Experimental research was conducted to clarify fundamental characters of high-strength concrete plasticized by air-entraining and high-range water-reducing admixture. Plasticized high-strength concrete was water cement ratio of 28 to 48%, and target slump-flow was fluidized to 50cm using air-entraining and high-range water-reducing admixture.

 Investigated items are adsorption characteristics of air-entraining and high-range water-reducing admixture and influence of the difference of various conditions of water cement ratio, unit water amount, slump before plasticization, environmental temperature, fluidization time on properties of plasticized high-strength concrete.

 As a result, air-entraining and high-range water-reducing admixture showed different adsorption characteristics during mixing and plasticization, and at plasticization, fluidity could be increased with less use than at mixing. Considering the adsorption characteristics of air-entraining and high-range water-reducing admixture, the agglomeration characteristics of cement is thought to change markedly within 180 minutes from 120 minutes after mixing. The decrease in slump-flow amount after plasticization of plasticized high-strength concrete is larger than that of standard high-strength concrete. Also, slower the plasticization time, larger the decrease in slump-flow amount. The setting time of plasticized high-strength concrete tends to be faster than standard high-strength concrete, and the difference increases as environmental temperature to colder. Among properties of hardened concrete, effect of plasticizing by air-entraining and high-range water-reducing admixture on compressive strength, Young's modulus and drying shrinkage ratio was small.

 From the above results, high-strength concrete plasticized by air-entraining and high-range water-reducing admixture has same points to be noted as conventional plasticized concrete, such as decrease in slump-flow amount after plasticization is larger and setting time faster. However, it is possible to obtain hardened concrete comparable to standard high-strength concrete while reliably obtaining necessary slump-flow.

 Hot-dipped zinc coating is well recognized as a standard method of long-term protection from corrosion and is widely applied to heavy steel products such as building structures in Japan. In this process, liquid metal assisted cracking (LMAC) has sometimes occurred at vent and drain holes which were located at the corner of welded beam-to-column connection during hot-dip galvanizing. However, the mechanism of the phenomenon has not been well understood. To develop the understanding of the phenomenon, This study focuses on the effect of geometrical parameters such as the flange-to-web thickness ratio, the web thickness, the diameter of the holes, and their position on LMAC to develop the understanding of the mechanisms through which LMAC that initiated from the edge of holes and seek to identify an effective way for preventing them in practice. As the pilot study, Immersion tests were performed with scaled test specimens to investigate the thermal and strain histories due to the thermal expansion and contraction in hot-dip galvanizing process. A companion analytical model shown in Fig. 4, which is capable of simulating the test results was developed. The verified numerical model was used to develop the helpful index for evaluating the effect of geometric parameters on LMAC, in term of the tensile strain at the point corresponding to LMAC occurrence shown in Fig, 11. From the investigations, the following conclusions were drawn:

 Immersion testing with the flange thickness as the test parameter showed that as the flange thickness increases, the strain at the circular hole edges also increases. For a web thickness of 9 mm, LMAC occurred at two locations at the top of the test specimen when 32 mm flange was attached. Furthermore, microscope images (SEM) of the cracks and the behavior of the strain in the vicinity of the circular holes showed that the LMAC occurred while the test specimen was immersed in the molten zinc.

 Increases in the tensile strain, which is defined as the difference between the maximum compressive strain at the strain concentration point at the edge of the circular holes during and the strain at steady state, cause significant increases in LMAC.

 As the flange-to-web thickness ratio increases, the tensile stress that occurs at the hole edges increases as well, and the occurrence of cracking also increases significantly. However, even for cases in which the flange-to-web thickness ratio is the same, it is possible to reduce the tensile strain by increasing the web thickness. Therefore, increasing the web thickness, as well as reducing the flange-to-web thickness ratio, is effective in preventing cracking.

 Increasing the diameter of the circular holes by a factor of 1.7 results in an approximately 20% reduction in the tensile strain. Increasing the distance between the circular hole edges and the weld toes within the range over which the holes still function as drain holes (L = 4 - 24 mm) has small effects on reducing the tensile strain.

 In earthquake risk management, it is important to conduct probabilistic seismic hazard analysis (PSHA) and determine the relationship between seismic hazard and probability. When evaluating seismic risk for a region, the spatial correlation of ground motion must be understood. In this study, we attempt a spatial PSHA with spatial correlation and correction based on earthquake observation records recorded in the target area.

