The climatic change such as the global warming is one of the important international problem. On the field of construction materials, environmental-conscious concrete which are CO₂ reduction from the concrete materials advance development and practical application. Mostly of these techniques use equivalent for the type C blended cement that are used a large amount of mineral admixtures as the substitution of cement. However, these concrete are used only underground structures of buildings due to properties of strength development, durability and construction performance. Moreover, the load for ready-mixed concrete plant is large on the manufacturing.

This research studied on the environmental-conscious concrete to use general purpose including above-ground structures. Type A portland blast-furnace slag cement is the blended cement which has the equality properties with ordinary portland cement. Nevertheless, Type A portland blast-furnace slag cement does not have production and distribution. Therefore, equivalent for the type A portland blast-furnace slag cement is made to blend the ordinary portland cement and the type B portland blast-furnace slag cement, and experimented on the indoor and the actual site.

On the indoor experiment, fresh concrete properties, strength development and durability were tested on 47 types concrete which were due to influenced mixing rate of blast-furnace slag, water-cement ratio and variability in qualities of the type B portland blast-furnace slag cement. As the result, concrete properties of blended the ordinary portland cement and the type B portland blast-furnace slag cement were similar to concrete adding granulated blast-furnace slag into ordinary portland cement. On 20% or less rate of blast-furnace slag, furthermore, compressive strength of since 28 days, setting times and carbonation resistance were equality with ordinary portland cement concrete.

On the actual experiment, time dependent change of fresh concrete, strength development of various curing conditions and model members, and construction performances were tested concrete which was made in ready-mixed concrete plant. As the results, method of mixture proportion design, quality control and curing showed can treat same with ordinary portland cement concrete.

In this matter, the type A portland blast-furnace slag cement equivalent which blend the ordinary portland cement and the type B portland blast-furnace slag cement has similar concrete characters with ordinary portland cement. Accordingly, this concrete is considered can use above-ground structures as the environmental-conscious concrete, and CO₂ is reduced approximately 18.9% from the trial calculate.

The indoor and outdoor surfaces of the peripheral walls of buildings are constantly exposed to different temperature and humidity conditions. Because various finishing materials are often applied to external reinforced concrete peripheral walls and rooftop slabs, carbonation-induced deterioration progresses at different rates on the indoor and outdoor sides of such structures. Therefore, an apparatus was developed to simulate environmental exposure by simultaneously applying different temperature and humidity conditions to the indoor and outdoor surfaces of mortar specimens and the effects of mortar temperature and moisture content on the rates of carbonation and rebar corrosion were investigated. In addition, two types of insulations were applied to external mortar surfaces and the effects of such materials on the rates of carbonation and rebar corrosion were examined.

The major findings of the investigations are summarized as follows:

(1) Carbonation can be controlled by applying an insulation layer and high-gas-permeability polystyrene insulation to external mortar surfaces. In addition, in the externally insulated test specimens, the summertime increase in concrete temperature was suppressed, and the moisture content was higher than that of the pristine mortar test specimen; therefore, carbonation was also controlled on the indoor side. However, in winter, all the specimens were maintained at a high temperature, and the resulting moisture content of both the indoor and outdoor sides were low owing to accelerated carbonation.

(2) In both summer and winter, rebar corrosion increased with both increasing mortar temperature and moisture content in the vicinity of the rebar, regardless of whether insulation materials had been applied to the external mortar surface.

(3) Even when the external mortar surface did not directly come into contact with rainwater, the wintertime mortar moisture content increased when the indoor surface was cold owing to condensation and water penetration from the outdoor surface. As a result, the half-cell potential of the rebar in the carbonated mortar decreased and the corrosion- current density increased.

(4) In summer, the insulation-induced suppression of the increasing mortar temperature effectively controls both outdoor and indoor concrete carbonation. In winter, on the other hand, the heat insulation effect may accelerate carbonation. However, preventing moisture condensation effectively controls rebar corrosion. Therefore, applying insulation materials to external surfaces can effectively extend the life of reinforced concrete structures.

When we create input ground motions for building design, it is necessary to consider multiple cases for source parameters such as strong motion generation area (SMGA) arrangement and its area size, and multiple faults. Presently, it is difficult to cover all the possible cases of faults and source parameters Therefore, we often considered the safety side ground motions with the faults and SMGA placement close to the planned site. However, it has been reported that long period ground motions do not necessarily directly relate to the distance between the epicenter and the site, and the ground motions vary with the direction of the source and the depth of the hypocenter.

