An ozone tester, equipped with apparatuses measuring either stress relaxation of specimens at constant strain or creep of them at constant stress, was made in order to investigate differences of ozone degradation caused by different loading conditions. Sheets of butyl rubber blended with EPDM in three levels were specially made in this study. Mechanical property of the sheet with each mix were varied by the degree of vulcanization. At first step, the specimens were exposed to ozone in the ozone tester controlled at a concentration of 50 pphm at 40℃, and stresses or strains of them were recorded during the test. The specimens were periodically inspected with cracks or ruptures if any. Good ozone resistance was observed with lower vulcanized samples in the test with constant strain, on the contrary, was observed with higher vulcanized samples in the test with constant stress. The results obtained were considered to depend on the visco-elastic behavior of samples, then a new factor expressing loading levels, an average of apparent strain energy, which can be obtained by integrating values of strain energy and dividing it by the period to failure. The availability of the new factor was approved by additional ozone tests with different ozone concentrations, temperatures and loading conditions. Keywords '. Ozone, strain energy crack, visco-elastic, creep, relaxation

View full abstract

Most seismic hazard analyses have been based on the stationary Poisson process for earthquake occurrences. However, it is more appropriate to assume nonstationary earthquake occurrences for estimating a seismic hazard in a short period. A model based on the time predictable model in area seismic sources and the Semi-Markov process for describing nonstationary occurrence of earthquakes is developed. The probability distribution of the maximum ground motion intensity in t years is calculated by the magnitude distribution obtained by the Semi-Markov process, the spatial probability of occurrence of earthquakes and the attenuation curve. The observed data of earthquakes in seven area sources in and near Japan show the time predictable model is adequate even in the small region such as one degree square in latitude and longitude. Those data show the scale, parameter of Weibull distribution applied to time intervals of earthquakes is about 1.5. The estimates of the maximum magnitude of earthquakes on an area seismic zone with the Poisson model are overestimated after a great earthquake has occurred, but are underestimated by about 5% after a seismic gap or after a small earthquake has occurred. The estimates of the maximum peak ground acceleration on an area seismic zone based on the Poisson model underestimates by about 10% after a long seismic gap. As an illustration of more realistic model, the seismic hazard analysis of a site in the Kanto district is presented.

View full abstract

In part 1 and 2 of this serial paper, we described the experimental results of the scaled models of multi-story, multi-span reinforced concrete frame under monotonic loading in consideration of the effect of axial elongation in bending yield beams. In this study, a cyclic loading experiment of four-story three-span reinforced concrete frame is performed to consider the effect of axial restriction of beam deformation. Finite segment method for the reinforced concrete frame with consideration of axial and bending deformations of beames and columns, shear deformations of beam-column joints and bond-slip effect of steel bars in concrete is used for analyzing the mechanical behaviors of the specimen. Main conclusions are: 1. Horizontal resistance force of this structure increases, because axial restriction imposed by the horizontal stiffness of columns causes the beam bending strength to increase. 2. When a reinforced concrete frame is subjected to loading in the right direction, the elongation of the beam in right span is larger. And loading in the left direction, the elongation of the beam in left span is larger. Then the elongation of the beam in center span is smaller than that of the beams in both-end span. 3. The Ibad deflection relationships obtained from the analysis are almost coincident with the experimental results. The analytical results provided us more detailed information such as bending moments, shear forces of all members.

View full abstract

Design procedures for reinforced concrete (R/C) members in shear have been studied by many. Researchers, since the truss analogy was developed by Ritter nearly a century 'ago. Recently, rational approaches to shear design were actively developed in Europe, America and also Japan. The typical approach is a macro-model approach based on the lower-bound theorem in the limit analysis procedure. The authors have previously proposed an evaluation method for the ultimate shear strength of R/C members under combined axial, bending and shear forces. This method is called "Cumulative Strength Theory (C. S. T.)", which also uses the concepts from plasticity. The most significant characteristic of this theory is that its solutions are obtained by accumulating data on the strength of two types of shear resistant mechanisms, the beam mechanism and the arch mechanism. The shear design quations based on C. S. T. are proposed in this paper, which are comparable in rationality and generality to the design equations for flexure and axial load. A relationship between the proposed, the C. S. T. and the flexural equations is designated. Physical significance v of the proposed equations is clarified by parametric calculations on the constitutive factors in R/C members. Predictions from, the proposed equations are compared against experimental results, and applicable scope is found. Excellent accuracy and great utility of the proposed equations are also indicated.

View full abstract

This paper discusses an analysis of the shearing forces around the drop panel of a flat slab structure. In this study, several shapes of the drop panel are investigated in case of an uniformly distributed load. The analytical solutions of the shearing forces are derived from the analogy between the equation of a slope of the membrane and the equation of a shearing forces of a plate. The results of the shearing forces among the several shapes are obtained by the experiments in membrane.

