Post-installed anchors are available in mechanical or adhesive type. The adhesive type is used often for seismic retrofitting, the installation of HVAC, etc. In the adhesion method, the post-installed anchor, especially the injection type, has high bond strength. Thus, nowadays there is significant research interest in establishing how to use post-installed anchors to support a long-term load.

 Some guidelines regulate the physical properties of the adhesion used for post-installed adhesive anchors. However, the physical properties of the adhesive prescribed in the guidelines differ depending on the guideline. In addition, the original data of properties is unknown. Therefore, in this study, we evaluated the quality and test methods for cartridge-type adhesives used for post-installed adhesive anchors with the injection method.

 We carried out experiments on epoxy resin type, radical reaction type, and cementitious type adhesives used for post-installed anchors. A total of 33 materials were tested. For these adhesives, we carried out three tests.

 First, we studied the new test method to measure adhesive strength. The test specimen is made of a screw bolt and a high nut, and suppose the stress load on the post-installed anchor. We confirmed the influence of the specimen shape on this test method. Test factors are influenced by the adhesion length, the bolt’s axis gap, and the axis angle. Experiments were carried out using two different nut diameters, M16 and M20.

 Second, we carried out compressive strength and compressive modulus tests. The tests were carried out according to the Japanese Industrial Standards. The test specimens were prepared using the method we proposed in a series of studies.

 Third, we carried out a bond test. This is a tensile test of the post-installed anchor as a base of concrete bound to a steel pipe. We carried out this test using five different adhesives.

 The results provided some knowledge. First, the longer the adhesion length, the higher the adhesive strength, and the higher the bolt’s gap axis or the higher the axis angle, the higher the adhesive strength. However, these are enough small to influence the adhesive strength, as long as the specimen is made precisely. Furthermore, the influence of the high nut diameter is small, so all of the tested nut types can be used. However, some adhesives are unfit for use with the M16 high nut, so we propose carrying out the adhesive strength test using an M20 high nut.

 Second, in this study, almost adhesion satisfied a compressive strength of 50 N/mm² and a compressive modulus of 980 N/mm². Thus, we propose that the quality standard for the compressive strength of cartridge-type adhesives should be 50 N/mm², and for compressive modulus 98 N/mm². In addition, a mutual relationship was identified between the results for adhesive strength and compressive strength, or compressive modulus.

 Third, there was a relationship between compressive strength and bond strength, as well as between adhesive strength and bond strength. The ratio of compressive strength and bond strength is about 1:1.5 when using deformed rebar for the bond test, so we propose a quality standard for adhesive strength of 1.5 times the criteria bond strength. In a series of studies, we proposed a standard bond strength of 10 N/mm² or 15 N/mm², so the quality standard for adhesive strength is thus 15 N/mm² or 22.5 N/mm².

 For the seismic design of super high-rise buildings in Japan, nonlinear response analyses are essential. In these analyses, the tangent stiffness proportional damping is often used. This damping model is suitable to express the change of the damping force corresponding to nonlinear condition of buildings, while it overestimates the damping ratios for higher modes. Then, this model underestimates the response and leads to the seismic design to the dangerous side.

 On the other hand, some damping models such as the Rayleigh damping and the modal damping, can treat damping ratios for higher modes properly, while they cannot change the damping force properly corresponding to the nonlinear condition. In this paper, new damping models based on the causal hysteretic damping were proposed to satisfy above demands without excessive calculation time.

 The proposed models are as follows;

 Model A: Applying the response displacement of the building to the causal hysteretic damping model directly, then multiplying the tangent stiffness of each time to evaluate the damping force.

 Model B: Projecting the response displacement to the initial stiffness to remove the residual displacement for the calculation of the causal hysteretic damping model, then multiplying the tangent stiffness of each time to evaluate the damping force.

 For the comparison with these models, the nonlinear damping model is used. This model performs response analyses updating the modal damping matrices for each time using eigenvalue analyses based on the tangent stiffness of each time. The accuracy of this method is very high, while its calculation time is excessive.

 Following results are obtained by the example analyses.

 ・In the case where the Takeda model is used as the nonlinear hysteresis damping, the response results of model A correspond well to those of the nonlinear modal damping model. On the other hands, the response results of model B is a little bit greater than them.

 ・In the case where the tri-linear model is used as the nonlinear hysteresis damping, the response results of linear modal damping, the nonlinear damping model, model A and model B are almost the same. In this case, the linear modal damping model or Rayleigh damping model can be applied as well as proposed models.

 ・For both an impulsive wave and a sinusoidal wave, model A and model B show almost the same tendency as above.

 Recently in Japan, a direction of housing policy is changing from newly construction oriented to stock oriented. In order to form a secondary market by supplying safe, secure and high-quality existing houses, it is desirable to establish a technology of "visualization" that appropriately evaluates the quality and performance of the housing and presents it in an easy way for consumers to understand.

