When a super tall building is designed, a member-to-member model is constructed to represent the assembly of columns and beams. The response history analysis using member-to-member model, however, is computationally expensive and produces excessive amount of data that are not necessary at the preliminary design stage. Therefore, an equivalent mass-spring model with a small number of elements and short analysis time is useful at the initial design stage. It promotes accurate and quick decision on the overall performance and economy of the super tall building at the critical stage where many different design options are explored. In addition to the shear deformation typically considered, bending deformation becomes large as building becomes taller, which must be simulated by the equivalent mass-spring model. In this paper, we propose a new modeling method for an equivalent mass-spring model, the so-called “bending-shear model” that can accurately reproduce the dynamic characteristics of the building by using the information obtained from the member-to-member model. The bending stiffness and shear stiffness are calculated so that the 1st mode frequency and vector as well as the 2nd mode frequency of the bending-shear model match with those of the member-to-member model. Instead of typical static lateral load, pure bending moment is applied to the member-to-member model, and bending stiffness is estimated and calibrated for the bending-shear-model. The displacement and acceleration responses of the bending-shear model agree well with those of the member-to-member model.

 After technical standards for CLT buildings were established in 2016, building structures using CLT are rapidly spreading in Japan. In Japanese CLT buildings, χ-marked connectors which conforms to national technical standards are commonly used for each CLT joint. There are two main types of χ-marked connector: screw joint and drift pin joint. Of these, the former is used for almost all kind of joints such as tension or shear joints between wall leg and concrete base, shear joints between wall panel and hanging or waist wall, or band joints that join floor panels together. Therefore, it is basically impossible to build CLT buildings without screw joints. Since screw joints can be reworked by screw re-driving, it is highly likely that screw re-driving will be carried out on site for reasons such as adjustment of construction accuracy, construction position, or incorrect construction order. Under Japanese building code, screw re-driving is not prohibited, but there is little research on the effect of screw re-driving on the strength and the stiffness of joints. Despite the fact that there is no clear evidence as to whether or not the screw re-driving impairs the structural soundness of the joint, the screw re-driving can be done based on on-site judgment due to experiences of structural designer or site manager. This situation is not very good, and experimental verification should be performed on the effect of screw re-driving on the structural soundness of the joint.

 In this study, tensile and compression tests are performed on specimens of TB-90 joints, which is a χ-marked tensile connector that connects the concrete base and the wall leg of the 3-layer 3-ply 90mm thick CLT panel. The effect of the screw re-driving on the strength and the stiffness of the CLT joint will be clarified by conducting a comparative experiment between “the normal test piece”, “the test piece at which screw re-driving is done” and “the test piece which has been repaired with epoxy resin after screw re-driving”.

 The following conclusions were obtained from the experimental results.

 • The screw re-driving at the CLT tensile joint (TB-90) has little effect on the strength and the fracture mode of the joint.

 • If the number of screw re-driving is one, the stiffness of the joint is hardly affected.

 • When the number of times of the screw re-driving is 10, the initial stiffness of the joint is lower than that of the specimen without screw re-driving due to the expansion of the screw hole.

 • Even if the repair material mainly composed of epoxy resin is applied to the screws, the strength of the joint is hardly affected.

 • When a repair material consisting mainly of epoxy resin is applied to the screws, the stiffness reduction due to the screw embedding is delayed and the initial stiffness is increased. In addition, the stiffness until the screw yields after screw embedding is also slightly increased.

 However, the experiment in this study was only conducted for TB-90. It is difficult to draw a universal conclusion because there are only a few specimens and only a static force experiment was conducted in this study. Since it is possible that screw re-driving on site can be done sufficiently for all kind of joints, more knowledge needs to be accumulated in the future.

 The purpose of this study is to propose a formula for estimating the vertical deflection of the floor when a concentrated load is applied on the beam of a thick-plywood subfloor.

 In Chapter 1, we described the background and purpose of the research, and steps of the research.

 In Chapter 2, we described the construction of an analytical model for a thick-plywood subfloor. First, in order to create an analytical model of the thick-plywood subfloor, we conducted an experiment in which a concentrated load of 1 kN was loaded on the beam for the thick-plywood subfloor with a planar shape of 5.46m x 3.64m. The vertical deflection distribution of the thick-plywood subfloor through the experiments was figured out. Next, an analytical model of the thick-plywood subfloor that reproduces the test results was generated with an analysis-software by finite element method. Finally, the constructed analytical model was verified that reproduced the vertical deformation of the thick-plywood subfloor.

 In Chapter 3, the influence of various factors affecting the vertical deflection of the thick-plywood subfloor was examined analytically. Fig. 4 shows the results.

 In Chapter 4, we conducted a parameter study. In the analysis, various influence factors such as the bending stiffness of the beam, the span of the beam, the Young's modulus of the plywood, and the shear stiffness of the wood screw that fastens the plywood to the beam were changed. Analytically, we examined the influence of various factors. Fig. 5 shows the effect of various parameters on the vertical deflection of the thick-plywood subfloor.

 In Chapter 5, based on the influence of various factors obtained from the analysis, we derived a formula for predicting vertical deflection of the thick-plywood subfloor that takes into account the effects of both composite action and two-way action. The prediction formulas for the vertical deflection in case concentrated load is applied on the beam of on the thick-plywood subfloor were shown as Equations (1) to (3).

 In Chapter 6, the applicability of the prediction formula for vertical deflection of the thick-plywood subfloors was examined for two types of floors in real houses and three types of full-scale floors. The estimated vertical deflections of the full-scale floors ranged from 0.81 to 1.16 for experimental results.

 In Chapter 7, a summary of this study was stated.

