Most fluid-applied polyurethane membrane systems are composed of two layers such as a main polyurethane layer and a top coating layer. It was sometimes found out in actual buildings that a top coating layer is ruptured by movement of cracks in substrate. This kind of defect hasn't been reported till now. The purpose of the paper is to make clear the reason why this kind defect occurred and to propose effective ways to improve durability of it. Three works such as a dynamic outdoor exposure test of polyurethane membrane systems, measurement of elongation of top coating layers of them and a fatigue test of dumbbell specimens of them were carried out in the study. Firstly test specimens of various thick polyurethane membrane systems applied on substrate were exposed outdoors for five years. The artificial cracks prepared at the center of the specimens were repeatedly moved once a week during test and top coating layers of the specimens were observed. After a few months later, fine cracks appeared in the coating layers of some specimens and then propagated into the polyurethane layers underneath them.

 To understand why cracking started in a top coating layer, it is needed the information on how a top coating layer is extended by joint movement of a substrate. Then, the specimens on which target scales for measuring elongation previously printed on the surface top coating layers were extended to joint width of 2.0mm in incremental steps, and measured elongation rate of the surface coating layers. The highest elongation was observed in the top coating layer just on the artificial crack of the substrate, and it increased according as thickness of a beneath polyurethane layer decreased.

 Secondly, repeated extension tests of dumbbell specimens of membrane systems were carried out at various elongation rates to know the fatigue properties of them. Any rupture didn't appeared in the specimens subjected to the elongation rate of 10％. However top coating layers ruptured in the specimens over 25% in elongation, which is almost equivalent to the elongation rate measured in the membrane system of 1mm thick at crack width of 0.5mm in the previous tests.

 Based on the above tests results, the three ways to avoid cracking in a top coating layer of a fluid-applied polyurethane membrane system were proposed, such as making a membrane system with sufficient thickness of polyurethane layer, improving the ability of fatigue resistance of a top coating layer, and laying an air-permeable sheet underneath of a polyurethane layer.

 Finally, the conclusion of the study was summarized follows,

 (1) It was found out through dynamic outdoor exposure tests that cracking first appeared in a top coating layer and propagated into underneath polyurethane layer.

 (2) This is because that fatigue resistance of a top coating layer is lower than that of a polyurethane layer.

 (3) Three ways to improve fatigue resistance of a polyurethane membrane system were proposed.

 In recent years, a lot of studies on seismic vibration control system using dynamic mass (D.M.) have been reported. Real buildings where D.M. was adopted are increasing. In our previous study, we have proposed a simplified placement method of Maxwell-type oil dampers (C type). After the study, this paper estimates performance of tuned dynamic mass system (MC type) and proposes combination system (C-MC type) of C type and MC type. Design examples of the optimum story-wise placement of C-MC type are shown for two MDOF building models. Time history response analyses are carried out for buildings installed with the necessary number of dampers subjected to the design earthquake. The results demonstrate the effectiveness of the proposed combination system.

 Based on these studies, the following results are obtained.

 1) Simplified design method can be applied to MC type and C-MC type. We can determine story-wise damper quantity against criteria for MDOF easily and rationally.

 2) Constant for all stories (Arrange2) is rational in MC type. But variability tends to be large by influence of higher modes. Assuming that structural damping of building frame is stiffness-proportional damping or if higher frequency components of the design earthquake are small, then we may adopt MC type.

 3) When additional damping factor is about 5 to 10 percent, we can reduce damper quantity and add damping to higher modes with C-MC type. The proposed combination is better placement method.

 Theoretical constructions in case of using hysteretic dampers and expansion to simplified method are made will be an issue to be addressed in the future.

 Seismic control has become generally used in super tall steel structure buildings. A combination of viscous dampers and hysteretic dampers is increasingly used nowadays to ensure high earthquake resistance of high-rise buildings – as such, this is called a “combination-system”. In design of buildings, simple model such as equivalent shear spring model is frequently used. However, there is no current research on simple model of “combination-system”, thus, this paper proposes a method to create an equivalent shear-spring model of “combination-system” consisting of its characteristic values, dampers and their support members, and frames.