 To model the spatial correlation of the ground motion, we create an earthquake ground motion dataset from the records of K-NET and KiK-net operated by NIED. Ground motion intensity measures in the dataset is peak ground velocity on engineering bedrock which are based on the records observed in Kanagawa prefecture for the target site in this study. There are 1316 records by 196 earthquakes including the 2011 Tohoku great earthquake in the dataset. We model the spatial correlation statistically using the dataset. We calculate the residuals between actual and estimated ground motion of dataset and obtain the relationship between the correlation factors of residuals and the distances for pairs of observation stations. When calculating the residuals, we correct the predicted ground motion intensity by the empirical attenuation equation with correction factor for earthquakes and observation stations obtained from dataset analysis.

 We apply the spatial correlation of ground motion and the correction at the observation points.to spatial PSHA. When conducting spatial PSHA, we employ Monte-Carlo simulation to determine the probabilistic characteristics of the spatial seismic hazard. First, we generate the samples of PGV on the engineering bedrock of the observation stations in consideration with spatial correlation and correction factor. Next, PGV of other places in the target area are interpolated with the PGV samples at observation stations by simple kriging method. Finally, we obtained PGV at ground surface by soil amplification factor calculated from VS30 and calculate JMA seismic intensity from PGV at ground surface. We repeat these calculation and aggregate into the hazard curve.

 Trial application for Kanagawa prefecture is studied. Target is rectangle area including Kanagawa prefecture. There are 83 observation stations in the area. And, seismic source model and ground motion prediction model are followed The National Seismic Hazard Map for Japan. Results of spatial PSHA is expressed by hazard curve which shows relationship between the exceedance probability and area ratio exceeding to given ground motion intensity relative to the total area of Kanagawa pref. By comparison the results to the hazard estimation result in the earthquake damage assessment by Kanagawa pref., maximum possible earthquake for regional disaster planning are discussed based on the quantitative hazard results for the region.

 Unbalanced snow accumulation resulting from wind flow represents a difficult problem when predicting snow loads on building roofs. Accumulation occurs because of the complex interaction between snow particle motion and fluid flow, which is affected by building geometry. Recently, computational fluid dynamics (CFD) was applied to the prediction of snowdrift around buildings in several studies. However, few studies have applied CFD to the prediction of roof snow. In this study, the CFD simulation method of the snowdrift around buildings as previously proposed by the authors was applied to the prediction of roof snow accumulation on a two-level flat-roof building. The performance of the prediction model was validated by comparing the results of field observations and wind tunnel experiments. In particular, the importance of the effects of snow distribution change on the flow field was investigated.

 The snowdrift model was implemented in ANSYS FLUENT 14.5 as a user-defined function. To model the suspension of snow, the transport equation for drifting snow density was solved. The deposition and erosion fluxes on the snow surface were determined as functions of the friction velocity at the snow surface. The proposed snowdrift model was applied to the prediction of snow distribution on the two-level flat-roof building placed perpendicular to the approaching wind flow. The building model used in this study was adopted from the detailed field measurements obtained in Hokkaido, Japan. In this study, the following two cases were considered: (a) lower roof located on the windward side (windward case) and (b) lower roof located on the leeward side (leeward case).

 First, the velocity vectors obtained from the wind tunnel experiment for the same configurations as well as the streamlines obtained by the present CFD were compared for each case. For the windward case, two vortices, i.e., a recirculation flow resulting from separation at the upwind corner and a standing vortex near the roof step, were observed on the lower roof. For the leeward case, considerable recirculation flow on the lower roof was present. The general patterns of the flow field were very similar between the experiment and CFD. Next, the snow depth ratios obtained by a single CFD, in which the effects of snow distribution change on the flow field were not considered, were compared with the field measurements. For the windward case, the peak of the snow depth near the upwind edge of the lower roof, which was not observed in the measurements, was observed in the CFD results. This peak appeared in CFD and was attributed to the recirculation flow at the windward corner of the lower roof. The snow depth on the lower roof was generally underestimated in CFD because of the large erosion caused by the strong separation flow. Furthermore, a large undulation near the roof step, which was observed in the measurements, was not reproduced in CFD. For the leeward case, CFD also underestimated the snow depth and failed to reproduce a large undulation near the roof step. Finally, a phased CFD, in which the snowdrift calculations were conducted stepwise by considering snow depth change on the flow field, was applied to the same target. The underestimation of snow depth observed in the single CFD was clearly improved by the phased CFD. The prediction accuracy of the large undulations near the roof step was also improved, and it was confirmed that reproducing the separation flow near the upwind corner of the building correctly by considering the snow depth distribution on the roof is essential.

 Use shape factors when solving shear deformation in Timoshenko’s beam theory. Many books show that shape factors can be derived by energy methods.

 Chapter 2 shows that a simple method can derive shape factors easier than energy methods. The shape factor of rectangular cross section obtained by the method is 1. However, the shape factor of rectangle obtained by energy method is 6/5.