 In this paper, a generalized method is proposed to construct an estimation model for story stiffness of new constructed houses based on design information and microtremor records of multiple houses. As the design information, story weights, wall type, wall length, opening type, etc. are extracted as known information. By referring these known information, story stiffness estimation model is constructed to estimate stiffness of each story and unknown model parameters can be determined so that estimated error to the actual primary eigenvalue frequencies is minimized.

 In this research, the validity of the proposed method is examined by using design information and tremor measurement data of 36 buildings of two-story industrialized houses as an example (total 72 data in both X and Y direction). Based on the identified unknown parameters, the vibration mechanism of real houses in the micro vibration range also can be examine.

 The results obtained are summarized as follows.

 (1) Inverse analysis method is proposed to determine unknown parameters, i.e., wall reference stiffness, stiffness reduction rate due to the wall opening and wall proportion effect to stiffness, etc., to establish stiffness estimation model of new constructed buildings by using known design information, i.e., each floor mass, wall type, wall length, wall opening type, etc.. This procedure can be performed by minimizing the estimated error between the actual measurement value and the theoretical value.

 (2) In addition to stiffness of frame structure, non-structural elements, i.e., partition walls, outer walls, wall openings, and wall proportion and so on, can be directly considered

 (3) It is confirmed that without giving any specifications, it is possible to obtain realistic parameters which is applicable to physical considerations on the reduction coefficient for the wall opening, the nonlinear effect of the wall proportion, the reference stiffness of the outer and partition wall.

 In June 2016, measure to Nankai trough long period ground motions was announced by MLIT. However, the structural calculation using the response spectrum method indicated by the Ministry of Construction notification Vol. 2009 is out of measure because it is difficult to inspect influence of long period ground motions. Also, we proposed a vertical distribution of seismic design load for seismically isolated buildings (SIBs) using the equivalent isolation ratio which has applicability to various dampers.

 In this paper, we analyze the influence of various structural parameters of SIBs on shear force coefficient distribution. In addition, we investigate whether the effect of adding the fluid damper on the shear force coefficient distribution to suppress the response displacement to the long-period ground motions can be expressed by using I_eq. Furthermore, we verify the applicability of the proposed formula for long-period ground motions amplified by the superficial grounds in the actual ground models and propose a new variation correction coefficient CV. The acquired conclusions are as follows;

 1) Based on the analysis results on the long-period ground motions at engineering base layer, it is found that the influence of the setting of the seismic isolation period on the relationship between I_eq. and the equivalent amplification factor of top floor β_eq is large. When the seismic isolation period is long, the moving average value of β_eq in SZ3 and CH3 obtained from the analysis result tend to exceed the Eq.(13). (Fig. 9).

 2) When adding fluid dampers to hysteresis dampers whose burden shear force coefficient α_s is 0.01, Eq.(13) cannot capture the change in β_eq. However, if the burden shear force coefficient α_s is 0.02 or more, Eq.(13) can capture the change in β_eq. (Fig. 10).

 3) The moving average of β_eq for long-period ground motions amplified by the superficial grounds in the actual ground models varies depending on the characteristics of the surface ground and the difference in the seismic motion. However, since the moving average of β_eq calculated collectively from the analytical results of the nine long-period ground motions in all the grounds of Site 1, Site2, and Site3 corresponds well to Eq.(13), it can be said to be able to grasp the response shear force coefficient for nine long-period ground motions on average. (Fig. 13).

 4) The variation coefficient CV of β_eq is determined by Eq.(17) for long-period ground motions amplified by the superficial grounds in the actual ground models. And as shown in Fig. 15, the applicability of amplification factor of vertical distribution of shear force coefficient for seismically isolated buildings to long-period ground motions is confirmed. (Fig. 15).

 1. Introduction

 Recently, the importance of the structural resiliency and functional continuity of buildings affected by earthquakes has become increasingly evident. Studies have reported that damage to the pile foundation often have major impacts on the continued usability of the buildings⁷⁾. However, these studies focused on the seismic performance and response characteristics affected by damages to superstructures⁵⁾. In this study, mid-rise reinforced concrete (RC) buildings having a pile foundation were simulated using the three-dimensional finite element method (3D FEM). Subsequently, the response characteristics of the superstructure and pile were evaluated after the pile foundation was damaged by a large earthquake.

 2. Overview of analysis

 A six-story mid-rise RC building with a foundation made of 5×5 pre-stressed high-strength concrete (PHC) pile was used as the study subject. The superstructure was modeled as a multi-degree-of-freedom system using beam element. Soil and pile were modeled using solid element and beam element, respectively. Linear superstructure behavior and non-linear pile and soil behaviors were modeled in order to focus on the influence of the damage to piles.

 The applied seismic motions were set to the wave notified in the Kobe and Random phase defined as from Level 1 to Level 2 earthquake in the Japanese seismic design codes. The seismic motions were continuously input once or twice by combining each seismic motion to vary the extent of damage to the piles from an earthquake.