The purpose of this study is to examine the arrange ment of the SMGAs and to study the effect of the arrangement by visualizing the Green’s function between the selected site and the source (hypocenter). This study is conducted by considering the building response to the seismic event. We conducted numerical simulations of the inter-plate earthquake along the Sagami Trough that is anticipated to occur in the Tokyo. In this study, we compared the results of the simulations with various SMGA arrangements. We also examined whether the spatial distribution can be interpreted based on the propagation path of the seismic waves generated from the source area where the Green’s functions were large.

From visualizing the Green’s function on the source area, the source area with large amplitude and long duration is located along the plain edge from the south to the west from Tokyo. Additionally, it was found that the results of waveform synthesis in case SMGA is arranged at this location, its amplitude becomes larger and duration becomes longer than the conventional calculation results. From the above, by visualizing the spatial distribution of the Green’s function and arranging the SMGA to be in the region where these amplitudes are large, it is possible to predict the ground motions considering the variation in the long period ground motions with source location. However, according to the calculated spatial distribution of the Green’s function when the building period, damping constant, and site was changed, the source area that showed a higher Green's function varied. Therefore, it is necessary to visualize the spatial distribution of the Green’s function for the buildings with various vibration characteristics in various city.

The limitation of this study is the results depend on the accuracy of the source model and the underground structure model and the results do not completely match the actual ground motions. In addition, there is room to study other source parameters such as the hypocenter (focus), rise time, and seismic moment. However, the present results of this paper can be used to consider the safety side in the structural design of the buildings.

Until now, we have set SMGAs based on the asperity location analyzed from the seismic intensity distribution of the past earthquake. However, this study approach which understand the source parameters that is a major influence the building structure, can be used for the ultimate designing of building and the facilities related to nuclear power that need to consider the maximum class seismic ground motions.

It is known that a pronounced torsional vibration is induced at buildings that are constructed on inclined bearing layers due to translational vibrations at the foundation with different amplitudes and phase characteristics during an earthquake. Although the seismic performances of the foundations of buildings constructed on inclined bearing layers have been discussed, only a few studies on the vibration characteristics of the whole building system have been performed based on observations and measurements. In this paper, the seismic-response characteristics of a seismically-isolated building constructed on an inclined bearing layer was analyzed using strong motion records and microtremor measurements. To understand the generation mechanism of the torsional vibration of the superstructure generated by the input motion from the foundation slabs, the relationships between the torsional response and input motions from the inclined bearing layer are examined. Firstly, the strong motion records are analyzed to confirm the vibration characteristics of the superstructure. Next, the steady-state vibration of the superstructure is analyzed from the microtremor record. The vibration characteristics, including the warping torsional vibration, are confirmed for each predominant frequency of the superstructure. Finally, we analyzed the torsional response of the building by a 3D-frame model consisting of the superstructure and seismically-isolated layer and a 2D-FEM of the subsurface structure with an inclined bearing layer. The effect of the amplitude and phase difference of the input motions on the generation of the torsional response is also evaluated. The conclusions of this study are summarized as follows:

1) From the analysis of the small-amplitude earthquake records, the superstructure has predominant frequencies of 1.2 and 6.8 Hz. The planar steady-state vibrations at these two frequencies are consistent with those of microtremors. It is possible that warping torsional vibrations as well as translation may occur during earthquakes.

2) From the microtremor measurement at the superstructure, there are a translation vibration at 1.3 Hz, a torsional vibration around the center axis of the building at 2.2 Hz, a bow-shaped vibration at 6.0 Hz (in which both ends and the center of the building are opposed to each other) and a warping torsional vibration of the building floor slab at 6.7 Hz. The warping torsional vibration is characterized by the twisting of the mat slab of the first floor, leading to an opposite rocking behavior at both ends of the building in the short-span direction and a consequent opposite horizontal movement at the top.

3) The effects of the amplitude and phase differences of the input motions on the warping torsional responses around 6.7 Hz of the superstructure were examined by the 3D-frame model and the 2D-FEM ground model. The influence of the amplitude difference of the input motion is small, while the phase difference between the two ends of the building has a large influence on the inducing warping torsional vibration.

