This paper presents a total Lagrangian nonlinear formulation of elastic trusses, in which the governing stiffness equations are described as the relations between the overall forces and positions. With this selection of spatial positions as basic unknowns, the specification of the initial configuration becomes unnecessary and the separation of rigid body motion is automatically attained by an appropriate selection of local coordinates. A simple two bar truss and a reticulated truss are investigated as numerical examples. In the former, the characteristic of the present formulation and the convergence by the successive substitutions have been demonstrated. The latter example is chosen to show the effectiveness of the present formulation and simple systematic procedure to trace the finite displacement equilibrium paths including the main path and paths after bifurcations.

View full abstract

This paper developes the strict theory on the lateral buckling of a through plate girder bridge, and simple approximate formulae for critical loads have been proposed. The validity of this theory has been examined by comparing with some experimental results, and the elastic lateral buckling loads and the load carrying capacities for the illustrative practical bridge models have been calculated.
The results have shown that the approximate formulae for critical values of symmetric and asymmetric elastic buckling under uniformly distributed load have to be useful, and that the load carrying capacity for long spans would depend on asymmetric overall lateral buckling. The symmetric buckling will be not so significant.

View full abstract

The method with the separation of rigid body displacements is a powerful tool for the finite displacement analysis of structures particularly with curved members, with variable cross sections, and/or with inelastic behaviors. Nevertheless, its theoretical equivalence to the solutions of the direct Lagrangian equations has not been examined so far.
This paper presents the theoretical convergence and accuracy of the method, where an original curved element of variable cress section is considered as well as an approximate straight element of constant cross section.

View full abstract

This paper is concerned with an interpretation and analysis on the elasto-plastic lateral torsional buckling of beams in a simplified manner by means of catastrophe theory. The numerical analysis does not require having recourse to the ordinary nonlinear numerical procedures, but can be performed using a relatively small microcomputer.
The proposed analysis takes into account the elasto-plastic behavior of the material, the geometric imperfections and the residual stresses. The ultimate strength of the beams is shown to be presented in the form of the bifurcation set in terms of the elasto-plastic buckling moment and the imperfection term that follows the 1/2 power law.
It is found from the study that the characteristics of the elasto-plastic lateral torsional buckling may be identified as the fold catastrophe just like the case of centrally compressed columns.

View full abstract

Based upon the direct measurement of bedrock motions, we proposed two models to predict the zero-damped velocity response spectra in the bedrock of the Tokyo metropolitan area for the period range from 0.1 to 10s. These models were derived separately analyzing acceleration records of category A and category B earthquakes. Category B earthquakes distributed in and around the Philippine Sea plate. The response spectra predicted from the two models differ several times at the periods around 0.5s for the same earthquake magnitude and hypocentral distance, while they come closer together at the periods longer than 1s. Proper prediction of the bedrock motion is impossible for this area so far as a single model is used, where all earthquakes are mixed up.

View full abstract

One box-section member whose dimensions are almost equal to the truss chord of Iwakurojima Bridge, (the world's longest cable-stayed bridge) has been fabricated. The fabrication process of the bridge has been evaluated throughout the construction and the member is used in the following fatigue test. The dimension of the member is 1000mm wide and 1000mm high, and a 600MPa class steel of 45mm thickness is used. The corner joints of this member contain groove welds with partial penetration from outside and fillet welds from inside. The fatigue test is carried out under the four point bending method by a fatigue testing machine with the dynamic capacity of 4MN. Fatigue cracks originate from the location of distinct surface ripples in the stop-and-start position of the inside fillet welds in the corner joints and in several positions in the member.

View full abstract

Polyhedral elastic membrane panels are studied with a finite element method. These panels are supposed to be elements of polygonal domes. Membranes generally suffer large deformation and wrinkling; their behaviour is highly nonlinear and introduction of numerical methods is indispensable.
The effect of the wrinkling is well formulated by means of the combination of the incremental loading method and the load transfer method. These two methods define a hierarchical iteration procedure. The method is validated by applying to simplest problems whose solutions have been established by the conventional analyses.
A hexagonal membrane panel is examined in detail which is one of the most fundamental elements of the polygons. Finally, the method presented is applied to a model dodecahedral dome and its mechanical properties are discussed.

