on the elimination ordering are also given in this paper.

View full abstract

The EK-WGI method was effectively applied to identification problems of a running load and beam system. By suitably formulating the state vector equation and the observation equation of the system, accurate estimation on the dynamic properties of both the running load and beam was obtained, where numerically simulated data were used as observation waves in order to examine the accuracy of results.

View full abstract

Fatigue tests were carried out on nine center notched specimens and six gusset attached specimens after an overload was applied. Fatigue life of these specimens were always longer than the average fatigue life of no overloaded specimens. Stress distribution near the notch or the gusset end revealed that there existed a point where the effective stress in tension became lower due to the compressive residual stress induced by the preoverloading. This stress can be related with the threshold stress intensity factor range and the improvement of the fatigue life due to preoverloading was partly explained. Fatigue crack propagation life was also computed using the effective stress range concept.

View full abstract

In these days of stable growth of economy in Japan, proper operation and maintenance of existing structures which have been used for many years have become an important subject. Furthermore, making improvements in the study of this problem, it is pointed out that replacement plans of the structures based on life term prediction should be applied. However, as compared with operation and maintenance, life term prediction is much more difficult because of necessity of data base on material changes with the passage of time.
The purpose of this paper is to describe the investigation results of life term prediction techniques focusing on hydropower steel structures. They are composed of simplified data acquisition and structural analysis, data base, expert system and others. Especially expert system plays an important role among them.

View full abstract

This paper presents a procedure of nonstationary spectrum analysis by using complex demodulation with which both accurate evolutionary spectrum and nonstationary phase variation can be directly estimated. Furthermore, the record of an actual strong seismic motion, that is, the N-S component of Muroran acceleration record (1968) is analysed to demonstrate the applicability of the present analytical method. The set of results indicates that the method of complex demodulation is an effective and reliable method of evaluating evolutionary power spectrum and phase of a strong earthquake motion.

View full abstract

It is indicated that the stiffness of bedrock under foundations evaluated analytically by FEM tends to be larger than that evaluated from in-situ vibration tests performed on the concrete block foundation. One reason may be considered to be owing to lowering of rigidity in this surface layer, another owing to imbalance of rigidity comming out between both parts of tension and compression sides in bedrock during rocking vibration, in extreme case, being followed by uplift. In this research some vibration tests are carried out with structural models on ground ones of silicon rubber, and the phenomena mentioned above and its characteristics are clarified. At the same time numerical simulations are carried out with the computer program developed here which can simulate nonlinear seismic response followed by sliding and separation. Accuracy of above numerical procedure is confirmed comparing the numerical results with experimental ones.

View full abstract

In this study, we investigate how the vertical distribution of shear wave velocity has effects on the shape of transfer function of subsurface layer. The effective coefficients for transfer function and its maximum peak value are defined on the basis of partial derivatives of transfer function regarding to shear wave velocity and thickness of a layer. The effective coefficients are calculated to several subsurface layer models in Japan. It is concluded that the effective coefficients of the shear wave velocity of upper layer and base rock are generally larger than those of the medium layers. However, in the subsurface layer model which has relatively large shear wave velocity and large thickness in medium layers, it is found that the effective coefficients in the medium layers are considerably large.

View full abstract

This paper presents a method of response analysis of ultra-high-voltage towers subjected to earthquakes, taking into consideration the interaction between the towers and the transmission lines.
Seismic responses of them are solved by the method of modal analysis using individual natural vibration characteristics of each tower and transmission line. This method has the advantages as follows: 1) The degrees of freedom remarkably decrease, as compared with those in the method applying the F. E. M. to the towers and the transmission lines. 2) This method can be applied, without difficulty, to the case where seismic waves differ at each foundation of tower and to the case where suspension towers are included. 3) The results obtained by this method are more accurate than those obtained by the method regarding the transmission lines as springs connecting the towers.

View full abstract

A numerical technique for determining a dynamic response factor, such as the impact factor of highway bridge, concerning the safety level of a stationarily randomly excited structure is presented by the use of the first-passage probability theory. From that aspect, the dynamic response factor is apparently different from what is defined only by the mean of the largest value of the random response. The effectiveness and versatility of the theory are demonstrated by its practical application to the determination of the gust response factor for a flexible steel tower exposed to a strong natural wind.

View full abstract

To repair lifeline systems, especially transportation systems, immediately after the seismic disaster, is very important for the whole restoration process. This paper presents and evaluates the optimum restoration process for damaged transportation systems. The theoretical optimum restoration process is calculated by using the avarage running time of all cars in the transportation network. And the procedure is applyed to the real network system of Izu-peninsula, which was damaged by Izu-Ohshima-kinkai earthquake (1978). The relationship between the characteristics of the roads and the proposed optimum restoration process is also analyzed, using Hayashi's quantification scaling (type 2).