 3. Response evaluation of static load analysis and single input analyses

 The stress of the piles varied according to the location of the pile, which is well-known as the pile group effect^15),22). First, the validity of that effect for the analysis model was confirmed by static load analysis. Subsequently, the response characteristics of the superstructure and pile were confirmed when a seismic motion was input only once. The response of the superstructure showed signs of leveling off, while that of the piles intensified with increase in the amplitude level of the input seismic motion. Piles burdened with a large stress due to the pile group effect could reach yield and ultimate state easily.

 4. Response evaluation of series input analyses

 In this chapter, response characteristics were confirmed according to the level of damage to piles when seismic motions were input twice continuously. Further, these characteristics were compared with those for no damage to piles, obtained in chapter 3. In particular, the relation between damage progression of the piles and maximum amplitude of the series seismic motion were studied. Also, the rate of increase in the response of the superstructure according to the damage state of the piles were studied.

 5. Conclusion

 In this study, the response characteristics of a mid-rise pile foundation building were evaluated after the pile foundation was damaged by a large earthquake. The main results obtained were as follows:

 1) Increasing damage progression of the piles was confirmed even though the amplitude of aftershock is less than or equal to that of foreshock because of increasing soil response and decreasing energy dissipation of pile with serious damage of a part of piles according to the foreshock.

 2) The natural period of the coupled soil-building system and that of the building conspicuously increased when 40% of the piles attained the ultimate state with fracturing even though the nonlinearities of the soil are considered.

 3) The response of superstructure increased and decreased with the changes of the natural period of the coupled soil-building system and frequency response characteristics of the series earthquake because of decline in influence of energy dissipation of pile with the fracturing of piles.

 1. Introduction

 Many types of steel pipe piles with steel spiral or flat wings welded to the pile tip or pile shaft have been developed and put to practical use. Piles with wings are generally driven into the ground by giving torque to the pile shaft during installation. Compared to cast-in-place concrete piles and embedded piles, piles with wings can reduce construction by-products because waste soil reduces. Recently, diffusion of piles with large wings has expanded the range of applications to large buildings. However, the wings that has diameter larger than the pile diameter are known to loosen soil around the pile shaft. The loosened soil may affect the lateral subgrade reaction of the pile. In this study, a cyclic lateral loading test was conducted on a pile with two flat plate-shaped wings in order to understand the following properties for lateral resistance. 1) Effect of the presence or absence of wings, 2) Effect of differences in wing diameter ratio, 3) Effect of aging after installation, 4) Grasp the secant stiffness and hysteresis damping of the steel pipe pile with wings subgrade reaction.

 2. Method

 This experiment was conducted in Ibaraki Prefecture. The soil deposit at the test site consisted of weak volcanic ash soil. From the ground surface to 0.55 m below ground level (GL) is andosol, from GL-0.55m to GL-2.2m is weathered volcanic ash, from GL-2.2m to GL-3.2m is volcanic ash, and beyond GL-3.2m is sand. Four types of piles were tested. For the test pile, a JIS G 3444 general structural carbon steel (STK) pipe of outer diameter φ 267.4 mm × material thickness t 9.3 mm (steel grade 490) was used. The pile with wings was fabricated by welding two flat plates to the tip of the steel pipe while preparing the edge such that the inclination angle of each wing became 10°. The test piles had a shaft length of 5 m. To measure the bending strain of the pile, strain gauges were attached on the inner surface of the pile. The loading device consisted of a test pile, reaction pile, and hydraulic jack. For the reaction force, the pile heads of these four reaction piles were combined in a square form with H-shaped steel. Test piles are allowed free rotation of the pile head. The lateral displacement of the test pile was measured at two points: the load height and GL+100 mm. The lateral displacement at the position of at GL+100 mm is referred to as "reference displacement". The experiment controlled by reference displacement.

 3. Conclusions

 The major findings obtained from this study are summarized as follows:

 (1) The lateral resistance of the pile with wings was smaller than that of the pile without wings because of loosening soil around pile by the wings passing.

 (2) The difference in wing diameter between 2.3D_p and 3.0D_p had little effect on the lateral resistance of the piles. It is considered that the peeling depth has a more dominant influence than the passing region of wings as a factor.

 (3) Even one year after construction, the loosened soil around the pile with wings affected its lateral resistance.

 (4) The secant stiffness of the piles with wings was smaller than that of the pile without wings. Secant stiffness ratio of those piles became larger depending on the pile displacement and was influenced by ground depth and soil type. The hysteresis damping of the pile with wings had little influence by the pile displacement and soil type.

 In 2011, some roof bearings of gymnasiums suffered severe damage by the pacific coast of Tohoku earthquake. Typical damage was the elongation and break of the anchor bolts or crush of base mortar. Since gymnasiums are used as shelters for the neighboring residents in earthquake disasters, damage prevention of roof bearings are required.

