“No-shoes-floor” mean the floors that person takes off his shoes in the popular Japanese houses or elderly facilities, we define it in our research. Due to behavior (walking, sitting, lying, etc.) of person on no-shoes-floor, body regions that touch to surface of floors or motions at each body regions are more multifarious than western style living. The final goal of our research is the establishment of evaluation methods of floor deformation properties considering comfortableness human behavior. We focused on human walking and examined evaluation methods of comfortableness for walking as the first step of our research.
 The final goal of our research is the establishment of evaluation methods of floor deformation properties considering comfortableness human behavior. We focused on human walking and examined evaluation methods of comfortableness for walking as the first step of our research.
 At the First, we selected typical 26 floors that have variety of deformation properties that are defined by our research. Using these floors, sensory tests had been carried out about comfortableness during various motions on the non-footwear floors, and psychological scales had been constructed using the test results. Then, the following 2 results were found due to examination of mutual relationships of the scales.
 1) From a viewpoint of the evaluation about the comfortableness, the motions on no-shoes-floor are classified into 2 types. One is walking which only sole is in contact the floor surface during, and the other is various motions except walking.
 2) The feelings of comfortableness about the various motions except walking are alike. Then, if a performance value corresponding to the psychological scale about a motion is set, this can also evaluate the comfortableness during the others various motions except walking.
 At the Next, we execute evaluation method of floor hardness and dynamic discrimination method those typical evaluation method for human comfortableness. However these previous methods are complicated. So we tried to define the evaluation method which is more simple and appropriate than the previous method. In addition, the new method is able to measure the more detailed data. We call it as “evaluation method of floor deformation properties” (for walking).
 The evaluation method of floor deformation properties are constructed from the following 2 factors. These factors are measured by our developed machine that can reproduce loading condition of simulated walking action.
 T’ : performance value intends floor hardness
 D : performance value intends degree of decline of evaluation by titubation on floors
 And, it was found that Y₁～Y₃ as the performance values calculated from compounding these 2 values are able to correspond to the psychological scales about the comfortableness for walking on the floors. Then, we presented the evaluation method for walking; the method is measuring Y₁～Y₃ by using device which can reproduce loading condition of simulated walking action, and applying to relationship diagram between the performance values and psychological scales.
 The evaluation methods for the other actions will develop in next paper.

View full abstract

 In order to make a decision of necessity of evacuation and repair of the buildings after large earthquakes, information about damaged structural member existence and its location are extremely significant. Since damage to the structure is inherently local, when considering the damage detection by sensor information deployed in buildings, the local scale damage detection framework with high density and large number of sensors can be expected high sensitivity to damages, and also identify damage location. Recent technological advancement of small and low-cost MEMS type sensors and wireless data transmission bring such the local damage detection framework into reality. So, we have proposed substructure-based local damage detection method, which assumes a dense array of sensor instrumentation to a structure. The proposed method divides the whole structure into a set of small virtual substructures consisting of beams and columns. Then, detects local damage existence inside of the substructures using the differences between observation output and simulated output, which called output error, at internal node of the substructure.
 This paper presents the characteristics of the output errors which are induced by structural damages inside the substructure, which used as a damage index of the substructure. From the analytical studies to general types of substructure, the output error of internal node has been formulated as a function of the boundary nodes inputs and internal node outputs itself. This function is composed of only the transfer functions inside the substructure. This means that this function can be configured with only the dynamical characteristics and damage state of the substructure and is independent of the characteristics of the other part of the whole structure, so the output error can be used as a damage index of substructure. Based on these functions, numerical studies to a simple beam and column model have been conducted. The results show that an increase of the output error by structural damages occurs most stable in the rotational direction, especially in the low frequency range which is dominant in vibration of buildings. Therefore, it is desirable to use a rotational direction component of the sensor output to damage detection of substructure. It also shows that the output error will increase when the substructure has multi or heavy damages. Using the seismic response analysis results for the whole structure, the output error of substructure has computed to verify this study. The simulation results show that the output error of rotational component is bigger than the translational component, and as a result that the rotational component is most tolerant to the observation noise.

View full abstract

 The Tohoku earthquake of magnitude 9.0 occurred on March 11, 2011. As a result, the need for Structural Health Monitoring (SHM) was highlighted. To assess the health of buildings, modal parameters such as natural frequencies and damping ratios are the basic and important damage indicators. In addition, maximum inter-story drift angle is also recognized as the important indicator. If we would like to estimate these indicators precisely, ideally we need to install accelerometers on all floors. However, this situation is not realistic due to the cost. Thus, in realistic SHM systems, the number of sensors is limited. In such cases, the excitation (input) is assumed to be measured and available. However, in some cases, some sensors including the input sensor may not be available. Thus, in this paper, we propose a method for estimating inter-story drift angle using only one accelerometer without input.
 In our method, we set two assumptions. One is that the response is represented by the superposition of the response of only lower modes. The other is that mode vectors and participation factors are available from the design model or by other means. Then, first, we estimate modal frequencies and damping ratios using the subspace method from obtained acceleration data. Second, we decompose observed acceleration data to each mode by band pass filtering. Third, we calculate mode response by solving on the vibration equation of each mode. The vibration equation allows us to obtain mode response using the observed output response as input to this equation. It was verified that this method could successfully estimate the modal response as well as the inter-story drift angles. The overall accuracy of the method was much better than the existing method.