View full abstract

Circular tubular struts with a compact section sustain local buckling after large deflection of these due to flexural buckling. Local buckling has significant effects on both post-buckling compressive and straightening-out tensile envelops during cyclic loading. Although the inelastic cyclic behavior of struts has been the subject of intensive study, its interactions with local buckling still remain to be identified more clearly. A point-hinge approach incorporating both inelastic and post-local buckling deformations was devised to calculate buckling, post-buckling, and inelastic tensile hysteretic envelops of tubular struts under reversed load. This method of analysis was found to be capable of simulating experimental results satisfactorily. Several numerical results were presented to evaluate post-buckling capacity of bracings of frames to resist seismic shear loads.

View full abstract

The authors had already examined the fire resistance of concrete-filled square steel tubular column, without fire protection, under constant axial load. These columns developed local buckling in the steel tube, but the fire resistance was maintained due to the transfer of the axial load from the steel tube to the concrete. In this paper, we have experimentally examined the mechanical behavior and fire resistance of these columns under axial load and bending moment. We had two types of heating tests. One type was carried out under eccentric axial load and another type was carried out under axial load and bending moment transfered from beam. The results obtained from the tests are summarized as follows: (1) Under an eccentric axial load, the end of the column members have sufficient rotation capacity. (2) Considering the framed structure at the occurence of fire, as the decrease of yield stress and rigidity of steel tube due to the temperature rise, the quantity of transfer of bending moment from beams to columns decreased. (3) Finally, the results were similar to those obtained in the heating test of the composite columns subjected to central compressive load, and consequently the fire resistance of these columns was mainly determined by axial load ratio.

View full abstract

Rapid reanalysis of eigensolutions after modifications is a problem of considerable practical importance. Several methods have been developed to compute eigenvalues by using modified parts. This paper proposes a numerical method for calculating eigenvalues of modified system using the numerical solutions of unmodified system with the Durand・Kerner-like method. The advantages of the proposed method are examined comparing with the solutions of House-holder-QL method by several numerical examples. The results obtained agree well with the exact solutions in short CPU time and this indicates that the proposed method provides efficient convergency.

View full abstract

Long period (2.5-12 sec) seismic motions on the Kanto sedimentary basin, Japan, are simulated by a two dimensional basin shaped underground model. We treat three shallow seismic sources located near the Izu Peninsula (their epicentral distances from Tokyo are 100-150 km). Simulations are carried out by the 2-D boundary element method with the Haskell-type dislocation source using known underground data and fault parameters. Comparing simulated motions with observed seismograms inside and outside of the basin, the following conclusions are obtained. 1) Main parts of the long period seismic motions in the basin are simulated well, with regard to their amplitudes and phases, by the 2-D basin shaped underground model. However, coda parts of the motions in the basin are not simulated. 2) The main phases in the basin are local Love waves excited in the layer of Vs=0.7 km/sec, and their characteristics are mainly controlled by the underground structure rather than the source mechanism.

View full abstract

Long period (2.5-12 sec) seismic motions on the A new method for evaluating the stiffness of the finite element with bilinear and trilinear displacement functions is proposed. In this method, an approximate solution of the differential equation in the displacement field of the finite element is used to obtain the 'stiffness matrix of the element. Several numerical examples of plates and shells are solved for investigating the effectivness of the analysis by the present method. The results show the method is better than those by the minimum potential energy method in the precision of analysis. It is also clarified that the reduced integration method proposed by Zienkiewicz and his co-workers and the method by using Wilson's nonconforming element are the same as the present method in the concept. Therfore these methods should be used with hybrid displacement method.

View full abstract

Viscosity of material is often explained with the use of "dashpot" in earthquake engineering nowadays. It follows that the attenuation of wave propagation in such material must be an exponential function. This limits us to search the real characteristic of earthquake wave propagation because the real attenuation can not only be an exponential function naturally. To realize the nature of earthquake wave, the fading memory theory is introduced in this paper. The attenuation property Q and theoretical solution of wave propagation in material with fading memory are presented. Some examples are calculated. The memory functions of dashpot system and creep function are also shown. The final result can be noted as: not like that in material of dashpot system, to find the dynamic behavior of real ground becomes possible if memory function is justified pertinently.

View full abstract

As part of a comprehensive study on developing the pseudo dynamic test using substructuring techniques, this paper proposes new integration techniques effective to the substructure pseudo dynamic test. First, conditions that the substructure pseudo dynamic test should fulfill in order to ensure accurate responses were identified. Second, various integration methods were surveyed with respect to their applicability to the substructure pseudo dynamic test, and based on the survey, algorithms using mixed integration methods and the operator-splitting (OS) method were proposed. Stability and accuracy of the algorithm using the OS method was examined analytically, and it was proven that this algorithm is perfectly adaptable to the substructure pseudo dynamic test and can provide unconditionally stable responses. Criteria for estimating the accuracy of this algorithm were also presented. Finally, effectiveness of this algorithm was calibrated for a structure with 20 degrees and a 3 story structure with base isolation devices, and advantages of the proposed algorithm over conventional integration procedures were demonstrated.