 Roof bearings are the connections between the lower RC frame and the steel roof. They are similar to conventional steel exposed column bases. However, the dominant stress and the mechanism to resist the horizontal forces of the roof bearings are different from those of steel exposed column bases. The dominant stress of steel exposed column bases is axial and bending stress while shear stress of the anchor bolts is dominant for roof bearings since the vertical load is extremely lower than that in steel exposed column bases in ordinary buildings.

 In present paper, the full-scale cyclic shear loading tests of roof bearings were carried out to investigate the fracture behavior of roof bearings and the restoring force characteristics were also evaluated. Three different height (20mm, 50mm or 100mm) of the base mortar were used to compare the fracture behavior and the restoring force characteristics. Moreover, the friction tests were conducted to investigate the friction coefficient between the base mortar and base plate.

 It was found from the results of friction tests that the friction coefficient between the base mortar and the base plate was about 0.5. This result agrees very well with the friction coefficient given in the AIJ recommendation for Design of Connections in Steel Structures.

 The fracture behavior and the restoring force characteristics of roof bearings are summarized as follows. The anchor bolts were leaned after the base plate contacted the anchor bolts. After that, the base mortar cracked since the deformed anchor bolts pushed it out. After the tests, the base mortar was crushed especially outside of anchor bolts. When the base mortar height is 100mm, the test body couldn’t support the vertical load since the base mortar located in inside of anchor bolts was crushed. However, the friction coefficient of base mortar was approximately 0.5, and we assumed that the restoring force characteristics of roof bearings is the summation of strength of anchor bolts and the friction force between the base mortar and base plate.

 The anchor bolts and the base mortar were already severely damaged before the horizontal load reached the yield shear strength. Meanwhile, we observed that the anchor bolts were bended at two parts, the bottom of the base plate and the slightly below of the surface of the concrete. For the small deformation, we proposed that the restoring force characteristics of roof bearings is evaluated by the yield bending strength considering this bending mechanism. For the large deformation, we proposed that the restoring force characteristics of roof bearings is evaluated by considering the bending and axial stress of anchor bolts. These stress of anchor bolts were calculated by the generalized plastic hinge theory applying kinematic hardening.

 Latticed shells made of H-shaped steel have been constructed all over the world. This study investigates elastic buckling of spherical latticed shells made of H-shaped steel. Out-of-plane shell-like buckling and in-plane member buckling dominantly occur in the elastic buckling of rigidly jointed latticed shells. An H-shaped section has strong and weak axes. Since bending stiffness on the strong axis is several times of that on the weak axis, in-plane member buckling is more likely to occur than the buckling of the shells made of pipes.

 In the present study, the latticed shells for study were on the square plan of 20m x 20m. The shells were relatively flat where the rise-span ratios of the two central arcs were between 0.075 and 0.125. In the linear-buckling (eigenvalue) analysis, in-plane member buckling appears in the 1^st buckling mode in all the FE models. The depth, area and second moment of inertia on the strong axis were uniformly set to 20cm, 25cm² and 2000cm⁴, respectively. On the other hand, the second moment of inertia on the weak axis was parametrically changed so that the higher order linear buckling load where shell-like buckling firstly appeared was 1.5, 2.0 and 2.5 times of the 1^st linear member buckling load. In addition, two boundary conditions, pin and roller were assumed since roller is often used in actually constructed shells.

 The higher order linear buckling loads agreed well with the classical shell-like buckling loads calculated using the continuum analogy.

 To estimate member buckling load itself, the buckling length is required. The buckling length calculated using the linear buckling analysis, is 0.7 and 0.5 times of the member length for the roller and pin supported shells, respectively. In the pin-supported shells, uniform compression is realized but in the roller-supported shells, compression and tension mix especially on the peripheral where the member buckling occurs. The tensioned members connected to the compressed member interfere the buckling of compressed member.

 In the elastic buckling analysis considering geometrical nonlinearity, shell-like buckling firstly occurred when the higher-order linear shell-like buckling loads were over 2 times of the 1^st member buckling load on the roller-supported shells and over 1.5 times on the pin-supported shells. This reverse of buckling type is due to significant geometric nonlinearity in the shell-like buckling. The knockdown factor was approximately 0.7 and 0.4 for the pin and roller supported shells, respectively. But in the member buckling, it was almost over 0.9.

 For latticed shells made of pipe, the shell-likeness factor was already proposed as an index to estimate the buckling type in elastic nonlinear buckling. Here it was modified for the application to those made of H-shaped steel. When it is over 4.86, shell-like buckling occurs.