In the near future, occurring the extreme ground motions exceeding the structural design level is predicted. If such the ground motions are input to a base-isolated building, the building may collide to the retaining wall and damage the building, so counterplans against the extreme ground motions are important issues. Aim to analyze the behavior of a base-isolated building when it collides with the retaining wall, we have proposed the simple modeling method for the lateral resistance of retaining wall with backfill soil using line elements. In this proposed method, it is modeled in detail with multi degree-of-freedom system, multiple springs and dashpots, it can represent behaviors close to the actual phenomenon. However, the number of elements is large, and the modeling load is a little heavy. At the time of preliminary structural design of buildings, simpler modeling is desired. Therefore, in this paper, we propose the reducing method for the multi degree-of-freedom system, multiple springs and dashpots of the simple modeling method into an equivalent single degree-of-freedom system, one spring and dashpot.

The height of the retaining wall in the equivalent single degree-of-freedom system is up to the collision position. We assume that the deformation of the retaining wall with backfill soil is equal to the displacement distribution according to the theoretical solution of the static loading of the cantilever, and perform the reduction according to the following four conditions; In both systems, “Sum of kinetic energy is equal”, “Sum of forces generated in dynamic interaction spring is equal”, “Sum of strain energy of dynamic interaction spring is equal”, “Sum of absorbed energy of dashpot is equal”.

As the results of comparing the lateral resistances of the retaining wall with backfill soil evaluated by the multi degree-of-freedom system and the single degree-of-freedom system, the hysteresis loop shapes of both of them corresponded well in the case where the excitation force was applied near the top of the retaining wall. Whereas, especially in the case of retaining wall thickness t = 0.2 m, the difference was large when the excitation force was applied near the base of the retaining wall. This is because the local plasticization caused deviation from the assumed displacement distribution. And the value which is multiplying the ratio of the equivalent rigidity of the retaining wall and the reduced dynamic interaction spring at the excitation position by the ratio of the excitation height and the equivalent thickness of the retaining wall is 5 or less, the accuracy may deteriorate.

Comparing the responses of the two systems by collision analysis, there was a large difference in the absolute acceleration of the superstructure, but the maximum response value distribution shapes and the time history of drift angle and shear force corresponded relatively well. In particular, when the retaining wall thickness t = 0.3 m, the difference in the upper of the superstructure is a little large, the maximum response of drift angle and shear force by the single degree-of-freedom system model was 1.10 to 1.66 times that of the multi degree-of-freedom system model. The seismic isolation layer response displacement was larger in the multi degree-of-freedom system model, which was 1.05 to 1.25 times that in the single degree-of-freedom system model.

In last few decades, numerous super high-rise RC buildings have been constructed in Japan. During some massive earthquakes, for example, the 2011 Tohoku Earthquake (Mw 9.0), the structural response characteristics of long-period buildings were affected by excessive shaking. In super high-rise buildings, flexural deformation induced by overturning moment significantly affects the lateral displacement, especially in the upper stories. Furthermore, flexural deformation is critical for evaluating the response including higher vibrational modes excited by pulse-like ground motion and grasping the performance of seismic damping devices placed between stories. However, the flexural deformation of super high-rise buildings is insufficiently validated, because there are a few strong motion records related to flexural deformation requiring vertical observation at plural corners in the same floor.

This study examines the dynamic behavior of a super high-rise RC building focusing on flexural deformation based on strong motion and microtremor records and seismic response analysis of a three-dimensional frame model. The building is a 22-story moment-resisting RC frame structure constructed in 1995 in Saitama Prefecture, Japan. Strong motion is observed at four points inside the building; three are at the corners of the building’s rooftop and one on the first floor.

First, using strong motion data, the flexural and shear deformation variations with nonlinear behavior of the building before and after the 2011 Tohoku Earthquake were investigated. The amplitude dependence related with decrease of natural frequency and increase of damping factor was made clear. Moreover, the decrease in normalized top rotational angle in the total deformation was also confirmed from the eigenmodes at the top of the building using the system identification based on the subspace method.

Second, simulation analyses of the strong motion records are performed using a three-dimensional frame model consisting of a superstructure and rocking spring at the base. The model is validated by comparing the simulated horizontal and rotational responses with the observed records at the top of the building during a small earthquake and the 2011 Tohoku Earthquake. The relationships between maximum response values, story shear forces, and inter-story displacements obtained from the pushover analysis confirmed the decrease in the inter-story flexural deformation rate because of the nonlinearity of the shear component during the 2011 Tohoku Earthquake. Obtained results are consistent with the decrease in the flexural deformation inferred from the strong motion records.