 In creating the shear spring model, the parameters are calculated by conducting static analysis of the building frame in four states – (i) state N for frame only state, (ii) state R in which an elastic spring with extremely high rigidity is placed in the damper’s installation position, (iii) state pN in which elastic springs are only placed in the hysteretic dampers’ installation stories, and (iv) state pR in which an elastic spring with extremely high rigidity is only placed in the hysteretic damper’s installation position.

 The parameters obtained from states pN and pR are important in evaluating seismic control effect of “combination-system”. The obtained characteristic values from state pN show that the upper limit of the effective deformation ratio in the viscous damper’s installation floor decreases. The obtained characteristic values from state pR show phase difference between the hysteretic damper load and viscous damper load.

 A 30-storey high-rise building with brace-type continuous arrangement of viscous dampers and hysteretic dampers is used to verify the proposed model. A total of 13 damper placement combinations are used in this study. This paper shows that the proposed model reproduces the results of time-history analysis of member models. With respect to energy absorption of damper, the proposed model is more accurate than the previous models because the former evaluates various characteristic values and can reproduce the damper load at each step of the time-history analysis.

 Roof bearings of many gymnasiums were destroyed by the 2011 off the Tacific coast of Tohoku Earthquake. Typical damages were crush of base mortar, elongation or break of anchor bolt. Because of the damage, the gymnasiums could not be used continuously as a shelter after the earthquake.

 After the quake, seismic response of steel roof gymnasiums has been energetically studied. However, there are few studies considering the detailed non-linear restoring force characteristics of the conventional steel roof bearings. In this study, we firstly propose a nonlinear model of the restoring force characteristics of the conventional roof bearings based on the results of cyclic shear loading tests of bearings in Reference 6, 7).

 The skeleton curve between the shear bearing force and the displacement of the base plate can be expressed in 4 stages. In the stage 1, the shear force increases with very high stiffness until the maximum static friction force. In the stage 2, the base plate slides with almost zero rigidity within the movable range of the anchor bolt hole. After the base plate contacts the anchor bolts, the stage transits into the 3rd where the elastic bending deformation of the anchor bolts occurs. After the shear force reaches the bending yield strength, the stage transits into the 4th. In the stage 4, the shear force is resisted by the axial force of the anchor bolts in consideration of geometrical nonlinearity. The axial forces are calculated using the law of plastic flow. The proposed skeleton curve model of the bearing well approximated the test results 6, 7). Furthermore, the hysteresis curve model was developed, that was able to approximate the test results well by using the slip type restoring force characteristics.

 We modeled a gymnasium damaged by the 2011 off the Pacific coast of Tohoku Earthquake, incorporating the nonlinear model of the bearing to conduct an earthquake response analysis. From the seismic response analysis, it was confirmed that the bending crack by the out-of-plane response occurred in the RC cantilever column as the actual damage. In addition, the bearing shear stress was greater in the in-plane direction of lower frame than in the out-of-plane direction. It is guessed that the bearings resisted the inertial force of the wall plus the steel roof in the in-plane direction of lower frame.

 It was confirmed that the shear stress in the pin bearing can be well evaluated by equation in the Reference 2, 3). In this study, it was found that, when the natural period of lower structure was greater than 0.25 seconds, the bearing shear stress was greater than the inertial force of RC cantilevered column when slide bearings were used. That is considered to be the effect of impact force when the base plate collides to the anchor bolts. It was found that when the range of motion of the slide bearing increased, the bearing stress became 1.4~1.5 times as large as that of the pin bearing.

 A lath-mortar wall is used on conventional wooden house, and the wall using furring strips for damp proof is particularly attracting the attention because of its good durability. However, a lot of severe damage of lath-mortar wall was reported during past earthquake disasters. Also, the past studies on seismic resistant performance of lath-mortar wall were limited. This paper investigates the resistant mechanism and restoring force characteristics, and the analytical method is studied.