 In recent years, there is a trend of wooden construction of non-residential building. We are also developing projects with industry-university collaboration, fusing CFRP and timber, and developing to realize better buildings. Timber is an anisotropic material. The grain direction has sufficient mechanical performances as a building structural materials, but the mechanical performances in the direction perpendicular to the grain is never sufficient as a building structural material. In buildings, members of shaft material are combined at right angles to assemble a rectangular parallelepiped structure with a space inside. For this reason, a force flow in the grain perpendicular direction always occurs locally at the joint portion between the members. In non-residential construction, it may be difficult to process the force flow in the direction perpendicular to the grain. How to improve the bearing performance in the direction perpendicular to the grain, which is a weak point of wooden buildings, is considered to be one of the key technologies for wooden non-residential buildings including middle- and high-rise buildings. There are only a few studies on how to prevent snap-in or improve bearing performance, and at the practical design level, processing is carried out by increasing the bearing stress areas.

 Therefore, in this research, we proposed a method of laminating and bonding carbon fiber reinforced plastic (CFRP) on the bearing pressure surface of wood by adhesive and reported verification experiments of its reinforcing effect. In addition, this reinforcement method is a technology that can be applied directly to AFRW (Advanced Fiber Reinforced Wood) reinforced with high performance fiber reinforced plastic that we are separately developing.

 We demonstrated the effectiveness of the reinforcement effect by experiments with various parameters set. We can grasp the trend of six conditions of thickness of wood, length of wood, length of pressure plate, thickness of reinforcing CFRP, length of CFRP for reinforcement, adhesion between CFRP for reinforcement and wood, test parameters were set. The following findings were obtained from the experimental results and their examinations.

 1) There was a large reinforcement effect on bearing yield strength, about 2 times when the pressure plate length was 100 mm, and about 4 times reinforcement effect when the bearing pressure plate length was 30 mm.

 2) When the thickness of the CFRP was 7 mm, there was a reinforcement effect with a bearing stiffness of about 1.2 times and ultimate strength of about 1.5 times.

 3) When the bearing pressure area is small, a particularly large reinforcing effect can be expected.

 4) By using the concept of effective penetration length, the tendency of wooden bearing performance reinforced with CFRP by the existing AIJ’s formula was roughly caught. Therefore, as a reinforcement mechanism, CFRP increases apparently effective bearing pressure area, it was thought that it was leading to an improvement.

 To mitigate seismic damages of traditional wooden buildings in Japan, it is important to evaluate aseismic performance of the traditional wooden buildings properly and conduct proper aseismic reinforcement. Major aseismic elements of many traditional wooden buildings are hanging walls. If strong ground motions occur, the columns next to the hanging walls may be broken and the buildings result in collapse. Therefore, it is important to evaluate whether the columns break or not using the material properties such as bending Young’s modulus and bending strength.

 To evaluate the material properties of each column in the traditional wooden buildings, non-destructive tests are desirable because many buildings are regarded as cultural properties. In our previous study²⁾, we have proposed a non-destructive test using impact elastic wave to measure wave velocity of wood. Moreover, we have found preferable correlation between the bending Young’s modulus and the dynamic modulus of elasticity which was calculated from the wave velocity and the density of the wood. Nevertheless, we did not propose a material property evaluation method to be used for the aseismic performance evaluation of traditional wooden buildings in our previous study. In this study, we propose the evaluation method of bending Young’s modulus and bending strength of columns.

 On the other hand, there are cases where the columns and beams in the traditional wooden buildings are replaced totally or partially because of degradation. Therefore, in our propose method, we make small specimens from these replaced wooden members and conduct both impact elastic wave test and bending test to the small specimens. From these test results, we conduct regression analysis and construct the relation expressions of the material properties. Then, the material properties of the columns in the buildings are evaluated to apply the wave velocity of the columns to the relation expressions.

 Our proposed method is summarized in three steps as follows.

 A) Field tests to columns in the buildings

 Impact elastic wave test to the columns in the building is conducted to obtain the wave velocity of the columns.

 B) Laboratory tests to small specimens

 Impact elastic wave test, bending test and density measurement are conducted to obtain the material properties.

 Regression analysis are also conducted to construct the relation expression.

 C) Evaluation of material properties of the columns

 The material properties are evaluated using the wave velocity from A) and the material properties from B).

 We will evaluate the material properties of the columns without laboratory tests to accumulate the relation expressions of the material properties of various kinds of tree species.

 The P-Δ effect caused by additional moment due to gravity load produces when the building responses by earthquakes, and produces the influence on the earthquake resistance properties particularly for buildings with a small base shear coefficient. The purpose of this paper is to investigate the P-Δ effect on the deformation of ordinary high-rise buildings by earthquake response analysis using three types of input seismic motions (sweep wave giving stationary loop, earthquakes specified in the notification and actual seismic wave), and to derive response evaluation formula considering P-Δ effect.

 The analytical model of the structure is the modified fish-bone frame model of a 14-story reinforced concrete building expressing infinite even-span. The investigation and comparison were made by changing the value of the base shear coefficient in the range of about 0.05 to 0.4.

 The major findings obtained in this paper were as follows.