View full abstract

 The near-fault ground motions have special characters and the effects of such near-fault ground motions on structural response have been investigated extensively so far. The fling-step (fault-parallel) and forward-directivity (fault normal) inputs have been characterized simply by two or three wavelets. For this class of ground motions, many sophisticated analyses have been conducted from various viewpoints. However, as far as a forced input is employed, both a free-vibration term and a forced-vibration term appear and the closed-form expression of the elastic-plastic response may be difficult. In order to overcome this difficulty, the double impulse input is used as a good substitute of the near-fault ground motion and the closed-form expression is derived of the bilinear elastic-plastic critical response of a structure under this double input.
 For guaranteeing the equivalence of the double impulse and the corresponding one-cycle sinusoidal input as a substitute of a near-fault ground motion, the equivalence of the maximum Fourier amplitudes of both inputs is introduced together with the equivalence of input frequency. Because only the free-vibration appears after each impulse of such double impulse input, the energy balance formulation using the kinetic energy and the strain and dissipated energy leads to the derivation of the closed-form expression of a complicated bilinear elastic-plastic response. It is shown that the maximum inelastic deformation of a structure with a positive post-yield stiffness can occur after the second impulse different from the case for elastic-perfectly plastic models. Furthermore it is made clear that, as the positive post-yield stiffness becomes larger, the correspondence between the double impulse and the corresponding sinusoidal input becomes better.
 The closed-form expression is also derived for the non-critical case. The introduction of the concept of critical excitation enabled this derivation of the closed-form expression of the maximum elastic-plastic response. The validity and accuracy of the proposed theory are investigated through the comparison with the response analysis to the corresponding one-cycle sinusoidal input and recorded ground motions (Imperial Valley 1979, Northridge 1994, Hyogoken-Nanbu 1995). It is demonstrated that the proposed method using the double impulse is applicable to actual recorded pulse-type ground motions within a reasonable accuracy.

View full abstract

 The oil damper is a typical example of a high-performance damping device for structural control. It is well known that the mechanical characteristics of an oil damper installed in a building with a bracing frame is expressed by a Maxwell model, and the energy absorption capacity or the control effect is limited by the stiffness of the model's spring. We have proposed a highly efficient oil damper that can break through this limitation by introducing an energy recovery system. The proposed device is equipped with an auxiliary oil tank outside the main cylinder, and the oil flow between the cylinder and the tank can be controlled by control valves. Conventional oil dampers, including variable types, always change vibration energy to heat. However, the proposed device recovers the vibration energy as oil strain energy in the auxiliary tank, and reuses it at an optimum timing to enlarge the damper stroke or to improve the control efficiency.
 This paper presents a newly developed control system based on a semi-active oil damper with an energy recovery system. First, we explain the basic configuration of the system, and show how the energy recovery system works according to the algorithm. The mechanical model of this device is expressed as a four-element model that consists of a Maxwell model and a Voight model in series. The spring element of the additional Voight model represents the equivalent stiffness of the oil in the tank. If we set the tank stiffness to be almost equivalent to the stiffness of the spring element of the Maxwell model, the energy dissipation capacity can reach 4 times that of a conventional oil damper.
 Second, we describe the design specification of the full scale control system, which consists of hydraulic device and a controller. The device houses an auxiliary oil tank, solenoid valves and sensors. The fundamental component parts, such as a cylinder, piston and ball-and-socket joint, are the same as those that have been applied to actual high rise buildings for 15 years. The controller controls the valves according to the signal processing flow based on the information from the pressure sensors and the stroke sensor housed in the hydraulic body. The controller has multiple failsafe functions in case of not only sensor problems but also CPU trouble. If trouble is detected, the current supply to the valves is cut off by the relay and the hydraulic circuit automatically switches to that of a passively controlled variable oil damper.
 Third, we show the results of dynamic loading tests conducted on a full-scale specimen. It is demonstrated that the proposed algorithm and the energy recovery system work stably, and the energy dissipation capacity is improved as expected. The results of simulation analyses based on the four-element mathematical model agree well with the test results including the change of the tank's inner pressure.
 Finally, we present the results of a seismic response analyses on a high-rise building model, and discuss the control effect of the proposed oil damper system compared with those of various other existing damping devices.