View full abstract

An effective stress method is presented for analysis of seismic response and liquefaction of irregular ground including soil-structure interaction, based on an implicit-explicit finite element method. A pore water pressure is computed with iteration from the total stress considering undrained condition. The simulated pore water pressure is in reasonably good agreement with experimental data. The proposed method of analysis is compared with other well-known methods for one-dimensional model, which is in good agreement. The present effective stress method is also applied to liquefaction problem involving two-dimensional soil-structure model. The numerical results are considered to be significant from the viewpoint of earthquake engineering.

View full abstract

In this study, a duration of earthquake ground motion was defined by the standard deviation of relative frequency distribution of Fourier phase differences. It was examined from comparison of the motions on surface ground with ones on base rock how the duration was influenced by soil condition. Then, the regression analysis was carried out to investigate the dependence of the duration to magnitude, epicentral distance and difference of datasets. It was concluded that present duration was not influenced by soil condition so much and the duration of motion on surface ground was equivalent to that of base rock. The difference of datasets had a large influence to the results of regression analysis for the duration, so the attention should be paid when a statistical analysis was carried out.

View full abstract

In this study, the effect of nonlinear boundary conditions at liquid surface on sloshing height of cylindrical tank under vertical and horizontal ground motions is discussed in comparison to linear analysis. It is found that this effect mainly depends on both dimensionless sloshing height and dimensionless liquid depth as far as smaller sloshing height is concerned, and simple formula estimating this effect is proposed. On the level of observed large sloshing height due to the 1983 Nihonkai-chubu earthquake, sloshing height calculated under nonlinear conditions are estimated about 10-25% larger than those under linear conditions. Taking into account this effect, response spectra of long period ground motions deduced from observed sloshing height are found nearly equal to two-dimensional response spectra calculated from strong motion seismograms.

View full abstract

In this paper, an attempt is made to formulate the optimum aseismic design using the concept of fuzzy mathematical programming. While the fuzzy mathematical programming can be applied to optimum aseismic design, it is necessary to consider carefully the aggregation of contradictory components for practical purposes. To find an adequate way of aggregating objective functions, we conducted a survey, during which several sets of questionnaires were distributed among experienced engineers to find their preference in the design of earthquake-resistant structures. Results of these questionnaires are collected and analyzed by using the pairwise comparison method. By using several numerical examples, the present method is illustrated and compared with the previous results using the minimum operator.

View full abstract

In this study, numerical analysis was carried out for SDOF systems with bilinear hysteretic properties by use of six earthquake records. On the basis of the average value of six earthquakes, we discussed the effect of parameters such as natural period T, damping factor h, plastic stiffness p and yield level R on hysteretic energy E_hp, input energy E_ip and the ratio of E_hp to E_ip, in combination with the inspection of previous works. We also examined the effect of these parameters on the ratio of the equivalent velocity obtained from the square root of 2 E_hp to the elastic spectral velocity V_emax and the ratio of E_hp to the elastic input energy E_ie and the regression equations of these ratios were derived regarding the four parameters T, h, p and R. It was found that there was a close relation between E_hp and both V_emax and E_ie.

View full abstract

Recent fatigue cracking on highway and railroad bridges urged the study on appropriate methods to repair and to reinforce the damaged structures. Fatigue tests were carried out on tensile plates: a) with stop-holes as crack arrester, b) with groove-welded repair at the center c) with stop-holes reinforced with high strength bolted splices, and d) with stop-holes reinforced with fillet welded splices.
The stop-holes were appropriate way for a temporary crack arrester, when the cracks were small. The groove welded repair showed wide range of fatigue life depending upon the weld quality. Relatively large weld defects were observed, when the groove welds were made by an unexperienced welder on vertical position. Short fatigue life was observed on the fillet welded splices, while high strength bolted repair showed longest fatigue life after repair.

View full abstract

This paper describes experimental studies on
(1) the repair works of the fillet welds between web plate and bearing stiffener plates,
(2) the repair works of the fillet welds between web plate and upper flange plate, and
(3) the reshape works from uniform plate girder to non-uniform one.
In these experiments, heating works such as gouging, welding, gas cutting and correction using gas were done under loading in order to study the safety during works, and ultimate strength after heating works was also obtained. Through these experiments, possibility and cares of repair or reshape works with heating works were investigated.

View full abstract

A semianalytical method is developed to study the free in-plane vibration of arches with any shape. The fundamental differential equations are first translated into the integral equations. By applying the approximate solution of integral equation, semianalytical solutions of the original differential equations are obtained. The solutions have discrete type expressions concerned with the equally spaced points of the arch axis or the points on the arch axis corresponding to the equally spaced points of the horizontal distance between the supports. The method is applied to analyze the free in-plane vibration of arches with nonsymmetrical axis because of the unequal height of the supports. Numerical results for the cases of nonsymmetrical, 2-hinged and fixed arches with parabolical axis are compared with the solution of the symmetric arches.

View full abstract