 1) For the sweep wave giving stationary loop, a clear reduction of the response due to the P-Δ effect was confirmed for all cases of the analytical models. The smaller the base shear coefficient was, the more significant this tendency was.

 2) For the earthquake ground motions for the notification and observed seismic waves, three were cases where the response displacement decreased or increased by considering P-Δ effect.

 3) The equivalent damping factor at the maximum response was calculated. Their values were greater when the P-Δ effect was considered because the potential energy of the equivalent linear model became smaller due to the decrease in the restoring shear force, although the energy dissipation was the same for both cases.

 4) The response deformations decreased for the sweep wave because the response reduction effect due to the increase in the equivalent damping constant was greater than the effect of decrease in strength due to the P-Δ effect. On the other hand, there were both cases in which the response displacement decreased or increased depending on the degree of increase in equivalent damping factor and decrease in restoring shear force, for non-stationary waves.

 5) The response evaluation formula considering P-Δ effect in the ordinary high-rise building was proposed. It can be said that the proposed evaluation formula gave appropriately evaluation of the response displacement considering P-Δ effect. However, P-Δ effect produced little difference in response deformation, in case of non-stationary response.

 In Tohoku earthquake (March 11, 2011), highrise steel buildings of the Tokyo downtown area shook for a long moment under the influence of long-period ground motion. To prevent damage under such ground motion, evaluation on deformation capacity of steel member (welded beam end, etc.) has been required under multi-cycle loading.

 In this paper, deformation capacity of the welded beam-end is assessed by the FE analysis, considering fracture rule, which adopts both cyclic damage rule and monotonic damage rule. The cyclic damage rule is a fatigue damage law based on Continuum Damage Mechanics (CDM), and the monotonic damage rule is a damage rule for a large ductility factor equivalent to monotonic loading. The validity of this fracture rule is verified by the simulation of past experiments under cyclic loading. Two past experiments are selected. One is an element experiment using specimens modeling beam flange and web, in which tensile and compressive repeated force is applied in the axial direction. The other is a partial frame experiment using field welding type specimen having a ¼ circular weld access hole of R23 at the beam-end, in which increasing cyclic force is applied. For an element experiment, simulation analysis of monotonic force and constant amplitude repeated force is performed using FEM with solid elements and shell elements. As a result of the FE analysis, we can simulate the reduction of peak load, and it is found that the deformation capacity under multiple cyclic loading conditions can be evaluated. For a partial frame experiment, simulation analysis using shell elements is carried out. As a result of the FE analysis, it is found that a crack occurs at the toe of the weld access hole on the lower flange, and propagates in the flange width direction. Using FEM it is possible to simulate the situation where the welded beam-end fractures, and it has been verified that the load and deformation relationship of FE analysis had good agreement with test results.

 In addition, we carries out a comparison between numerical results and the performance curves of the welded beam-end proposed by MLIT project (the project name; 27-2) and JSSC committee. It is demonstrated that a performance curve is possible to be evaluated considering effect of beam-end details using the proposed numerical method.

 In this study, T shaped tensile tests and 3-point tensile tests are conducted. The purpose of these experiments is to reveal the relationships between brittle fracture of spheroidal graphite cast iron and fracture critical stress in various stress conditions. Fig. 1 shows the application method of fracture critical stress for brittle fracture. Bending specimens are made of the T shaped castings shown in Fig. 2 by cutting. Specimens of round bar tensile tests is shown in the Fig. 3. The results are indicated in the Table1 and Fig. 4. Fig. 4 is Stress-strain relationships of T shaped castings.

 Fig. 5 presents the T-shaped tensile bending test specimen and displacement measuring position. It is measured vertical displacement of the end of specimen, and calculated the deformation angle of the end. Fig. 6 presents the specimen case of three-point bending. PR shaped specimen has the protuberance. It is 5 mm. PL shaped specimen has no protuberances. All of these shaped specimens are composed of width 35 mm and 125 mm. The narrow width specimens are supposed to be plane stress condition, and the wide width specimens are supposed to be plane strain conditions.

 Fig. 7 shows the Lord-deformation relationships and Bending stress-deformation relationships. Bending stress was calculated by the formula (1) and (2). The wide width specimens broke earlier deformation angle than the narrow width specimen. The list of bending test results are indicated in the table2.

 Notch round bar tensile test was conducted to calculate fracture critical stress. The test specimen and analysis model are shown in the Fig. 9 and Fig. 10. The list of notch round bar tensile test is shown in the table3.

 Fig. 12 shows the analysis model of T shaped specimen. As indicated in the Fig. 13, the Lord-deformation relationships are relative to experimental results. The difference between the maximum first principal stress of bending analysis and fracture critical stress was under 15%. It is shown in the Table4.

 From the results of test and analysis, following findings were obtained.

 (1) The wide width specimens were broken smaller deformation than the narrow width ones. That is because the wide width specimen is likely plane strain. This is why, when the same deformation with narrow width specimens, the wide ones has higher stress.