 Buckling-restrained braces (BRBs) pose predictable and stable hysteretic behavior with excellent energy dissipation capacity. However, when such BRBs are used in steel frames with pinned connections, damage concentration and large residual deformations may occur due to their low post-yield stiffness. In overseas, post-tensioned cables are added on the BRBs to provide the self-centering force and eliminate residual deformations following a strong earthquake. Hysteretic curves of BRBs with post-tensioned cables can be defined with two parameters, namely post-yield stiffness ratio (α) and energy dissipation ratio (β). Previous studies have aimed to change β from the bilinear (β=2.0) to flag-shaped (β=1.0) hysteresis to have zero residual deformations. However, the flag-shaped behavior may require a significant post-tensioning force due to strain and compressive hardening. As such, limited energy dissipation capacity increase the peak displacement responses. Although previous numerical studies have found that residual deformations are small enough even hysteresis is not flag-shaped, an optimal range of α and β is not well studied. A newly developed BRB named as PT-BRB is introduced in this paper to provide self-centering force by adding post-tensioned cables while the BRB core dissipates input energy. Carbon Fiber Composite Cables (CFCCs) are used as post-tensioned cables and performance of the brace is evaluated experimentally. Section 2 presents the developed PT-BRB configuration and mechanism, and discusses the appropriate α and β ranges by conducting some numerical analysis. In Section 3, a 1/3 scale specimen of the PT-BRBs are actually manufactured, and cyclic loading tests with various post-tensioning forces are performed to confirm the hysteretic properties and deformation capacities. A numerical model that reproduces the behavior of each part of PT-BRB for application to the 3D model is proposed in Section 4. It is found that numerical values agree very well with the experimental results. Note this paper presents the results of a joint research between Japan and Turkey, and the contents conform to the US standards and design guidelines.

 In summary, the following results were obtained:

 1) In the SDOF model, the increase of α is effective in reducing residual deformations in the range of α<10%, but the slope of decrease is small when α is higher than 10%. Peak acceleration and brace axial force increases when α increases. Additionally, when the energy dissipation ratio β is decreased from β=2.0 to β=1.5, the residual deformations decrease significantly.

 2) In the 6-story MDOF model, it was shown that partially self-centering (α=5.6%, β=1.8) can minimize the residual deformation even though the maximum deformation and brace axial force are the same as bilinear case. It was also confirmed that a uniform story drift distribution can be expected by increasing the restoring force for this model.

 3) Partially self-centering behaviors (1.0<β<2.0) are obtained from the cyclic loading tests and, all PT-BRBs showed stable hysteresis. PT-BRB-16 with the smallest post-tensioning force reached up to 3% story drift. In all specimens, CFCCs remained elastic until the core plate has fractured.

 4) A numerical model of PT-BRB is constructed by connecting the Core, Inner Tube, Outer Tube and CFCC as individual element with contact element. This model accurately captures the experimental results and track the actual hysteretic behavior.

 1. Introduction

 For the purpose of not only simplifying fabrication of steel members but also improving the efficiency of transportation of steel members to the construction site, high strength steel box-section column-to-beam connections which have small projection exterior diaphragms using thicker steel plate than conventional one have been developed. By now, mechanical behavior of the connections was verified based on loading tests and finite element analysis. However, these studies targeted the connections without eccentricity of centroid lines between column and beam. Thus, in this paper, the behavior of the connections with eccentricity is examined by loading tests and finite element analysis, and further, the calculation method of elastic stiffness and yield strength of the connection is proposed.

 2. Calculation method of elastic stiffness and yield strength

 To develop the calculation method of the elastic stiffness and yield strength, only beam flange-to-column connection is focused on because stress transfer at the beam web connection is negligible. The beam flange-to-column connection is modeled dividing into diaphragm element and column element 1). Further, each element is separated as outside and inside parts to consider the effect of the eccentricity of beam. The elastic stiffness of the diaphragm element is obtained in consideration of flexural, shear and axial deformation 3). The elastic stiffness of the column element is determined using Rigid Body-Spring Model 7). The elastic stiffness of connection is obtained combining these elements. Meanwhile, the yield strength of the connection is determined by the lower strength of the two elements 3).

 3. Monotonic loading test and FEM analysis of beam flange-to-column connection

 To verify the fundamental behavior of the beam flange-to-column connections, experimental and numerical investigations were conducted. Test specimens consisted of a beam flange, a welded square hollow section column, and an exterior diaphragm. Both ends of the column were simply supported, and a tensile force was applied on the beam flange monotonically. On the other hand, the numerical model by finite element method has two beam flanges at both sides of the diaphragm, and anti-symmetric axial forces were applied on the beam flanges. Based on the numerical results, approximate equations to determine the unknown parameters included in the calculation method, proposed in Chapter 2, were made up. As a result, it is clarified that the calculation results with beam eccentricity can correspond with experimental and numerical results as well as those without beam eccentricity.

 4. Cycric loading test of the cruciform beam-to-column subassemblies

 To confirm elasto-plastic behavior of the whole beam-to-column connections, cyclic loading test and finite element analysis of cruciform subassemblies were conducted. Test parameters were eccentricity of beam and dimensions of the exterior diaphragm. Through the loading test, as the beam eccentricity becomes larger, it is confirmed that elastic stiffness and yield strength increase and the outside panel zone yields earlier because of the torsion about the longitudinal axis of the column.

 5. Conclusions

 In this paper, the calculation method of elastic stiffness and yield strength of beam-to-column connection using exterior diaphragms with small projection with beam eccentricity was proposed. Further, it is clarified that the calculation method has as well accuracy as that of connection without beam eccentricity, based on the loading tests and finite element analysis.