View full abstract

The ultimate strength analysis of the Warren type of rigidly jointed truss girder structure, in which the influence of the finite displacements, yielding of material, residual stusses and initial crookedness is taken into consideration, is carried out in this paper. From the analysed results, the generating mechannism of the secondary moments in the chord members, the effect of the secondary moments on the ultimate strength of truss girders, the ultimate strength due to the failure of the chord members which behave as an end-restrained column in the ultimate state are discussed herein. It is shown that trusses have much more strength than that estimated by the end-hinged column strength.

View full abstract

The fatigue strength of transverse-fillet-welds in the attachment of diaphragms is a decisive factor in the design of box-section truss members. The fatigue strength of this joint is influenced by the profile of the fillet-welds at the scallop and the residual stress created by the corner weld. In order to study the fatigue strength at the ends of the scallops, a real size cruciform model was fabricated, where a pair of rib plates were welded on both sides of a base plate containing four longitudinal weld-beads. The size of the base plate was 45mm thick and 300mm wide. Fatigue tests were performed with an electro-hydraulic fatigue testing machine of a dynamic capacity of 4 MN. The tests were performed with stress rations at O, -1, -2 and -5.

View full abstract

This paper describes a data-base approach to the ultimate compressive strength of unstiffened plates loaded in uniaxial compression. A total 793 individual test results of plates is surveyed and stored into the Numerical Data-Base for the Ultimate Strength of Steel Structures (NDSS). Statistical assessments are made for the plate test results to compare with the available strength formulas. And the effects of initial imperfections such as residual stress and initial out-of-flatness are discussed. Then strength formulas are proposed for two groups, “with residual stress” and “without residual stress”. to explain more accurately the test data.

View full abstract

A beam theory including the effect of shear deformation is formulated in finite deformation. The governing equations are organized in terms of two non-dimensional parameters, one of which is the slenderness ratio. In order to quantify the contribution of shear, the elastic and inelastic buckling of a simple beam is analysed, and results are examined with respect to those two parameters. The significant reduction of critical stresses is observed for shorter columns, and the range of the slenderness ratio is obtained, in which the difference between this theory and the Bernoulli-Euler beam theory becomes prominent. Moreover, judging from the fact that shorter columns buckle inelastically, this theory becomes more important for the inelastic analyses of such deep beams.

View full abstract

Accurate equilibrium equations and appropriate static and geometric boundary conditions are derived for the geometrically nonlinear theory of shells undergoing finite rotations without restriction to small strains. The principle of virtual work is used to obtain the shell equations in which a nonrational tensor of change of curvature is employed. The introduction of variations of displacement vectors instead of those of displacement components makes it possible to reduce computational efforts for deriving the shell equations. The effects of finite rotations at the shell boundary are strictly taken into account utilizing the total finite rotation vector for the boundary.

View full abstract

An analysis of identifying the multiple rupture process of the 1979 Imperial Valley, California earthquake is made utilizing distinct phases in the near-field horizontal accelerograms. Two methods are employed in the identification of the distinct phases; one is the visual inspection method and the other the quantitative method. The location of the source corresponding to the distinct phases is determined using the S-wave travel time curve. It is found that the earthquake is a multiple event with three smaller events in the period of less than 1 to 2 seconds and that the main energy is released from a localized portion about 8 to 13 kilometers from the epicenter to northwest along the fault. The results are compared with the spatial dislocation model of this earthquake obtained by Hartzell and Helmberger.

View full abstract

Fourier phase differences of earthquake motions are investigated by deriving the relationship between the phase differences and the envelope function of narrow band components containing a certain frequency of the Fourier spectra. By making use of the relationship, it is verified that the probability distribution of phase differences is closely related to the envelope of earthquake motion. The relationship is also used to analyze the nonstationary frequency content of recorded accelerograms, the result of which is compared with evolutionary spectra. It is concluded that the phase differences are useful to evaluate the nonstationary frequency content of earthquake motions. It is suggested that the concept is also useful for the simulation of earthquake motions.

View full abstract

In recent years, the system damping effects on wind-induced response of cable-stayed girder bridges have been attracted attention. The authors previously defined a governing cause due to the so-called internal resonance, and reported that the system damping would have good effects on dynamic response to moving design live loads.
In this paper, an analysis technique of time series response of bending aeolian oscillations is proposed taking into consideration the internal resonance. Furthermore, time series response analyses of an actual multi-cable-stayed bridge are performed from the the viewpoints of dynamic safety. Finally, the authors attempt to obtain data on the application to design for wind resistant stability.