 (2) The difference between the maximum first principal stress of bending analysis and fracture critical stress was under 15%. Therefore, in the various stress conditions, fracture critical stress is relative to brittle fracture of spheroidal graphite cast iron. It is necessary to set the safety factor of fracture critical stress when fracture critical stress is applied to actual structures.

 (3) CS specimens (the tensile side is casting surface) were broken earlier than MS specimens (the tensile side is machining surface). However, it is necessary to consider the effects of cast iron surface for fracture, because the number of CS and MS specimens are not enough.

 In this study, a hysteresis damper using bent steel plates for base-isolated structures (BSPD) is proposed. The BSPD comprise four steel plates bent into a trapezoid. The bent steel plates are connected together in the shape of a cross to reduce the influence of the loading direction on mechanical behavior. In chapter 2 and 3, the cyclic shear loading tests are conducted to investigate the influence of the shape of the bent steel plates on mechanical behavior of the BSPD. The conclusions of this chapter are as follows.

 1. The loading direction has little influence on the load-deformation relationship including the yield strength and initial stiffness.

 2. Failure mode are classified according to the load amplitude and direction. In all failure modes, the crack occurs near the bolt holes.

 1) d=100mm, 0°: The bent steel plates parallel to the load are broken.

 2) d=100mm, 45°: The crack propagates perpendicularly to the load.

 3) d=100mm, 0°: The steel plates parallel or perpendicular to the load are broken.

 4) d=200mm, 45°: The steel plates perpendicular to the load are broken.

 3. The yield strength _iQ_y and initial stiffness _iK_ini increase with thickness t. The deformation capacity decreases as t increases. The diagonal plate length h has little influence on _iQ_y and _iK_ini. The deformation capacity increases with h. The diagonal plate height g has little influence on _iQ_y, _iK_ini and the deformation capacity.

 4. The specimen of the steel type SN490B has larger strength as the specimen of the steel type LY225. However, the specimen of the steel type LY225 has larger deformation capacity as the specimen of the steel type SN490B.

 5. The deformation capacity n_t is correlated with the plate spring stiffness _iK, and n_t increases as _iK decreases. Due to the original experimental plan, there are bolt holes in the bent steel plate. However, in the experiment, the crack occurs near the bolt holes, which led to the strength deterioration. Therefore, in chapter 4, finite elements (FE) analysis are conducted to investigate the influence of the bolt holes on deformation capacity. The conclusion of this chapter is as follow.

 6. The FE analysis indicates that the bolt holes make the local strain increase, and to eliminate the bolt holes is effective for improving the deformation capacity.

 1. Introduction

 Nowadays, there is a move to legislate the examination the performance of the buildings against long-period ground motion earthquake. Systematic experimental study of CFT columns has been desired for constructing the evaluation method for the strength deterioration behavior because there are many parameters affect the structural performance and strength deterioration of CFT columns. The objective of this study is to clarify the cyclic behavior of square CFT beam-columns subjected to cyclic lateral load under constant axial load with constant displacement.

 2. Outline of test

 Test parameters are effective length-section depth ratio l_k/D (20, 14, 10, 8), axial force ratio n (0.15, 0.3 0.45 0.6) and lateral rotational angle amplitude R₀ (0.5, 0.75, 1, 1.25, 1.5, 2, 3). The loading program is monotonic loading and cyclic loading. The number of test specimens is fourteen for monotonic loading and thirty four for cyclic loading. The width-thickness ratio is 25 and the material of the steel tube is BCR295. The compressive strength of concrete is about 60~78N/mm².

 3. Test results and discussion

 Lateral load and rotation angle relationships were shown and the effects of the test parameters on strength deterioration behavior were clarified. The relationship between axial strain behavior and the strength deterioration behavior was discussed. When the rotation angle R₀=1%, l_k/D =20 and 14 with n=0.6 specimens and l_k/D =10 and 8 with n ≧ 0.45 specimens, the axial strain dramatically decreased and the lateral strength decreased to 80% of the maximum strength.

 4. Discussion of test strength

 The strength when the part of the specimen was yielding was compared with the allowable strength for short-term loading shown in CFT Recommendations. The yield strength agrees well to the allowable strength. The relationship between the strength and the strength deterioration was discussed.

 5. Conclusions

 The conclusions derived from this study are as follows:

 1) In monotonic loading, test strength reached the full plastic moment except for the l_k/D=14 and n=0.6 specimen (Fig. 8).

 2) Strength deterioration is larger as the axial force ratio becomes larger and the effective length-section depth ratio becomes smaller. As the rotation angle increased, the strength decreased more. Strength of hollow steel tube specimen decreased more than that of CFT specimens in case of n=0.6 (Fig. 9).

 3) When the rotation angle R₀=1%, l_k/D =20 and 14 with n=0.6 specimens and l_k/D =10 and 8 with n ≧ 0.45 specimens, the axial strain dramatically decreased.