 First, to clarify the resistant mechanism and inelastic behavior of lath-mortar wall, the horizontal loading tests are done as parameter with the load distribution pattern (see Table 2). 1^st story and 2^nd story wall test specimen are prepared (see Figs. 1, 2), which wall is made of ventilation construction method.

 From test result in case of 1^st story wall specimen, the slip and pinching behavior are observed (see Fig. 5). From the observation of test specimen, the remarkable damage is not appeared on mortar, however, the pull-out of staple and damage on wooden frame are observed (see Fig. 7). The rotation of lath-mortar is occurred following the shear deformation of wooden frame. And the same ratio of each deformation is presented until 1/3- rad story drift. During the ultimate state after 1/30 rad, the rotation of lath-mortar is not progressed because of pull-out and fracture of staple joint (see Fig. 6).

 From test results in case of 2^nd story wall specimens, the slip and pinching behavior are presented (see Fig. 8). From the observation of test specimens, the lath-mortar does not show the remarkable damage, however, the pull-out of staple are occurred (see Figs. 9, 14). From the comparison of skeleton curve, the stiffness and strength show different each story (see Fig. 10, Table 3). It means that the relative displacement between lath-mortar and wooden frame presents the different trend due to load pattern (see Fig. 15).

 And the analytical study is done to discuss the deformation of lath-mortar wall (see Fig. 16). From the consideration of analytical expression and test results of 2^nd story wall specimen, the relative displacement of staple joint presents different from each other due to the center of lath-mortar wall which is calculated with equilibrium condition.

 To chase the resistant mechanism of lath-mortar wall, the inelastic behavior of staple joint becomes important. Here, the elementally loading test on partial staple joint test specimen with lath-mortar is performed (see Figs. 17, 18). From test results, the load – displacement curve presents the ductile restoring force characteristics due to suppressing by lath and mortar. Also, the ultimate state presents the pull out of staple, which shows the same mode with lath-mortar wall specimen (see Fig. 20).

 Finally, the analytical method of restoring force characteristics is studied. The analytical model of restoring force characteristics of staple joint is assumed with elementally loading test results (see Fig. 21). From analytical results of center of rotation of lath-mortar, it shows good agreement with test results (Fig. 23). Also, the relation of rotation of lath-mortar vs. horizontal displacement of wooden frame show good agreement with test results. (see Fig. 24) From these considerations, the analytical method can chase the deformation of lath-mortar and wooden frame. Furthermore, the analytical results of restoring force characteristics present good agreement with test results (see Figs. 25, 26). So the analytical method can estimate the test results well.

 To evaluate the performance of buildings constructed by the traditional construction method, it is necessary to accurately evaluate the hysteresis model of mud walls. By combining the hysteresis model of the mud wall with the frame analysis method, the numerical analysis of the frame with the mud walls can be realized. The hysteresis model of the mud wall has three curves of virgin load curve, return curve and reloading curve. In the previous model, it was necessary to adjust the parameter of the curve equation individually based on the experiment.

 In this paper, we propose a hysteresis model with the maximum load point as a parameter. The analyzing data were generated by subtracting the experimental load of the frame only from the experimental load of the mud wall. We used the experimental data of 7 specimens of Kyoto mud walls.

 The proposed model is based on the statistical processing of the experimental results of the Kyoto mud walls, and it becomes a regression curve using the load-deformation angle relation normalized based on the maximum load point and unloading point. The virgin loading curve consists of two polynomial and hyperbolic curves up to the maximum load point and in the strength lowering region. The return curve is modeled as a polynomial, and the reloading curve is modeled as a hyperbola. These equations are covered by dividing the deformation angle θ into three regions, θ < 1/90 rad, 1/90 rad ≤ θ ≤ 1/45 rad and 1/45 rad < θ.