View full abstract

This paper presents the experimental researches on the static and dynamic behaviors of concrete filled steel columns. It is clarified from a series of model tests that the concrete filled steel columns have not only sufficient strength, rigidity and ductility against static forces, but also various merits such as the simplified structural details of stiffening ribs at the base plate and the corrosion treatments without painting at the inner surface of steel column. Through the field tests of a bridge pier made of concrete filled steel column, it is, moreover, confirmed that the normal and shearing stresses due to the shear by phenomena at the corner parts of column and beam of pier as well as the anchor part of column to the foundation can significantly decrease after composited. The dynamic properties such as natural frequency and damping coefficient of the concrete filled steel pier can also remarkably reduce to greater than those of non-composited-column.

View full abstract

When constructing different-type bridges on one route, bridge engineers tend to design each of them separately according to the specification. However, their efforts often yield ill-balanced, inharmonious works for lack of knowledge of aesthetic design methods. This is because there is no aesthetic rule to follow on this kind of problem. This paper proposes an aesthetic design method and its principle to unify designs of such different bridges, it also treats of the actual design-planning of Honshu-Shikoku Bridges on Kojima-Sakaide route. Using the method, new forms of piers, towers of the bridges designed by the author are compared with the forms planned by Honshu-Shikoku Bridge Authority in the beginning.

View full abstract

The three-dimensional rotations of finite magnitude are well known for being out of a linear space, but to be dealt with in the nonlinear field, more generally. Further, in the formulation of variational problems such as the present elastic deformation, the way of describing moments is significantly associated with that for finite rotations. In this paper, according to the theory which has been proposed to decompose a moment into the components in such a manner that the inner products between its components and differentials of the rotation parameters represents the real infinitesimal work, an actual discretization of thin-walled beams will be developed without any restriction on the magnitude of displacements themselves, under the condition of small strains.

View full abstract

Frequency characteristics of spring constants at both ends of a suspended transmission line subjected to horizontal periodic dis placements at one end are obtained theoretically by using the method of modal analysis and by analyzing the equations of motions directly. Both results agree well with each other and with the experimental results of a transmission line model. They show that the spring constants at both ends are not always equal and positive, but have frequency domains in which the spring constants are negative. The time histories of additional tensions in a transmission line subjected to earthquake-like excitations are computed by using the former method, and it is presented that only a few vibration modes which have large tension modes contribute to the seismic responses for tensions in a transmission line.

View full abstract

The heaving aeroelastic responses of rectangular cylinders are well known as vortex-induced oscillation and galloping. But these responses are not always observed separately. The response of a 1:2 rectangular cylinder in higher winds above U_cr(=1/S_t), concerned in this paper, is this case. And it was considered as a “mixed-type” response. This paper shows that two aerodynamic forces, self-induced one and vortex-exciting one, which induce this response, can be estimated independently by their unsteady pressure characteristics. Consequently we result that this response should be classified not as a mixed-type response, but as a galloping response.

View full abstract

Pipe beam roof is a supporting and shielding system primarily used for tunneling through under railways without interrupting railway services. The pipe beam roof is formed by connecting the pipes over the full length with the interlocking joints after driving them horizontally below rails. By these joints, loads are distributed to adjacent pipes, and the rigidity of this structure is fairly improved for the moving train.
In this study, theoretical investigations are made on the load distribution by joint in order to propose a design method for the structure. The governing paramaters as well as their numerical range corresponding to real structures are presented for the load distribution. Then, within this range, the effect of these parameters are examined in detail.

View full abstract

The purpose of this paper is to analyze the irregularities of seismic motion imparted into structures in reference with its acceleration amplitude, period and phase content, and to make clear these effects on the response of structures.
In this paper, it is discussed that of all irregular waves imparted into a certain elasto-plastic system, the wave having the same period component as apparent natural period T_e, which is calculated by structural characteristics, has the most influence on its system. As a result of numerical analysis, it is possible to estimate the response of elasto-plastic structures easily and effectively by making linearized system and sinusoidal excitation wave having apparent natural period T_e.

View full abstract

A computational method is presented for the optimum design of structures subject to external forces from uncertain directions. Finite element method is used for structural analysis so as to make use of the algorithm for optimizing various types of structures.
As the loading direction changes, the member forces will also vary accordingly. The direction most unfavorable to each member does not necessarily coincide each other. Hence in each member the critical member force and its corresponding direction need to be found before each optimization step. Constraints on stresses, bucklings, dispacements and member sizes are considered. Also presented is the case in which multiple external forces vary their directions independently. Example problems are solved to show the validity of the optimization algorithm.
The concept presented here is useful for the optimum design of towers and marine structures.