View full abstract

 1. Introduction
 Pile foundation response during earthquakes is strongly affected by nonlinear soil-pile foundation interaction. The damages to pile foundations during the 1995 Hyogo-ken Nanbu Earthquake and the 2011 Tohoku-Chiho Taiheiyo-Oki Earthquake were obviously attributed to nonlinear interaction of soil-pile foundation-superstructure. The dynamic nonlinear behavior of soil around each pile affects lateral load distribution and displacement of each pile. Therefore, shaking table tests were conducted to clarify the dynamic nonlinear behavior of soil around piles, further earthquake response analyses were conducted to simulate the effect of the nonlinear soil-pile interaction system on performance of the superstructure supported by pile group.
 2. Outline of shaking table tests
 25-pile group foundation model was set up in Toyoura sand deposits within a laminar container. The sand deposits were prepared by air pluviation method in the laminar container. The model piles were acryl cylinder with 12mm in diameter and 400mm long. 25 piles were arranged in squares and pile spacing was 2.0 times pile diameter. The superstructure was modeled as a rigid body. The input waves were seismic waves in notification with random phase of Japan and the Hyogo-ken Nambu earthquake. 3 input acceleration levels were used to investigate the influence of input motion level. Tests with and without the mass of superstructure were conducted to investigate lateral load distribution of each pile.
 3. Outline of shaking table tests simulation
 Two analysis methods were conducted for simulation of tests, one is nonlinear 3D-FEM, and the other is lumped mass model. Two case of Lumped mass model are conducted, one is with soil springs around pile group which were different according to pile arrangements and direction of loading, and is with soil springs without pile group effect. Soil springs were calculated by using nonlinear 3-D FEM, which were modeled appropriately according to the location of pile.
 4. Conclusions
 The concluding remarks of these shaking tests and analytical studies are as follows.
 (1) In shaking tests, lateral subgrade reaction around piles depends on the location of each pile in pile group and the direction of loading. That of the corner pile is remarkably larger than that of the middle pile.
 (2) In shaking tests, hysteresis curve of soil spring at the corner pile is found to be asymmetric loop, while that at the middle pile be symmetric loop.
 (3) In shaking tests, bending moment at each pile head depends on the location of each pile in pile group. That of corner pile head is remarkably larger than that of middle pile head. As the input level of acceleration increase, bending moment at each pile head is equalized.
 (4) The proposed lumped mass model of the soil-pile foundation system, which consists of beam elements and nonlinear interaction springs considering the location of pile, is verified validation. It is confirmed that this proposed model well represents response of the superstructure supported by pile group.

View full abstract

 This paper presents seismic retrofitting technique of existing mid-rise SRC buildings with newly added stories utilizing mass damper effect. This proposed retrofit method is summarized as follows. A number of the stories are reduced from the top of the building. A lightweight super structure of steel moment resisting frames is newly built on the rooftop through a mid-isolation system with lead rubber bearings (LRB). Total number of the stories is assumed to increase than ever as far as the total weight of the retrofitted building is below the former one. The downsizing of the building has effect on mitigation of inertia force induced by ground motion. On the other hand, the extension part is expected to work as a tuned mass damper having large mass. The objective of this study is to find optimum design parameters, which consist of required number of reduction stories, strength and stiffness of the mid-isolation.
 Firstly, an equivalent two degrees of freedom (2DOF) system consisting of the existing part and the extension part is employed for computational efficiency of nonlinear time history analysis. As a result, design formulae of the optimum strength and stiffness are presented through a response surface method in terms of the maximum drift angle. By using the proposed optimal strength of the isolation, relationship between inter story drift of the isolation and the acceleration of the super structure are clarified in terms of the mass ratio and yield strength of the substructure. Thus, these numerical results are accurately estimated by introducing the deformation ratio of the substructure to the super structure without time history analysis, which ratios are derived based on the stochastic vibration theory.
 Secondary, a performance curve is formulated based on the previous results. This diagram approach enables structural engineers to decide required number of reduction stories, stiffness and strength of the isolation in a similar manner to that of the other popular passive control systems. A retrofit procedure is presented with the formulae of the design parameters and the performance curve accordingly.
 Finally, examples of the retrofit are shown for multistory buildings having seven or ten stories. The buildings are retrofitted in accordance with the proposed procedure. Three historical ground motions are used to verify improvement of seismic performance through time history analysis of stick models having all degree of freedom of the stories. Maximum inter story drift angles of the existing part are confirmed to be less than 1/200 radian, which requires no additional reinforcement for the column members. Both the maximum deformation of the isolation and the maximum acceleration of the super structure meet a specified seismic criteria and show good agreement with estimated values based on the performance curve.