 In spite of such a simple modeling method, it was confirmed that the result of the cyclic loading experiments of Kyoto mud walls could be expressed well. The experimental result of other researchers was also expressed well. The area of the hysteresis loop of the proposed model was 15% larger than that of the experimental results. The reloading curve had almost the same history as the experimental result, and the return curve was well reproduced after 1/45 rad, although the load of the model was underestimated in the area of small deformation angle.

 The variations of the virgin loading curve model concerning the maximum load points were shown in Fig. 16 and 17. By these combinations, virgin loading curve models of various mud walls were drawn. By giving the maximum load point, the envelope curve of the load-deformation angle of the experimental results of other researchers could be estimated well up to the vicinity of the maximum load point. Furthermore, the virgin loading curve model also showed the correspondence with the load-deformation angle relations for the design which were shown in reference 22) to 24).

 It is also possible to construct a hysteresis model corresponding to different wall lengths or other regions' mud walls by collecting cyclic loading experimental data.

 Techniques for large wooden constructions are actively developed for the purpose of reduction of environmental load. One of the solutions for fire and seismic resistance is composite structure, which utilizes different structures like steel or concrete. Wooden structure laterally connected to different structure is discussed, which is called "wooden horizontal hybrid structure". Since it involves stiff cores, most of seismic force acting on wooden parts is sustained by the cores, and wide open space with less shear walls is realized. However, such force transmission complicates the displacement mode as well as the stress distribution in the structure. The authors have proposed dynamics continuous model of horizontal hybrid structure assuming that core part is infinitely-rigid. Although the model can simulate displacement mode of horizontal hybrid structure having various arrangement of shear walls, the structure is assumed to be linearly elastic. While allowable strength design is conducted using elastic model, wooden structures generally have non-linearity even in small deformation. The equivalent stiffness is required to be determined depending on the target deformation. Horizontal hybrid structure, however, shows complicated displacement mode, and it makes it difficult to define deformation and the corresponding equivalent stiffness of each element. Therefore applicability of equivalent linear model is discussed by comparison of analysis result with test result, and simplified evaluation of member force is addressed.

 In this research, prediction by the proposed method is compared to test results conducted by the authors⁹⁾. The referred shake table test on one third scaled model of wooden horizontal hybrid structure is introduced. One side of the specimen is laterally connected to steel jig which represents stiff core. The specimen is 3-story 3-span wooden frame with shear walls. Uni-directional input motions with various intensity are applied to the specimen. Three specimens were tested and the parameters were wall arrangement and nail interval of floor diaphragm which characterizes the stiffness and the strength. The followings are findings of this research.

 1) The prediction method of displacement mode which the authors had proposed was applied to the specimens. Since non-uniform stiffness balance of each element was observed, the balance could be idealized to fit the assumption of the dynamics model. It showed close agreement with test result.

 2) Although stiffness balance of each element was changed according to earthquake intensity, displacement mode and the corresponding equivalent mass ratio were relatively stable. It insists that the existence of core is likely to prevent concentration of deformation on particular element.

 3) 30 to 50% of seismic force acting on wooden part was transmitted to core part through floor diaphragm. No. 2 having stiffer floor diaphragm showed the higher ratio compared to the other specimens. They corresponded to predicted values with acceptable accuracy.

 4) Connection moment between wooden part and core part could be conservatively evaluated using tensile force of bolt and assumed stress condition.

 5) The ratios of connection moment to connection shear force, which represents equivalent floor length, could be conservatively evaluated by assuming triangular distribution.

 6) Based on findings 3), 4) and 5), connection moment between wooden part and core part which is obtained as connection shear force multiplied by equivalent floor length gives tensile force of bolt with some safety margin.

 The disk shear-key is a composite shear resistance system composed of a steel disk and an anchor bolt. The disk shear-key has high stiffness and shear strength as compared with a general post-installed anchor. In the previous study, it is suggested that the expected shear strength can not be demonstrate when tensile force acts on the disk shear-key. Therefore, in the current design and construction method for the external seismic retrofitting, the disk shear-keys are placed in the center of the span and it is assumed that the shear force is to be borne on average. Then, for the tensile force due to the eccentric moment, post-installed anchors are placed at both ends of span. In order to make effective use of the disk shear-key, it is necessary to grasp the mechanical behavior under combined stress and arrange it efficiently. Therefore, in this study, the disk shear-key was tested under cyclic shear force along with constant tensile force.