View full abstract

Numerous degrees of freedom are necessary for dynamic analysis of electrical towers-transmission lines system by the F. E. M.. An analytical theory for simply and practically calculating the vibration characteristics and earthquake responses of the system in the direction of transmission lines is proposed with consideration of the dynamic interactions between towers and transmission lines. In the proposed theory, the number of degrees of freedom is considerably reduced by means of regarding the transmission lines as springs connecting the towers and by using the substructure method. The earthquake responses of three typical models of the system are calculated. One of the principal conclusions is the following. In every case, earthquake response values of the towers obtained by taking into consideration the interactions between towers and transmission lines are smaller than those of a single tower.

View full abstract

Node reordering algorithm for decreasing the matrix profile is newly proposed, and its efficiency is surveyed through a number of test examples by comparing with the results by well-known RCM and GPS algorithms. Proposed algorithm consists of three steps; Firstly any graph obtained from the objective matrix is dissected into gatherings of simple subgraphs, secondly all nodes in each subgraph are ordered independently so as to decrease the fill-ins, and finally all nodes are rearranged in order to decrease residual zero entries in the profile area. The results show that the proposed method not only requires less computation time comparing to the GPS algorithm but also can give better reorderings, especially for systems with rather complex toplogical structures.

View full abstract

This paper studies an efficient method of estimation of dynamic properties such as natural frequencies, modal damping coefficients and participation factors, of multiple degree of freedom linear system. The EK-WGI method which was previously proposed by the present authors is used for the solution of a system identification problem. To solve a multiple degree of freedom linear system, the modal analysis technique is corporated into the formulation of the state space equation. It is found that dynamic properties can be estimated even if an excitation and a single response time history at any point of mass are available. It is also found that if observed response at each mass are available for the identification, they can be used for the verification of estimated dynamic properties.

View full abstract

Collapse of structures during strong earthquakes is studied using monte-carlo simulation on one mass system with the hysteretic restoring force. Decision of collapse is given by deformation of the system. Model system which collapse just due to static 0.2G acceleration (=design seismic coefficient popular in Japan) is made in computer and is applied by natural or simulated earthquakes with various duration and acceleration. Averaged acceleration of earthquake, duration time of earthquake, and ductility of the system have much effect on the collapse. In same cases, the maximum acceleration to collapse the model system with much ductility applied by short duration earthquake is several times more than 0.2G.

View full abstract

The object of this paper is twofold: (1) To produce synthetic ground motions at near-source region using fault dislocation theory, and (2) To present a simplified method for ground strain estimation. The records of the 1978 Imperial Valley earthquake are used. Nonuniform slip on fault is considered by the three point source model. Syntheses are compared with records and results obtained by Hartzell et al. With respect to waveform and duration, the synthesized ground motions resemble the recorded ones rather than results calculated from the uniform rupture model. Further, simple formula for estimating maximum ground strain is proposed after discussion on influence of separation distance on average and local strains, and apparent propagation velocity. In particular, the method should be useful for practical application.

View full abstract

The proposed method of structural analysis is suitable for the use of a small computer. The structures considered here consist of one way continuous rectangular plates with two opposite free edges and arbitrary intermediate supporting conditions. One of the features of this method is a drastic reduction in the number of degrees of freedom needed for numerical analysis in comparison with the FEM, and the other is the efficient formulation of a stiffness matrix method combined with a relaxation technique. As for the detailed procedure, the displacement functions in series form and the point-matching method are adopted to derive the stiffness matrix of large-size rectangular plate panels. The relaxation technique is developed in a way that is compatible with the stiffness matrix method. In addition, the initial unbalanced quantities required in the iteration process are given by an elementary beam theory.

View full abstract

In this study, it is investigated how the seismic damage of lifeline systems interacts on each other during and after severe earthquake. For this purpose, items associated with seismic damage are carefully extracted on the basis of past earthquake damage data. Then, by appling both the interpretive structural modeling method and the fuzzy structural modeling method on these items, a seismic damage influence model is constructed in the form of an oriented graph, whereby the mutual relationship among lifeline damages are evaluated qualitatively.