View full abstract

 The author has proposed the following calculation formulas on the rectangular flat slabs of the arbitrary width. Those are the primary natural frequency on the rectangular flat slabs, the new restraints along edges, the effective mass coefficient and the deflections of the live load.
 Formerly, the calculation formulas of the primary natural frequency of the both all edges simply supported and all edges built-in on the rectangular slabs have only been proposed. So Kojima(2012) had derived the calculation formulas of the primary natural frequency on the generalized support conditions of the rectangular flat slabs. The subjects of studies for support conditions on the rectangular flat slabs are all edges simply supported, all edges built-in, two opposite edges simply supported and the other two edges clamped, one edges simply supported and three edges built-in, three edges simply supported and one edge built-in, two adjacent edges simply supported and the other edges clamped, all edges semi-clamped. Since the primary natural frequency of the ordinary flat slabs is intermediate between the simple support and the fix along edges, Kojima(2012) had defined restraints along edges of the flat slabs which are set to 1 by simple support along edges of the flat slabs and set to about 2 by fix along edges. Through the definition of the new restraints along edges, Kojima(2012) had estimated the primary natural frequency of the ordinary flat slabs is intermediate between all edges simply supported and built-in on the rectangular flat slabs, that is the numerical value is 1.5 times of all edges simply supported slabs. Kojima(2013a) had proposed the calculation formulas of the effective mass coefficient that the concentrated load affected the rectangular flat slabs. As the support conditions, Kojima(2013a) had calculated the simply support along edges, the fixed support along edges, two adjacent edges simply supported and the other edges clamped, or all edges semi-clamped. Through a series of these processes, Kojima(2013b) had derived the calculation formulas of the maximum deflections from the concentrated load on the rectangular flat slabs.
 Timoshenko(1959) proposed the accurate solutions of the deflections on the rectangular flat slabs through both the all edges simply supported and built-in with the both the concentrated load and the uniformly distributed load. In this paper, the author has proposed the accurate solutions, the polynomial solutions and the solutions with arbitrary size incorporated effective load by Kojima on the deflections based by Timoshenko, s solutions on the rectangular flat slabs. Moreover, these solutions with arbitrary size incorporated effective load by Kojima(2013b) have been expanded not only the central concentrated load, but also the uniformly distributed load. The author compares and verifies among the accurate solutions, the polynomial solutions and the solutions with arbitrary size incorporated effective load by Kojima(2013b).

View full abstract

 An accurate plastic hinge type beam elememt has been developed by the authors for three-dimensional (3-D) elastoplastic large deformation analysis of frames which contain all kinds of members: i.e. steel members, RC members, SRC members, CFT members, PC members, steel damper braces and tension braces. The element was originally proposed for pure steel frames by the first author (Shugyo 2003) and named Fibered Plastic Hinge Model (FPHM). The formulation procedure is a combination of the modified incremental stiffness method, the updated Lagrangian formulation, and numerical integration of fiber stiffnesses about the sections at the plastic hinges. The following assumptions are made to form the elastoplastic tangent stiffness matrix of the pure steel element: (1)members have thin walled closed or open sections, and cross sections remain plane and do not distort in the absence of cross-sectional warping, (2)deflection is large but elastic strain is small, (3)axial stress and the shear stress due to St. Venant torsion participate in yielding of fibers of members with closed sections, while only axial stress participates for members with open sections, (4)plastic deformation consists of only four components that correspond to axial force, biaxial bending moments, and torsional moment or bimoment, (5)there is no local buckling, (6)although an actual generalized plastic strain increments in a short element generally distribute nonlinearly, it is idealized as generalized plastic strain increments distribute linearly with the values at element nodes i and j, (7)incremental plastic deformations in the two half portions occur concentrically in the plastic hinge of zero length at element nodes i and j respectively. Because of the above mentioned assumption (6), the element requires at least four-element approximation for a frame member. This causes considerable increase of total degrees of freedom in the analysis of multi-bay multi-story frames.
 In this paper a method to reduce the total degrees of freedom in frame analysis by the FPHM is presented. Introducing the plastic deformation reduction coefficient, the assumption (6) is modified as one-element approximation for a frame member has sufficient accuracy for practical use. The optimum value of the plastic deformation reduction coefficient is examined by using quasi-static analysis of four kinds of one-bay one-story plane portal frames, 3-D quasi-static analysis of two-bay four-story steel frame which contains composite beams and semirigid column bases, and 3-D quasi-static analysis of twenty-story eccentric steel frame with H-shaped steel columns.