View full abstract

In order to make clear the PULSE RESPONSE BEHAVIOURS of asymmetric building structures, a parametric response study on lateral-torsional responses of inelastic single-story mono-eccentric building structures has been performed by using MODE-SEPARATING PULSE RE-SPONSE ANALYSIS METHOD proposed in the previous paper presented by the authors. In this paper, the single-story mono-eccentric building structures are idealized as a system consisting of a rigid deck suported by 4 frame and/or wall assemblages situated at the periphery which are assumed to have force-displacement characteristics represented by a bi-linear model. An unidirectional excitation is considered in the response analysis. In this research, the system parameters and their numerical ranges are : (1) eccentricity ratio (0.0-1.0) (2) aspect ratio (1.0-10.0) (3) initial elastic period (0.1-10.0 sec) (4) initial base coefficient (0.1-1.0) Then, the effects of the system parameters on the lateral-torsional response have been examined through the parametric response study. The results show very interesting conclusions among which the followings are important for an earthquake-resistant design of building structures with the eccentricity in strength as well as stiffness. 1) The lateral displacement of the center of mass does not vary with changing the eccentricity ratio at al^ for the system having small aspect ratio, but increases largely with increasing the eccentricity ratio for the system having large aspect ratio. 2) The torsional displacement of the system increases almost linearly with increasingthe eccentricity ratio for the system having small aspect ratio, but not linearly for the system having large aspect ratio. 3) The maximum displacement of element located on the weak stiff edge in the eccentric direcition increases with increasing in eccentricity within the almost entire period domain. 4) The system shows some what constant in the total deformation energy despite of changing eccentricity and aspect ratio in an asymmetric building structure. But, the deformation energy distributed to every resisting element in the X or y direction varies largely wiht changing the eccentricity ratio. 5) The effects of the eccetricity ratio on the responses may be divided into 3 regions: (1) small eccentricity ratio: 0.0-0.2 (2) medium eccentricity ratio: 0.2-0.5 (3) large eccentricity ratio: 0.5-1.0 6) The increse of the maximum displacements trends toward unchanging despite of increasing the aspect ratio when the aspect ratio is larger than 3.0 -4.0.

View full abstract

Theoretical and empirical knowledge on the dynamic response of a pile group is none the less limited. Effective tools for the dynamic analysis of a pile group should be provided immediately and the dynamic characteristics should be clarified. In this paper, the dynamic analysis method of a pile group considering the pile-soil-pile interaction effects is presented. Herein, the effects are closely included by using a point load solution derived from the thin layer formulation. Group effects on pile spacing ratios, a number of piles and soil structures are studied from the impedance functions of a pile-foundation obtained by this method. Furthermore, the reducing values on the spring constant and the damping factor of a grouped pile against those of a single pile, which are respectively defined by eqs. (26) and (27), are quantitatively evaluated from the impedance functions, and then the reducing effects on the spacing ratios and a number of piles are discussed. The resultant values would be useful for the dynamic response analysis of the structures supported by a pile group.

View full abstract

This paper is a follow-up of the paper with the same subtitle in the previous Journal of Structural and Construction Engineering (Transactions of AIJ, No. 413, July, 1990). In this paper, a practical calculation method for the stiffness matrix elements, pertaining to the shear type stiffness and the rotational stiffness of the beam-to-column connections assumed to be rigid zone in elastic framed shear walls, is proposed. The shear type stiffness of framed shear walls without floor" slabs can not be evaluated satisfactorily by the I-beam solutions itself. Also, ' the rotational stiffness of the beam-to-column connections can not be evaluated by using I-beam theory. This is because the rotational stiffness of the connections is different from that of horizontal or vertical cross section of the shear wall. Therefore, the practical calculation charts based on the two-dimensional theory of elasticity are proposed in order to obtain the shear type -stiffness and the rotational stiffness of the connections with a high accuracy. These proposed stiffnesses will be adopted in the formulation of a practical nodal stiffness matrix of the framed shear walls together with the flexural and axial stiffnesses which were proposed in the previous paper. This practical nodal stiffness matrix will be useful in analyzing the elastic behaviour of the frame structures stiffened with shear walls by matrix method of structural analysis.

View full abstract

If a designer can use a computer language which can directly represent a data flow diagram, some design solutions can be systematically generated and easily tested. We arranged such a computer language and discussed the programming technique using the language from a standpoint of an architectural designer. Our discussions are summarized as follows: (1) Because the procedure of generate-and-test can be directly expressed by the data flow formulation, the designer can easily describe the calculation procedure. (2) Programming the generate-and-test procedure using a procedural computer language such as FORTRAN cannot make good use of the above advantages (1), because the programming re-quires to translate the data flow formulation into a control flow diagram. (3) The above defect is overcome by the use of a computer language which is able to correspond to the data flow formulation directly. We arranged such a computer language and called it DSP in this paper. The design calculation system is easily constructed and maintained the designer himself using such a computer language. , (4) DSP has some debugging aids such as a data flow analysis and a data type checking, and these are helpful to formalize the design calculation procedure.

View full abstract