 In Chapter 2, the outline of the experiment is indicated. The eight specimens were built with various parameters, which is, diameter of the steel disk R_d, embedded length of the anchor bolt l_e, and tensile force ratio η’. Here, the tensile force ratio is the ratio of the tensile axial force to the tensile strength of the disk shear-key.

 The experiment results are described in Chapter 3. The shear strength of the specimens in this experiment was less than the ultimate shear strength according to the current evaluation formula. This is considered to be caused by the different loading condition in the vertical direction from the previous experiment. When the vertical stress generated at the joint surface is zero (η’=0), the shear strength of the disk shear-key can be roughly estimated by the design shear strength according to the current design evaluation formula. However, in the range of η’=0 to 0.5 conducted in this experiment, the relative horizontal displacement of joint at maximum shear strength was within 2mm, which is the allowable value of the external seismic retrofitting joints. Also, the relative vertical displacement of joint increases as the tensile force is applied. Furthermore, in the case of application to external seismic retrofitting, the relative vertical displacement becomes large even when no tensile force is applied. It is suggested that the steel disk slips out, and it is believed that the shear resistance decreases due to the strength reduction in the bearing stress area of the concrete. In addition, the curvature of the anchor bolt is maximized at the position near 2d_a away from the joint surface on the existing side, and the value reaches the yielding region. However, curvatures within the elastic range occur at the other measurement points, and bending deformation hardly occurs at the retrofitting part.

 In Chapter 4, the shear strength evaluation formula under constant tensile force was constructed. The evaluation formula was modified in consideration of slipping out of the steel disk and the reduction of restraint effect by anchor bolt due to the tensile force. It was evaluated almost accurately with the experimental results.

 This paper investigated experimentally the hysteretic behavior of steel panels confined by two pieces of polypropylene plates (hereinafter referred to as PP plates) and bolts, and proposed methods to evaluate ultimate strength and deformation capacity of the confined steel panels. Thirty one 1/2-scale and/or 1/4-scale specimens were made and tested under reversed cyclical lateral loading, with the size, the thickness and the type of steel plates, the thickness of PP plates, and the loading program as main experimental variables.

 Test results have indicated that when confined by PP plates thick enough to prevent the panel from out-of-plane buckling, the steel panels exhibited high lateral resistance and sound energy absorption performance until the drift angle of 0.07 rad without resistance degradation. For the bare steel panels and those confined by thin PP plates, out-of-plane buckling occurred before the drift level of 0.05 rad, but the deformation capacity increased along with the thicknesses of steel plates and PP plates. The specimen having sufficient confinement cracked at the corners of the steel panel at large drift beyond 0.07 rad, and the deformation capacity was independent of the thickness of PP plates. Test results also shown that the loading history might affect the cumulative plastic drift angle at the ultimate state.

 The full plastic moment of the plate section can predict ultimate strength of the steel panels confined by PP plates with sufficient thickness on the safe side, and increasing the yield strength of steels by 10 percent to take into account the effect of strain hardening can enhance accuracy of the theoretical ultimate strength. An equation based on the elastic lateral buckling strength of the steel-PP composite plate was proposed to predict the lateral buckling load of the confined panel without sufficient confinement by PP plates, and gave very good prediction of the measured buckling loads. It is noteworthy that one can determine the minimum thickness of PP plates to prevent the panel from the premature out-of-plane buckling just by letting the lateral resistance (_calQ_bu2), defined by Eq. (3), corresponding to the 10% enhanced full plastic moment, be smaller than the calculated buckling load (_calQ_cr2), defined by Eq. (10).