View full abstract

Nonlinear responses of the cable through bifurcation under inplane forcing are analyzed in the present paper. Two different responses, that is the anti-symmetric response of the cable which is supported at the same level under symmetric forcing and subharmonic responses are treated. The equation of motion of the cable is solved by the Galerkin method and the harmonic balance method. From the results of numerical examples, the frequency range where the anti-symmetric response exists varies with the sag-to-span ratio and its width is broad in the particular sag-to span ratio. The secondary unstable region is important compared with the primary unstable region. Subharmonic responses with order 1/2 and 1/3 are also obtained. The subharmonic response with order 1/2 only in the case of the cable. This response is important for the cable whose quadratic nonlinear terms are predominant.

View full abstract

The connectivity of the lifeline system is studied based on the two ratios: (a) “damage ratio” which is pertaining to the degree of physical damage and is defined as an expected number of failures per unit length; and (b) “normal-supply ratio” which is pertaining to the system connectivity and is defined as the ratio of the length of the portion being connected with the supply node to the total link length of the system. The purpose of this research is (a) to develop a methodology for examining the relationship between the two ratios just after the earthquake and during the recovery period for several simple network systems, and (b) to clarify the change of the reliability of the system in accordance with the change of the scale and configuration of the network due to the growth of the community.

View full abstract

Proposed is a modification coefficient which is to be multiplied to earthquake response spectra with 5% damping ratio of critical to obtain response spectra of arbitrary damping ratio h. A modification coefficient ξ_SA was defined as S_A(T, h)/S_A(T, 0.05) in which S_A(T, h) represents absolute acceleration response spectral amplitude for natural period T and damping ratio h. Studied was a variation of ξ_SA in accordance with natural period T and frequency characteristics of ground motions represented in terms of acceleration response spectral ratio β(T, 0.05) (β=S_A(T, 0.05)/a_max, in which a_max represents the peak acceleration). Horizontal strong motion records of 206 components obtained at free field sites in Japan were used for the analysis. It was concluded that coefficient ξ_SA can be approximated as 1.5/(40h+1)+0.5.

View full abstract

General nonlinear displacement fields for a plate segment comprising the cross section of a prismatic member are derived from Lagrangian strain-displacement relations by abandonning the assumption of no distortion of cross section. Second order displacement fields are also derived from those displacement fields for the buckling analysis. On the basis of the displacement fields and strain components obtained, four simplified differential equations governing the buckled state of I-beams in pure bending are presented, considering the initial curvature due to bending about their major axsis. The solutions obtained using the simplified equations compare favorably with results obtained by a finite strip method with a larger degree of freedoms. The effects of initial curvature of bending are remarkably appeared in the region between local buckling and distortional lateral buckling, and decrease buckling stress of bending in the case of no initial curvature to about 89%.

View full abstract

The theory of the filtered Poisson process is applied to analyze the response of highway bridges to stochastic live loads. For each lane of the bridge, the live loads are idealized as a train of uniformly moving vehicles for which a statistical data base exists so that the probability distribution functions of the vehicular headways and weights can be established. In particular, the headways are assumed to have exponential distributions. The cumulants, and hence the mean value and variance, of any desired response quantity at any bridge location can then be obtained with the aid of the associated influence lines. On the basis of the cumulants thus obtained, the characteristic function of the response quantity can be constructed and used to obtain the corresponding probability density function by means of the Fast Fourier Transform (FFT) technique. Examples are given for the cases of a simply supported girder bridge and a three-span continuous girder bridge subjected to such stochastic live loads.

View full abstract

Fatigue crack growth rates were presented for the various weld metals of 600MPa and 800MPa class steels. The weldings were done by submerged arc processes and MIG processes under large and small heat input conditions. For the weldings of 600MPa class steels, two weld materials of different strengths were used. The tests were performed on the modified compact-type specimens of 12.5mm thick. A fatigue crack propagated in the weld metal in the long-itudinal direction or the transverse direction to the weld line of the specimen. Fatigue crack growth rates in the weld metals did not vary so much according to the direction of the fatigue crack, the differences in the strength of weld materials and the variations of the heat input conditions.

View full abstract

Field observations are carried out on the load distribution by joint in a pipe beam roof which is used for constructing a passage under the Tohoku main line of Japanese National Railways.
The field observations are composed of two stages. The strains and the displacements of the pipes are measured under the dead load as well as the moving train load in order to confirm the accuracy of the proposed structural analysis and to examine the impact coefficients for the structure. Then, the natural frequency and the damping ratio are measured by the free vibration test to obtain the fundamental data for the dynamic properties of the structure.

View full abstract