Finally, the detailed distribution of vertical displacement regarding the flexural deformation is investigated by microtremor measurements including plural points at the rooftop. The distribution of vertical displacement at the top of the frame along the building perimeter is different from those along the center line with a central void, which is also indicated by the simulated results using the three-dimensional frame model.

Recently, large scale time history response analyses are increasing mainly due to the large improvement of computer performance. Since it is well known that the internal damping of many materials is comparatively frequency independent, the Rayleigh damping model is often used for such kind of analyses. However, the accuracy and the availability of the model are not high when the users want to fit the damping ratio from low to high frequency modes.

On the other hand, modal damping shows frequency independence in a wide frequency range, but since the damping matrix is dense, it is difficult to use due to the problem of storage area in large-scale analysis.

Hysteretic damping is known as damping that does not depend on frequency, but it is difficult to use it for time history analysis because of causal failure. However, recent studies have devised a causal hysteresis model that approximately satisfies the causality within a limited frequency range. Furthermore, as a more practical damping model with a reduced computer load, a damping model combining Rayleigh damping and causal damping was devised.

In this paper show its practicality and verify, Considering the application of frequency-independent damping to large-scale time history analysis that has not been feasible until now, comparing the required memory area, calculation time, and modal viscous damping with the classical damping model.

In order to confirm the validity and effectiveness of extended Rayleigh damping, this damping model is applied to a realistic building as an example, and the damping constant for each mode is evaluated from the free vibration analysis and seismic response analysis results. Based on the results, we first examine the fitness of the damping amount in the assumed adaptive frequency range using a simple multi-mass system model, then perform free vibration analysis, compare the response with the conventional damping model, and Benchmark test required memory area / calculation time.

As a result, followings were obtained.

1) For only horizontal or only vertical seismic motion, the accuracy and the availability of ordinal Rayleigh damping is not bad. However, for simultaneously horizontal and vertical or 3-dimensional seismic motions, the accuracy of ordinal Rayleigh damping is not enough.

2) On the contrary, the accuracy and availability of extend Rayleigh damping are confirmed for these seismic motions since its responses correspond well to those of modal damping.

3) For large scale models, computational load of extend Rayleigh damping are almost the same as that of ordinary Rayleigh damping and much less than that of modal damping.

4) From above, the availability of extended Rayleigh damping for large scale models are confirmed.

In Japanese Recommendation for Design of Building Foundation, the method to estimate steel pile’s ultimate strength in the liquefied soil at pile’s collapsing due to the significant earthquake is proposed. However, there are some cases, in which steel piles do not fail and continue to carry the dead load, even though they become plastic due to the previous seismic motion. The ultimate strength of steel piles, which have already accumulated the damage due to the multiple earthquakes, may decrease and the piles may collapse when they experience the subsequent massive earthquake. In this paper, centrifugal tests for the case where piles are subjected to multiple earthquakes are conducted to clarify pile’s collapse mechanism in the liquefied soil and pile’s ultimate strength. Furthermore, the estimation method of pile’s cumulative damage is presented based on results of centrifugal tests and numerical analyses.

Fig. 1 shows the model and instruments. The specimen consisted of a superstructure and a footing beam with mass, two bending plates, four piles, and a saturated sand layer. Table 2 shows specimen parameters, which are pile’s diameter, the relative density of the soil and the height-to-width aspect ratio of the superstructure. The centrifugal tests were performed under the centrifugal acceleration of 40 g.

Figs. 11(b), (c)-16(b), (c) show response time histories of Case 4. The bending strain at the pile head for the first shaking reached ε_lc after the soil liquefied and superstructure’s inertial force became maximum. The piles became inelastic although they kept their ability to carry the dead load even after shaking. The maximum value of superstructure’s acceleration and pile’s varying axial force during the second shaking were almost the same as that of first shaking. On the other hand, the bending strain gradually increased and reached the maximum value, ε_b,max, at 35 s. As shown in Fig. 19, the bending strain exceeded ε_yc for Case 2, but did not reach it in Case 1, Case 3, and Case 4. For all specimens, the local buckling occurred at the pile head, as well as at the bottom and the center of the pile as presented in Photo 1.

Fig. 20 shows the relationship between pile’s strength on centrifugal tests and the M-N interaction curves of design criteria. For Case 2, Case 5-Case 8, which collapsed during the first shaking, results at ε_b,max exceeded the M-N interaction curve of Japanese Recommendation for Design of Building Foundation and were distributed roughly following the ultimate strength curve. However, for Case 1, Case 3-Case 4, which collapsed subjected to multiple shaking events, the values at ε_b,max did not reach the ultimate strength curve and were smaller than those of Case 2, Case 5-Case 8.