 Furthermore, the equivalently converted width-thickness ratio ((b/t)_BE, see Eq. (14)) and the balanced (b/t)_BE (=_eq(b/t)_BE, see Eq. (16)) were introduced to evaluate the ultimate accumulative plastic drift angles (γ_u) and the equivalent shear buckling deformation angle (γ_B) of the confined steel panels. Very strong correlations existed between the measured drift angles at ultimate states and the theoretical ones calculated by Eq. (15) and Eq. (20), respectively. The cumulative plastic drift angles at ultimate states of the specimens failing in out-of-plane buckling increased with the decrease of the (d/t)_BE, while those of the specimens failing with crack at the corners of steel plates were almost constant. These observations imply the rationality of the introduced two terms ((b/t)_BE and _eq(b/t)_BE) and the accuracies of the empirical formulae (Eq. (15) and Eq. (20)).

 High-panelized steel sheet shear walls with cold formed edge stiffened burring holes are applied to flat offices, stores, and warehouses in seismically active and typhoon or hurricane prone regions. A 3.53~4.53 m long × 0.455 m wide sheet containing the vertically aligned burring holes (dia.: 200mm) with a pitch of 320mm is hot-dip zinc–alumi–magnesium alloy-coated steel (nominal yield stress: 295N/mm2, thickness: 1.2mm). The burring holes are created by cold pressing a sheet with small-radius holes and contain ribs (curvature radius: 10mm) and cylinders. The edges of the sheets are connected to studs and tracks using drilling screw (dia.: 4.8mm). The end studs are built-up members (□−75×75×2.2: two members + C−150×75×15×3.0) and connected to anchor bolts via tension load connectors. The cross-rails are [−110×50×2.2 and connected to studs to be designed to strengthen the bearing capacities. A configuration with burrs on the inside and smooth on the outside enables the construction of omitting the machining of holes for equipments and thinner walls with simplified attachments of finishings.

 The shear load-story angle relations by in-plane shear experiments showed that the values of shear stiffness without bending effects of 3.53~4.53-m-height walls were almost same and gradually decreasing according to the deformation increasing. The values of ultimate strength of the walls were also almost same, using cross-rails which were installed at equal pitches. The walls that receive the in-plane shear force allowed shear stress to concentrate in intervals between the burring holes and showed no local deformations at story angles until 1/200. The walls at 1/100 story angles showed local waveforms created on the tangents in the intervals between the burring holes, and no large out-of-plane waveform which was effectively prevented by burring ribs.

 Design methods for evaluating the allowable shear loads of the walls at story angles from zero to 1/200, are developed, using the idea of decreasing the band-width of the inclined tension field in the intervals between the edge-stiffened burring holes, which is like post shear buckling behavior that Dr. Basler proposed for plate-girder designs. The design formulas to evaluate the shear loads of the wall at story angle from 1/200 to 1/100 are also developed, using the idea of balanced forces between the horizontal components of the tension in an interval between the edge stiffened burring holes, the compressions of the resisted holes and the horizontal shear forces at a group of screw connections of the sheet and the stud, which are derived from preventing stud deformations by cross-rails.

 In this research, the behavior of steel floor structures, which are composed of steel beams, concrete slabs and bolted connections, under elevated temperatures in fires are numerically simulated using detailed FEM models. The simulations account for the inelastic temperature-dependent material properties and thermal elongations, considering the large deformation. The findings are as follows:

 1) The simulation models of the components of the floor structures, such as beams, slabs and bolted connections were independently validated by comparing with existing experimental data at elevated temperatures. Deflections of composite beams agreed between the results of the experiments and simulations, where beams and slabs are modeled with shell elements.

 2) The created FEM models of reinforced concrete slabs simulate the experiments well. The deflections at the elevated temperatures agreed, taking into account the thermal distribution in the sections. The simulated stress of the reinforcement significantly increased in the center and corners of the slab, where the slab cracked and ruptured in the experiment.