 4) When the rotation angle R₀=1%, l_k/D =20 and 14 with n=0.6 specimens and l_k/D =10 and 8 with n ≧ 0.45 specimens, the lateral strength decreased to 80% of the maximum strength.

 5) Strength could keep 95% of the maximum strength after 100 cycles when the specimens R₀/R_pc (R_pc is the elastic limit rotation angle) is smaller than one and n is equal and smaller than 0.45.

 6) Strength of the specimen with l_k/D=20, n=0.15 and R=1% did not reach the allowable strength for short-term loading and did not decrease even after 300 cycle loading. In specimens whose maximum bending moment at the bottom of the column reaches the full plastic moment, the lateral strength decreased to 80% of the maximum strength.

 The advantages of concrete-filled steel-tube (CFT) columns include high strength and remarkable ductility, since the steel tube provides confinement to the concrete while the concrete prevents local buckling of the steel tube. CFT column composite frame systems with steel beams have been widely used in moment-resisting frame systems for mainly office buildings. The CFT columns at typical floors are short and have small ratios of buckling length to cross section depth, and those at entrances of lower floors are slender and have large ratios of buckling length to cross section depth. CFT columns from short to slender have been used in buildings. In order to verify the structural safety of CFT structures in the ultimate state against huge earthquakes, etc., a restoring force characteristic model incorporating a strength reduction after ultimate strength is required. The stress states of CFT columns applied to buildings are in a state of flexural-shear stress, and the structures are designed for the predominant flexural yielding. It is thought that accurate evaluation of a shearing force-deformation relationship under axial force leads to design of columns that have the features of CFT columns. For a shearing force-rotation angle relationship, which is the basis of the restoring force characteristic model of CFT columns from short to slender, it is important to devise a simple model incorporating a reduction of strength after ultimate strength.

 This paper proposes a new simplified model of a shearing force-rotation angle relationship for square CFT beam-columns under monotonic flexural-shear loading to estimate the elasto-plastic flexural-shear behavior of beam-columns from short to slender. The square CFT beam-column model incorporates a reduction in strength after ultimate strength. The proposed simplified model is a multi-linear model having a crack strength point, a yield strength point, an ultimate strength point and strength reduction points. In order to evaluate flexural-shear behavior of short to slender columns with a single model, the column member is divided into elasto-plastic parts at the column ends and a remaining elastic part. For the elasto-plastic part, the rotation angle of each point is calculated using the bending moment-curvature relation model proposed by the author and a newly proposed evaluation formula for curvature magnification. Predictions from the proposed model of the shearing force-rotation angle are found to agree approximately with experimental results up to large rotation angles.

 To investigate stress and deformation behavior of steel structure in fire, it is important to consider behavior of friction type high strength bolted joints during not only heating phase but also cooling phase. The purpose of this study is to clarify stress and deformation behavior of the joints during cooling phase in fire. Failure of the joints may occur due to tensile forces of the connected beam under cooling phase during decay period in fire. Furthermore, strength degradation of the high strength bolt steel which experienced high temperature cause the bolt in shear failure in fire. However, there was few research with the test which focused on the deformation behavior of the joint during cooling phase.

 This paper reports on results of tensile tests of friction type high strength bolted joints during cooling phase and the numerical analysis on the basis of component based model. Post-fire-heating tensile tests were carried out in order to obtain the maximum strength of the joint after fire. This test was carried out on 6 specimens of the joint. The main test parameters were steel temperature and main plate thickness. Cooling-phase restraint tests, which simulated the development of tensile force during cooling phase, were carried out on 2 specimens of the joint. In addition, finite element analysis on the basis of component based model were carried out to discuss the behavior of the joint during cooling phase. The follow conclusions could be indicated from this research.

 1) Result of post-fire-heating tensile tests indicated that failure mode of the friction type high strength bolted joint could be changed from plate tear-out failure to bolt in shear failure. This is caused that strength of the bolt decrease considerably after fire heating. Maximum tensile resistance of the tests approximately agreed with the calculated result on the basis of the tensile tests of the employed steel plate and bolt.

 2) In case of the specimen which designed to investigate the behavior of plate tear-out failure, the sufficient deformation capacity was confirmed because of the ductility of the plates with a bolt hole. To prevent failure of the joint during cooling phase in fire, the tear-out failure of the steel plate is better than the bolt in shear failure.

 3) In case of the specimen which has much thickness of the main plate, the specimen failed with bolt in shear failure due to development of the tensile force at the joint during cooling phase. Damage of the bolt at high temperature affected maximum strength of the joint during cooling phase.

 4) The numerical analysis result of the joint with plate tear-out failure approximately agreed with the test result on increasement of the tensile force due to shrinkage of steel and restraint of axial displacement at both ends during cooling phase.

 5) Influence of shear deformation of the bolt at elevated temperature on the resistance of the joint during cooling phase was indicated by the numerical analysis results of the joint with bolt in shear failure.