Fig. 21 represents the calculation flow of pile’s accumulated plastic strain amplitude, ∑ε_pa, which is defined as pile’s cumulative damage indicator in this paper. Mechanical characteristics’ relationships (as shown in Figs. 26(a)-(c)), as well as regression equations (Eqs. (15)-(18)) are obtained based on analysis results of Figs. 22-25. Eq. (19), which is derived from Eqs. (15)-(18), shows the relationship between normalized ∑ε_pa and plastic deformation capacity μ_cmax. Fig. 27 describes the comparison between the values of centrifugal tests, numerical analyses, and Eq. (19). It is concluded that the cumulative damage of the steel pile, which collapses subjected to multiple earthquakes, can be evaluated using Eq. (19).

The topology optimization method using voxel finite element method is an effect Five method to create various morphologies from rectangular parallelepiped design domain. Fujii et al. [1,2] created morphologies of building structures using such method. Also, Fujii and Yamashita et al. [3] have developed a method that can consider finite deformation by using the particle method (HMPS method) instead of the finite element method.

By the way, in the topology optimization using such a voxel analysis method, the individual element densities in the initial morphology are generally uniform. However, in 3D analysis, the number of elements (voxels) in the fixed design domain is hundreds of thousands, whereas the number of elements of the target solution is often about 10% of it. Therefore, if an optimal solution is obtained starting from an initial morphology with a small number of elements, the computational efficiency can be greatly improved especially in the analysis using the particle method. In addition, in the optimization problem, there is also a problem of making modifications while respecting the initial morphology set by the designer. For this reason, it is desirable that the topology optimization method obtain an optimal solution from any initial morphology. In this paper, we propose CA-IESO, which is a combination of IESO (Improved ESO) method proposed in Refs.[1,2] and CA method, as a topology optimization method that can evolve from various initial morphologies.

In this paper, in order to verify the effectiveness of the proposed method, we first analyze the examples shown in Refs. [8,9]. Then, by comparing the solutions obtained from the three initial morphologies, we verify that they do not fall into local optimal solutions. In the two-dimensional problem, we also compare the solutions with the solutions obtained by CA-ESO method proposed in the past. Next, we analyze an example of creating a structure that supports a floor slab subjected to vertical loads and seismic loads.

The conclusions are as follows.

(1) From the initial morphology in which the initial density of all design target elements is 1, many optimal solutions near the global optimal solution can be obtained.

(2) When evolving from the initial morphology with small number of elements, the solutions near the global optimal solution can be obtained by setting the optimization parameters appropriately. However, depending on the initial morphology, no matter how the optimization parameters are adjusted, a local optimal solution may be obtained.

(3) By comparing the proposed method with CA-ESO method, the optimal solutions similar to the proposed method can be obtained by CA-ESO method. However, CA-ESO method does not provide as diverse solutions as the proposed method.

(4) In the case of morphological creation of the structure that supports the floor slab, the proposed method give an organic morphology like a tree. In addition, in the proposed method, the evolution process from the initial morphology is obserbed, so it seems to be effective in the problem of correcting the initial morphology.

(5) Since the proposed method can analyze with smaller number of elements, the calculation efficiency at each optimization step is good. However, in order to converge to a solution near the global optimal solution, it requires two to five times as many steps. Therefore, the calculation efficiency is not always good when obtaining a solution near the global optimal solution.

The target surfaces of the frame supported membrane structures are generated by using the Gaussian and mean curvatures. The surface is discretized into a triangular mesh, and the Gaussian and mean curvatures are defined at the interior vertices of the surface based on the formulations of discrete differential geometry. The minimal surface with zero mean curvature, which is equivalent to the uniform stressed surface, is often used as the target surface because uniform stress distribution is desirable. However, while the membrane structure is generated by connecting planar membrane sheets; i.e. cutting patterns, the minimal surface cannot be developed to a plane without out-of-plane deformation. In this respect, the developable surface with zero Gaussian curvature may be desirable; however, it cannot realize uniform stress distribution. Therefore, in this study, the curved surface with a geometric property intermediate between those of developable and minimal surfaces is generated and used as the target surface of a membrane structure.