View full abstract

 This paper proposes an estimation method of the impact force acting on ceiling safety nets in consideration of the distance between ceiling and safety nets. The proposed method utilizes the balance of the potential energy of the falling objects including ceiling boards and the strain energy stored in safety nets or wire ropes after their deformation, in order to obtain the maximum displacement of the safety nets or wire ropes. The structures targeted by the proposed method can be categorized into two types: 1) safety nets or wire ropes solely lying in one way, and 2) safety nets and wire ropes which interact with each other in two dimensions. The collapse patterns of ceiling considered in this paper are the one that can be treated as concentrated load and the one that can be treated as uniformly distributed load.
 The results obtained in this study are as follows:

 (1) It has been confirmed that the finite element analyses can reproduce the result of the static loading experiment in sufficient accuracy, in which the load-displacement relation of the ceiling safety nets has nonlinear characteristic.

 (2) The impact force acting on safety nets or wire ropes solely lying in one way can be estimated as the vertical reaction force by adapting the maximum displacement on the relation of the vertical force and the displacement. Due to the difficulty originating from the geometrical nonlinearity, the vertical force-displacement relation and the horizontal force of the distributed load are obtained by finite element analyses, whereas those of the concentrated load are obtained by theory.

 (3) The impact force acting on safety nets and wire ropes which interact with each other in two dimensions can be estimated as done in the above case. The maximum displacement of the structure is calculated by adding the maximum displacement of each structural element.

 The proposed method has been applied to some examples to examine the influence of the distance between ceiling and safety nets on the impact force. It has been concluded by the results obtained through these examples that the distance between ceiling and safety nets seriously affects the impact force and should be considered in estimation.

View full abstract

 1. introduction
 The main object of this study is the scale effect with the destruction of the wood. In the steel, J_i value is a material characteristic value so it does not depend on size. The more the size of specimen is large, the more the stress at J_i is small. So it is easy to make the dangerous judgment of destruction. This property can be found in the wood, we make the JIC test with Japanese larch.

 2. Experimental Method
 We made experiments is referred to in The Japan Society of Mechanical Engineers. The ratio of specimen's width, height and span is a one-two- eight ratio(Fig. 1). Their notches are height of half and added 1mm wide notch at first and infinitesimal wide notch at last 1mm. (Fig. 2) If the crack initiation is observed frequently three times within each 2seconds intervals, we defined the third sound as initial point of frequent sound. (Fig. 3) We stopped loading and colored the crack of specimens when we heard the sound of crack initiation. Also we stopped loading and colored it before or after the sound on each specimens. (Fig. 4) We aimed at the load to be static and monotonic loading at a speed of 0.5mm/min. We measured load, COD (Crack Opening Displacement), displacement of force and length of crack colored. And also we observed the sound of crack initiation.

 3. Experimental Results
 We measured the growth of internal crack by coloring. (Photo. 3) There are two types of the crack growth as wavelike or parabolic in any dimension. We show the results in Table. 2. We calculated the J integral value from the relation of load and displacement by the simple equation⁴⁾.

 4. J_i value worked out by coloring
 We plotted the relationship between the J integral value and the average of the internal crack length (Δa_ave) on each series. (Fig. 8) We defined the J integral value at the y-intercept as Ji. We compared with Ji to the J integral value at the first sound, and we considered that it is useful to determine the fracture toughness value simply. (Fig. 12)

 5. Scale effect in the fracture mechanics
 It was observed generally good response that comparing the experimental value and the theoretical formula in applied the fracture mechanics. The scale effect was confirmed not just with the steel but also with the wood. So it can be said that the application of fracture mechanics to the destruction of wood is useful for crack propagation status confirmation.

 6. Conclusion
 1) It was observed generally good response that comparing the experimental value and the theoretical formula in applied the fracture mechanics. The scale effect was visible not just with the steel but also with the wood. So it can be said that the application of fracture mechanics to the destruction of wood is useful for crack propagation status confirmation.
 2) We worked out J_i value with the average value of crack growth measured by coloring. We draw an approximate straight line by the least squares method with the plot of J-Δa_ave relationships, We established the J integral value of the y-intercept as J_i value.
 3) We paid attention the J integral value at the first sound with the sound of the crack initiation. The average value of them was 69.0N/m. And the difference between the average value of each size and the average value of all was within 1%. It is useful for estimating J_i from this value.