 3) Shear failure of high-strength bolts is a major failure mode of simple bolted connections in fires. The inelastic shear force displacement relationships at elevated temperatures are defined by reviewing the existing experimental data and are applied for the springs of bolts. The principle direction of the shear forces of bolts in beam ends may change with respect to beam sagging and consequently generate tensile axial forces of the beams at elevated temperatures. The created bolt springs return the reactions in the changing principle shear directions. The behavior and shear failure of simple bolted connections at elevated temperatures in the existing experiment are reasonably simulated using the models.

 4) FEM models of a one-bay floor structure composed of primary and secondary beams and concrete slabs are modeled with shell elements. Insulated primary beams located on the perimeter of the rectangular bay are rigidly supported at the ends, and two uninsulated secondary beams are supported by the primary beams. Shear springs for the bolts are used for the simple bolted connections at the secondary beam ends. In the case that the bay is internally located in the plan of a building and the slab edges are rigidly supported at the perimeter, shear failure of the bolted connections of the secondary beams may occur due to thermal expansion under elevated temperatures. Sagging increases, and the slab cracks at the center and corner of the bay. However, catenary effect can maintain the vertical load carrying capacity associated with yielding of the reinforcement, given that the reinforcement is perfectly bonded to the slabs.

 5) The simulations of the one-bay floor structure are conducted under various conditions. It is found that the uninsulated secondary beams may reduce the deflection at the beam temperatures of 600℃ or lower, and the catenary effect of slabs play an important role in carrying the vertical load at higher temperatures. The quantity of slab reinforcement influences the deflection in the catenary effect.

 One of the attractiveness of flat slab structures and flat plate structures is that the structural types without beams have the advantage of a high degree of freedom to use of indoor spaces. The new construction method that we will describe in this study is a flat slab mixed structure with steel capitals in RC slab. The distinctive of this construction is that the columns are generally designed in RC but changed to small diameter steel round bars. In addition, by replacing the RC drop panel and the RC capital with a steel capital in RC slab, the structure will provide a space with a high degree of freedom as well as flexibility.

 The purpose of this research was to experimentally grasp the failure in RC slab by punching shear, the strength and the deformation capacity under lateral load. Furthermore, it is a characteristic of this structure that the decrease in the rigidity of RC slab due to cracking has a great effect on the properties of the structure. Therefore, we will also describe the serviceability and reparability limits by the observation of the residual crack width of the specimens.

 The following conclusions were obtained from this study.

 1) If the circumference obtained by adding the effective depth of RC slab to the capital plate diameter is taken as the critical section for shear in RC slab, it is confirmed that the ultimate vertical force when only the vertical force is transmitted can be calculated with a safety factor of about 10% by using the AIJ standard for RC structure⁶⁾. Here, we use the effective depth is calculated by subtracting the capital plate thickness from the effective depth of the AIJ standard.

 2) By setting the long-term allowable load to about 1/3 of the ultimate vertical force when only the vertical force, it was confirmed from the observation of the crack width that there is no problem in the serviceability of RC slab under the long-term load.

 3) Under the 1/3 (inner column) or 1/6 (outer column) of ultimate vertical force on RC slab, in the lateral loading test, regardless of the inner and outer columns in all the loading directions, it was confirmed that the resistance for lateral force did not decrease up to the story drift angle (“R”) become 20×10^-3 rad, and it was confirmed the high deformation capacity.

 4) The lateral initial stiffness of this method was almost equal when the inner column and outer column specimens were applied in the positive direction. Then the outer column was applied in the negative direction, it was about half that of the positive direction. Furthermore, as the story drift angle increased, the lateral stiffness further reduced due to the damage of RC slab. We also confirmed that the story drift angles of all the specimens were about 10 to 20% of the initial stiffness at R=40×10^-3 rad.

 5) At the story drift angles (R=10×10^-3 rad) often used as the design criteria for large earthquakes, high reparability of this structure was confirmed from the observation of the residual crack width.

 6) The shear failure strength around the column was verified that the correlation equation Vu/Vo+ Mu/Mo≧1.0 of the AIJ standard using RC column is established when punching failure occurs.