The Gaussian curvature at an inner vertex is defined using the angle defect, which is the difference between 2□ and the sum of the angles between edges connecting to the vertex. The mean curvature is defined using a cotangent formula for the mean curvature vector at the vertices. The developable and minimal surfaces are generated by minimizing the sum of the squares of Gaussian and mean curvatures, respectively. The positions of the vertices on the outer boundary of the surface are fixed, and the optimization problem is solved by using the z-coordinates of the interior vertices as variables. In addition, the intermediate surface can be obtained by solving a multi-objective optimization problem that minimizes the sum of squares of both Gaussian and mean curvatures. The constraint approach is applied to obtain a Pareto solution; i.e., the sum of squares of Gaussian curvature is minimized under the upper bound constraint on the sum of squares of the mean curvature.

After obtaining the target surface, it is flattened by minimizing the sum of squares of differences between the edge lengths of on the surface in three-dimensional space and those in its development diagram on a plane. The cutting pattern is obtained by shrinking the obtained development diagram according to the target stress. Then, the equilibrium shape of the membrane structure is obtained by installing the obtained cutting patterns to the frame. The equilibrium state is achieved by solving the optimization problem that minimizes the strain energy of the membrane elements.

The effectiveness of the proposed method is demonstrated in the examples of curved surfaces with cylindrical boundary shapes of various heights. The membrane structure is constructed with four cutting patterns. The intermediate surfaces obtained by the proposed method have shown to have intermediate properties between the minimal and developable surfaces with respect to Gaussian curvature, mean curvature, and surface area in every height. The intermediate surfaces have realized the most preferable distributions of stresses in two directions while the developable and minimal surfaces have realized the uniform stresses in only one direction. The average values of the stresses in two directions are also closest to the target stresses when the intermediate surface is used. Therefore, when the intermediate surface is used as the target surface of the membrane structure, the preferable equilibrium state can be obtained easily without optimization of the cutting pattern.

The folded plate structure is a structural form that increases bending rigidity and strength by folding the plate. When a plane is bent in waves, the in-plane and bending stiffness increases in the wave orthogonal direction, but the in-plane stiffness extremely decreases because the plane is meandering like a spring in the wave direction. Since shell structures resist external forces mainly by in-plane forces, it is considered that there is an appropriate arrangement of the corrugations according to the flow of forces when the shell surface is made into a corrugated plate to improve the structural performance.

In this paper, an optimization method is used to investigate optimal wave directions of corrugated shells.

The finite element method with triangular planar shell elements was used for the structural analysis of corrugated shells. However, we do not directly discretize the corrugated plate shell surface, but approximate the corrugated plate as an anisotropic plate and treat the constitutive law of the anisotropic plate in this study.

The objective function is defined as the strain energy under the self-weight, and the wave direction of each finite element of the shell structure is design variable. The quasi-newton method was used as the optimization method.

In the paper, the wave directions are optimized for two models with various boundary conditions, cylindrical shell and EP shell.

In all models, the wave directions, which were all given in the same direction in the initial solution, changed to a smooth distribution of forces flowing from the apex to supports in optimal solution. The strain energy was significantly reduced from the initial value. The membrane forces and bending moments, which occur in two directions in each element in the initial solution, changed to occur in only one direction in the optimal solution. With the proposed optimization method, we can obtain the wave directions with low strain energy and high stiffness under the self-weight. In all regions of the shell, the membrane forces and bending moments occurred in only one direction, and the wave directions tended to coincide with the direction of the principal stress.

To see the mechanical effect of the corrugated shells, we compared the mechanical properties of the optimal solution with those of a normal shell of uniform thickness without waviness. The constitutive law was replaced with that for an isotropic plate and the analysis was performed separately. The maximum displacements and the strain energies of the corrugated shells were smaller than those of the normal shells, but the normal shells were more advantageous in the case with many support points without free edge. The comparison of the corrugated shells and the normal shells on the maximum displacement and strain energies confirms the effect of applying corrugated plate to shell surface, although they may be disadvantageous depending on the support conditions.

Drift pinned joints are generally used in glulam or sawn timber structure, and evaluation methods for calculating characteristic values such as ultimate load, yield load and initial stiffness have been established.

On the other hand, there were a few previous researches for yield load based on European yield theory for CLT structure. For the initial stiffness, it is considered that the beam theory on the elastic foundation can be applied, but there are few researches on the calculation method. There is no research on the evaluation method of the ultimate load with CLT failure.