View full abstract

 Cross Laminated Timber (CLT) panel is consisted of several layers of lumber stacked crosswise and glued together on their surfaces. In Europe, mid-to-high-rise building using CLT has been already built. Recently, CLT structure attracts attention in Japan as one of the means to promote Wooden Public Buildings. Since the lamina can utilize small or middle diameter wood as the raw materials, expansion of CLT demand would promote local forestry and wood industry. There are various joint methods of CLT such as LSB joints, screw joints with steel side-plates, and so on. In this study, we examined tensile bolt joint which is one of the effective joint methods of the end of panel when the wall panel showed rocking behavior, assuming architectural structure using the panels of around 1-2m in width. Research process of this study is as follows:
 1) Comparison for calculated results and test results on partial compression of CLT materials.
 2) Test results of tensile bolt joint with CLT.
 3) Proposal of design method for estimating strength properties of tensile bolt joint with CLT and the verification of its suitability.
 In study 1), partial compressive strength of CLT panel was calculated by combining estimated results of the longitudinal and lateral compression stiffness of lumber (Eq. 1). These estimated results were compared with experimental values of the paper (ref. 12), and estimated results were calculated slightly lower.
 In study 2), cyclic loading tests were conducted tensile test on tensile bolt joint with CLT. The parameters of the experiment are as follows:
 1) Vertical end distance between lower end of the CLT and rectangular through-hole (l_s), horizontal edge distance between side edge of the CLT and rectangular through-hole (l_n), strong-axis direction or weak-axis direction and positioning of edge-glued lamina (Table 2, Fig. 7)
 2) Thickness of the CLT and number of the bolt (Table 3, Fig. 23)
 In consequence, it was found that the above all parameters of experiment had an influence on the strength performance, particularly, the end distance (l_s) and the edge distance (l_n) were important elements for affecting the change of the fracture mode.
 In study 3), the estimating results calculated with proposed design method were compared with the experimental results. It was quoted that the stiffness of joint is obtained by sum of the compression of CLT, tension of CLT and tension of bolt, those stiffness in series (Fig. 30, Eq. 1～Eq. 4). The yield strength of joint assumed it compressive strength of parallel layers (Fig. 31, Eq. 5). Fig. 32 shows the three kinds of fracture mode, and the maximum strength of joint was assumed to the minimum of those maximum strength (Fig. 32, Eq. 6～10). As a result, maximum tensile strength, fracture mode, yield strength and stiffness were close to these estimations. However, estimated stiffness of test specimen ”P77” was higher than experimental result. It was estimated that the reduction rate of the stiffness per unit area accompanying increase in the compressive area of the parallel layers was higher than expected.

View full abstract

 Many RC buildings built before 1981 need seismic retrofit at present in Japan. Installing new RC shear walls or steel braces in RC frame is one of the seismic retrofit techniques. In this paper, the new type of seismic retrofit method using CLT panels as shear walls is proposed. In this method, setting small CLT panels in RC frame and bonding each panel and panel to RC frame with epoxy resin, panels compose shear walls. The advantages of this technique are: There are less dust, noise, and vibration during construction; Light weight panels enable easy construction and short construction period; Light weight panels also cause small seismic force.
 In this research, cyclic loading tests for 5 types of reinforced specimens and 2 types of plain RC frames as control were conducted. 2 types of RC frame were made in 1/3 scale and their spans are 1540mm and 2940mm . 3 layer/3-ply Japanese cedar CLT panels were used. 4, 7, or 10 pairs of panels are piled and set in a line within the RC frame. And these are bonded to each and to the RC frame with epoxy resin on-site. Every adhesive thickness is 5mm thick.
 Every reinforced specimen is stiffer, stronger, and more ductile than the plain RC frame. And test results indicate that the bond strength in each specimen is around 3 N/mm², the maximum strength of panels in each specimen is under the shear strength, and the stiffness depends on the fiber orientation of CLT to columns. So the maximum bond load can be bigger if the adhesive area becomes larger. And this seismic retrofitting can become stiffer if the inclined fiber orientation of CLT is set in the RC frame. The failure of the bonding connection between concrete and panels caused a small drop of strength at the end of elastic region in every specimen. The failure of bonding connection occurred inside of concrete. Then curve of the specimens immediately followed by secondary linear part, which was caused by the compression of the part of panels, the friction and shear key resistance between concrete and panels.
 The stress analysis showed that the bond strength between CLT and RC and shear modulus of CLT in these specimens match the result of element tests. So the specimen strength could be divided into the RC frame strength and the CLT strength until the initial deformation. As the bond strength between CLT and RC was smaller than the shear strength of CLT, the specimens can be stronger by increasing the adhesive area.