Therefore, we focused on the fact that the in-plane anisotropy of CLT is small, and proposed the calculation method of the ultimate load for CLT drift pin joint. In addition, a simple calculation method for the yield load and the initial stiffness is summarized based on the previous research. The accuracy of the evaluation methods was confirmed by the moment resistance test of the CLT wall leg joints using Sugi CLT (S60-5-5) and drift pins (φ16, SS400).

The conclusions were;

(1) The calculated values of the ultimate load considering group shear failure were in good agreement with the test results

(2) The evaluation method of ultimate load considering group shear failure was expressed as the sum of tensile strength and shear strength on the assumed failure cross sections.

(3) The calculated yield loads based on the yield theory were in good agreement with the test results.

(4) The calculated initial stiffnesses based on the beam theory on the elastic foundation were almost twice as large as the test results.

(5) In the scope of application (CLT thickness 90 mm to 270 mm, drift pin diameter 12 mm to 20 mm), the maximum deference of calculated values of the initial stiffness considering the multi-layer and that of the simplified single-layer was about ±20%.

Plywood-sheathed shear walls are widely used in timber structures as resistance elements against horizontal loads such as seismic and wind forces. When an in-plane shear force acts on a plywood-sheathed shear wall, the nails/wood screw joints between the plywood and post-and-beam resist the shear force, thereby securing the load bearing capacity and toughness required for the shear wall. However, the toughness cannot be sufficiently secured in some cases, for example, positive and negative cyclic loads such as the seismic force. In such a case, cyclic bending deformation occurs in the fastener and fatigue failure occurs because of low cycle fatigue. Because this phenomenon causes vulnerability to the mechanical behavior of the joints or the shear wall, the low cycle fatigue characteristics of the fastener in the load-bearing performance evaluation should be considered. Therefore, in this study, the load–deformation characteristics of the shear wall and its joints were estimated considering the low cycle fatigue characteristics.

The performance of the joints was estimated by modeling the load–deformation characteristics of the joints and multiplying the reduction coefficient considering the low cycle fatigue characteristics. The performance of the shear wall was estimated by the finite element analysis using this model. The results of the constant-amplitude cyclic bending test of the fastener were used for the low cycle fatigue characteristics of the fastener.

The single shear test of the joints and static lateral loading test of the shear wall were performed using different loading protocols (such as those according to ISO and Japanese Industrial Standards) for comparison with the estimated results. The fastener used in the test entailed 4.1×38 mm and 4.5×50 mm wood screws and CN50 nails. In the joints test, Japanese cedar lumber was used as the main material and coniferous plywood as the side material. In the shear wall test, Japanese cedar lumber, Douglas fir lumber, and coniferous plywood were used for the columns and base, beams, and face material, respectively. Additionally, the nails and wood screws were fastened at intervals of 150 mm for the face material.

The load–deformation relationship could be roughly estimated in the joint estimation. In the estimation of the shear wall, when the fracture properties in the experiment such as punching out or pulling out were different from the fracture of the fastener, the estimated toughness was sometimes lower than that of the experimental results. The properties that often indicate the breakage of the fastener could be roughly estimated.

The de-bonding of longitudinal rebars is a method of suppressing the damage range of reinforced concrete (RC) structures. According to previous studies, in RC beam members, the damage can be concentrated on the hinge position by de-bonding the longitudinal rebars. However, the restoring force characteristic becomes a slip type, and energy dissipation becomes poor. Therefore, in this study, to secure energy dissipation while suppressing the damage range, a structural system was adopted in which the buckling restraint brace (BRB) was attached to the RC frame with de-bonding of the beam longitudinal rebars. Structural tests and finite element analysis were conducted. The purpose was to grasp the effects of the de-bonding of beam longitudinal rebars on the structural behavior of buckling restrained braced RC frames with a guaranteed hinge position.

The specimen was one-half scale, assuming a high-rise RC building. The longitudinal rebars at the end of the beam were arranged in two layers, and the hinge position was separated from the beam–column joints by cutting off the longitudinal rebars in the second layer in the middle of the span. The RC frame and BRB were joined by a gusset plate installed on the beam. The yield axial force of the BRB was set such that the horizontal component was equivalent to the yield strength of the RC frame. Two specimens were used, and the presence or absence of de-bonding of the beam longitudinal rebars on the side where the moment was smaller than the assumed hinge position was used as a parameter. For loading, a horizontal force was applied, while an axial force was applied to the column.

From the test results, in the specimen with de-bonding of the beam longitudinal rebars, the damage to the beam could be concentrated near the cutoff position of the longitudinal rebars, and a spindle-shaped restoring force characteristic was obtained. However, the story shear force decreased by approximately 30 % compared with the specimen with bonded beam longitudinal rebars. This is because the tensile force of the tensile longitudinal rebars is small in the de-bonded section.

Next, a finite element analysis was performed to reproduce the two test specimens. In addition, to improve the decrease in story shear force of the specimen with de-bonded longitudinal rebars, a model in which the longitudinal rebars of the center of the span were fixed was analyzed. It was confirmed in the analysis that the decrease in the story shear force of the model that de-bonded the longitudinal rebars occurred because the tensile force of the tensile longitudinal rebars is small. In addition, in the model in which the longitudinal rebars in the center of the span were fixed, the tensile longitudinal rebars bear the same tensile force as in the model without de-bonding, and the same story shear force as in the model without de-bonded longitudinal rebars could be obtained. Furthermore, it was confirmed that, even if the longitudinal rebars in the center of the span were fixed, the de-bonding section would not be damaged, and the damage range could be suppressed.

Concrete filled steel tubular (CFT) columns are rational columns in which local buckling of the steel tube is suppressed by the filled concrete and the filled concrete is confined by the steel tube; the steel tube and concrete complement each other’s disadvantages. Therefore, CFT columns have structural characteristics of high strength and high toughness. For this reason, CFT columns are widely used in high-rise buildings because they allow the column section to be reduced and the effective space of the building to be increased compared to steel reinforced concrete columns or reinforced concrete columns.

Recently, taller building and wider living space have required, sustained load and seismic force of column are increasing. Therefore, the application of high-strength materials to CFT columns is desired. Some studies on CFT columns using high-strength materials have been conducted and useful data is being accumulated. However, there are still many unclear points about the basic structural performance required for the design, such as ultimate flexural strength, limited displacement and so on.

To grasp the structural performance of square CFT columns using high-strength materials (590N/mm² class steel and Fc150 N/mm² concrete), six specimens were tested in this study. The test variables were width-to-thickness ratio and clear span-to-depth ratio. Using the results from this test and past tests, ultimate flexural strength, limit displacement and so on are discussed. In addition, a formula for evaluating the ultimate flexural strength and limit displacement was proposed. The major findings from this study are follows:

1) In the high-strength square CFT column, after the maximum flexural moment, axial displacement progressed due to local buckling of the flange/web of the steel tube at the hinge region and crushing of concrete. In addition, the flexural moment gradually decreased accordingly.

2) The lower limit of ultimate flexural strength can be evaluated by correcting the concrete strength reduction coefficient using the normalized width-to-thickness ratio of the steel tube as a parameter.

3) As an index to evaluate the limit displacement, the ratio of the concrete stress at the ultimate flexural strength (_cN_u) and the axial compressive strength of the column (N₀) was proposed. In addition, a formula for evaluating the limit displacement using _cN_u / N₀ as a parameter was proposed.

Many studies have been conducted to clarify unknown input/output systems based on measured data obtained via experiments and observations. Specifically, neural networks, which have been significantly developed in recent years, can be used for equation discovery. However, the identified networks are extremely large and difficult to understand. Although many studies have been conducted on mathematical expressions that are easy to understand, there is a paucity of studies on methods for determining a simple equation by eliminating unnecessary terms.

In this study, a novel method is proposed for identifying unknown equations by using mode extraction via singular value decomposition. This method is based on the modal iterative error correction method, which is effective for solving inverse problems with strong discontinuities. Additionally, the proposed method includes a process for removing null modes to obtain a simple equation without unnecessary terms.

The key findings of this study are as follows:

(1) We consider equation discovery as an inverse problem of unknowns such as coefficients and exponents. Thus, a novel method is proposed wherein the “process of improving the precision of an equation” and the “process of deleting unnecessary terms” are alternately iterated based on singular value decomposition results of Jacobian matrices of unknown parameters. This method enables the identification of functions consisting of only essential terms and ensures the reproducibility of unknown input/output systems.

(2) The applicability of the proposed equation discovery method is confirmed via a sample problem on cantilever deflection. Thus, an equation of an input/output system composed of only essential terms (i.e., excluding unnecessary terms) can be identified with high precision by using correct answer values that are provided as training data.

(3) It is confirmed that the proposed method is effective even in the presence of input variables that are unrelated to the input/output system. Furthermore, it is confirmed that “setting tolerance errors” and “expansion of the range for null modes” are necessary, when measurement error is included in the training data.