View full abstract

 Lateral braces are usually connected to upper flanges of H-shaped beams in the moment resisting frames. Continuous braces such as folded-roof plates may restrain the lateral deformation of H-shaped beams, when they are jointed to the upper flanges of H-shaped beams. If the stress of the flange with continuous braces becomes a tensile, the rotational rigidities of the lateral braces must be increased to obtain sufficient amount of the lateral buckling strength of the beams.
 Our previous research (Kimura, Yoshino and Ogawa 2012) clarified the relation between the lateral buckling strength of H-shaped beams and the demands of the lateral and rotational rigidities for continuous braces when a beam is subjected to uniform moment distribution. The elastic buckling strength of the H-shaped beams by formulating energy conservation equations was estimated, considering the lateral and torsional deformation of bracing points of beams.
 Then our previous research revealed that the lateral buckling strength of H-shaped beams is effectively increased when lateral braces are attached to compressive flanges, while it is not increased when lateral braces is attached to the tensile flanges.
 To consider more realistic conditions, this paper takes the effect of moment gradient along beams into account on the elastic lateral buckling strength of H-shaped beams with continuous braces.
 In this paper, three groups of H-shaped beams are adopted as simulation models with a ratio of a flange width to a web depth “b/h” of 0.33, 0.5, 0.78. Two types of loading conditions are considered: one is that the stress of upper flange become tensile (designated as Type A), and the other is that the stress of upper flange become compressive (designated as Type B), as shown fig, 3, when H-shaped beams is subjected to uniform moment.
 The following procedure is adopted: First, the elastic buckling strength of the beams under moment gradient is estimated by formulating energy conservation equations, which are verified by eigenvalue analysis. Then, the effect of lateral and rotational restraint of continuous braces on the lateral buckling strength for the H-shaped beams is examined. In addition, the elasto-plastic buckling behavior of the beams is further simulated by elasto-plastic large deformation analyses.
 Finally, the elasto-plastic buckling moment of the beams are evaluated according to the buckling curve provided by Recommendation for Limit State Design of Steel Structure (Architectural Institute of Japan (AIJ)) using the proposing modified equivalent slenderness ratio.
 Major findings of this study can be summarized as follows:
 1. The elastic lateral buckling strength for H-shaped beams restrained by continuous braces under moment gradient is obtained from Eqs. (12), (15), (19) and (22).
 2. For Type A as shown Fig. 3, Eqs. (12), (15) and (19) is applicable as the elastic buckling strength for these beams. Herein, the ratio of bending moment at the ends M₂/M₁ is expressed as the symbol of m as Fig. 3.
 3. For -1.0≦m<0.4 of Type B as shown Fig. 3, Eqs. (12) and (22) is applicable as the elastic buckling strength for these beams. For m>0.4, the lower flange with no lateral braces are attached are buckled laterally, and this buckling mode for Type B is same as that for Type A. Therefore, the lateral buckling strength of Type B can be approximately evaluated by that of Type A.
 4. The elasto-plastic buckling moment of beams under moment gradient can be sufficiently estimated by the buckling curves provided by Recommendation for Limit State Design of Steel Structure (AIJ) when the proposing modified equivalent slenderness ratio from Eq. (24) is applied.

View full abstract

 H-shaped steel beams connected to steel dampers are subjected to compressive axial forces in addition to flexural moment, so that its buckling behavior is more unstable than that to only flexural moment. On the other hand, it is effective for continuous braces to prevent the lateral bucking of H-shaped beams. In this paper, the lateral buckling behavior for H-shaped beams with continuous braces under flexural moment and compressive axial force is clarified, and elasto-plastic lateral buckling stresses of the beams or lateral stiffing force and rotational stiffing moment of continuous braces are evaluated by the energy method and numerical analyses.

View full abstract

 1. Introduction
 The objective of this study is to examine the relationship between battened built-up columns and frames about buckling. Especially, the characteristics of the Bleich's formula are investigated with respect to the upper bound theory and moment distribution. The buckling strengths, buckling modes and bending moment of one bay －n stories frame with rigid beams are calculated by the energy method. The results are compared with those of obtained by the Bleich's formula and correct solutions.

 2. Analytical work
 The analytical model is shown in figure 3. Buckling strengths are calculated by the three methods. Firstly, the buckling formula presented by Bleich is shown, and analytical method using the energy method is described, where the buckling modes are assumed by the equations (12). The buckling equation is presented as equations (22). Moreover it is shown that the correct buckling strengths can be calculated by the equations (23) and (24).

 3. Results and Discussion
 As the analytical parameters, the number of stories n and the value of 4Ic/A_cb² are selected. The values of n are 1, 2, 3, 5, 10 and the value of 4Ic/A_cb² ranges from zero to one. Figure 5 and 6 show the correct dimensionless buckling strength p^*-4Ic/A_cb² relations and p_g^*-4Ic/A_cb² relations. Figure 8 shows the ratio of the Bleich's strength to the correct value and the strength obtained by the energy method to the correct value.
 Although the buckling strengths by Bleich agree fairly well with correct values, it is shown that the strengths obtained by Bleich equation are neither the upper bound nor the lower bound. Moreover, Figure 12 shows that Bleich's moment distributions correspond to those of obtained by the energy method.

 4. Conclusions
 The conclusions derived from this study are as follows:
 1) Governing parameters of the problem are the number of story n and the value of 4Ic/A_cb².
 2) Although the buckling strengths by Bleich agree fairly well with correct values, the strengths by Bleich are neither the upper bound nor the lower bound.
 3) Buckling strengths calculated by the energy method estimate the correct strength within 5% error in case of n>2. As is well known, the buckling strength by the energy method gives the upper bound.
 4) Bleich's moment distributions correspond to those of obtained by the energy method.

View full abstract

 Transmission steel towers and radio communication steel towers, where the size is less than middle, are usually constructed by angle members. Angles in the towers are commonly connected by bolts and lap joint is used. Lap joint will generate an eccentricity; secondary bending moment due to this eccentricity will be subjected to the member and affect to the flexural buckling strength. Currently, structural design of the steel towers follows two Standards. Transmission steel towers follows Design Standard on Structures for Transmissions (JEC-127); radio communication steel tower follows Design Standard on Structures for Radio Communication in Power System (JEC-144). In these Standards, effect of eccentricity is categorized in three classes; flexural buckling strength of the members are evaluated. However, the categories shown in the Standards are qualitative and not evaluating the eccentricity quantitatively.
 To clarity the effect of eccentricity due to lap joint, firstly, full scale testing were conducted in this study. Moreover, difference of bending stiffness of connected angles will affect to the strength; therefore, this parameter is also included in the testing. Test parameters are location of the lap joint, difference of the bending stiffness of connected angles, and member length (i.e., slenderness ratio). From the testing, following results were observed.
 (1) Location of the lap joint affected the flexural buckling strength even if the connected angles were same sizes. The specimen where the joint was located L/4 in the member showed smaller strength than the specimen where the joint was located at the middle (L/2) of the member (L is member length).
 (2) Difference of bending stiffness of connected angles affected the flexural buckling strength. The strength of the member which had different size of angles decreased compared to the strength where the member was composed with same size. Difference of angle sizes will change the amount of eccentricity at the joint and also changes the equivalent bending stiffness of the member.
 (3) Flexural buckling deformation in elastic range showed two patterns. Specimens where the joint was in the middle of the member showed maximum deformation at the end of the lap joint. On the other hand, specimens where the joint was in L/4 of the member showed maximum deformation at the middle of the member.
 Observation from the testing showed that difference of bending stiffness of connected angles, location of lap joint, and eccentricity according to the lap joint shall be evaluated to estimate the flexural buckling strength. To evaluate these parameters, Rayleigh-Ritz method and energy equilibrium theory were applied. Based on Rayleigh-Ritz method, firstly, difference of bending stiffness of connected angles without eccentricity was evaluated; coefficient α that can be used to evaluate this effect was derived. Secondary, based on energy equilibrium theory, effect of eccentricity at the lap joint and location of the joint was evaluated; coefficient β that can be used to evaluate these effects was derived. Finally, coefficients α and β are incorporated to evaluate the flexural buckling strength of angle member with lap joint.
 Utilizing the derived flexural buckling strength, effective slenderness ratio which is useful to evaluate the flexural buckling strength is proposed. Proposed formula was validated with the test results, and reasonable results were confirmed.

View full abstract

 Timber elements, which are different from other structural elements, have a characteristic problem in that the load bearing capacity decreases due to self-burning in the case of a fire, and this self-burning may continue after other fuel in the room has been exhausted. Therefore, the structural fire performance of timber elements should be clarified during not only the heating phase, but also the cooling phase. In the present paper, deflection behaviour and failure mode of larch glued laminated timber beams exposed to fire heating and natural cooling is discussed based on load-bearing fire tests in which load level is the test parameter.

 The present paper described the fire performance, including the cooling phase, of structural glued laminated timber beams whose section was 210mm (width) x 420mm (height). The main conclusions were:

 (1) Failure time was 79 minutes, for the long-term allowable load was loading. It was satisfied the Quasi-fire resistance of 1 hour.

 (2) In the cooling phase, it may continue to support the load, when loaded below 0.4 times of long-term allowable load.

 (3) On the other hand, in the cooling phase after 1 hour heating, it was confirmed that the deflection is increased. Compared with the deflection just after heating for 1 hour. LF-0.4 (1) in the cooling after 7 hours, deflection than during the just after 1 hour heating was about 6.7 times. LF-0.2 In the cooling After 24 hours, deflection than during the just after 1 hour heating was about 7.2 times.

 (4) In the case of bending failure, strength of cooling 7 hours and 24 hours was about the same as the strength of the cooling 3 hours of previous report.

 (5) The failure mode, bending failure and due to the laminar layer tensile breaking of the beam bottom, the shear failure due to the slip in the growth ring centroid axis near of the beam cross section was observed. Shear failure was seen in up to 1 to 3 hours from after heating.

 (6) Bending strength in the cooling phase of the residual cross-sectional area, calculated value that gave the strength reduction rate due to Eurocode 5 is roughly agreement with the experimental value. On the other hand, the calculated value for the shear strength was significantly higher than the experimental values.

 In the future, investigate the relationship between the strength of evaporation and the timber of water from the high-temperature material testing, it is expected to further consider the strength reduction in the cooling phase.

View full abstract