 Chapter 3 shows how to obtain the shape factor by a conventional energy method. Here, the author organizes the relationship between external and internal work on element of micro length. Shape factors can be calculated as external work equals internal work, but external forces used for external work do not satisfy equilibrium of forces.

 Therefore, in Chapter 4, the author proposes to add a moment load as external force and how to obtain shape factors considering its external work. According to the method, the shape factor of rectangular cross section is 1.

 In Chapter 5, the author proposes the method to separate bending and shearing of a beam which satisfies principle of superposition. This shows that the moment load applied in Chapter 4 interposes in that separation. Shear deformations of the beam also show that shape factor requires the same correction in Chapter 4.

 In Chapter 6, the author verifies that shape factor of rectangular cross section is 1 by FEM analysis.

 When obtaining angle and deflection in shear deformations, it is necessary to correct shape factors obtained by energy method. For a rectangular cross section, the shape factor changes from 6/5 to 1.

 During the 2011 off the Pacific coast of Tohoku Earthquake, many buildings suffered destruction of nonstructural components, especially fall of ceiling. Cantilevered roofs with ceiling have similar risk. The building standard law of japan determines the seismic vertical design loads as 1G, but the basis is not shown. Large span cantilevered roofs may severely vibrate not only horizontally but also vertically and the response characteristics are strongly affected by the main structure.

 On the above backgrounds, firstly in this study, we attempted to identify the vibration characteristics of an actual cantilevered roof by performing a microtremor measurement. The building has a 14m×26m cantilevered roof which was built in Tokyo. The cantilevered roof has a suspended ceiling and the beam section is tapered toward the tip.

 By analyzing the data, we found that cantilevered roof had two major vertical vibration modes, the vibration of cantilever itself and the vibration exited by the sway movement of main building. When these two modes resonate, vertical response acceleration of cantilevered roof might be very severe even if the building is subjected to only horizontal ground motion.

 Secondly, seismic response of cantilevered roofs was evaluated using parametric analyses of simplified two-dimensional frame models. Vertical deformation of the cantilever appeared commonly in the first mode of building sway and the second mode of cantilever vibration. The natural periods ratio between the first mode and the second mode is defined as T₂/T₁. When T₁ and T₂ are close to each other, the sway deformation of main building and the vertical vibration of cantilever appear the same time. For understanding the resonance in seismic response between the cantilever and the main building, a study with time-history analysis is conducted. It was found that the peak vertical response can be estimated as a function of the natural period ratio of the roof to that of the main structures, as found on shell-roofs on main structures. Amplification characteristic by the effect of the main structure is analyzed. It is evaluated as a function of R_T. R_T is the natural period ratio of an equivalent SDOF of main structure T_eq to the cantilever’s own natural period T_R. Next, we investigated the effect of R_M on the response amplification of cantilever around R_T=1 by changing M_eq. R_M factor is a mass ratio of equivalent main structure M_eq to the cantilever’s total mass M_R. Where R_M factor is large, the amplification of the cantilevered roofs increased.

 Finally, the acceleration response of cantilevered roofs is analyzed using response spectrum analysis with SRSS method of the simplified the inverted L-shaped two-degree-of-freedom model (L2DOF model). The building is converted into the L2DOF model based on push-over analysis. The two masses represent the equivalent main structure and the cantilevered roof. By using the L2DOF model, the response of large-span cantilevered roof can be expressed very simply but the calculation was found close to the numerical analysis.

 The authors have proposed effective steel connections achieving high bending stiffness and strength for timber grid-shell structures, and confirmed their performances through real size mock-up tests. In this research, reflecting the test results, the buckling strength of 24m-span timber grid-shell with and without diagonal bracing roofs are discussed. Their theoretical buckling strength including the rotational stiffness at connections are derived using continuum shell analogy, and compared with the results of discrete FEM analyses. Finally, the reduction factor equations of buckling strength as the functions of in-plane / out-of-plane bending stiffness ratio and two directional rotational stiffness are proposed, followed by confirming their validity.

 Dimensions of the studied single-layer timber latticed grid-shell are assumed to be 24 m × 24 m. The discrete FEM analysis models as Fig. 3, 4 are constructed using the fiber models with two-directional rotational springs at connections. The results of the real size mock-up bending tests are reproduced with the constructed FEM model. They showed good agreement and the validity of the FEM models are confirmed. Using the constructed FEM model, the grid-shell roof with and without diagonal bracings, with various proportions of timber sections, and various in-plane / out-of-plane rotational stiffness are constructed and their elastic buckling strength including geometrical nonlinearity are investigated. Generally, the bending stiffness of timber members and rotational stiffness of the connections are lower in in-plane directions (along z-axis) than in out-plane directions (along x-axis) in the test results. As the results of analyses, the buckling strength of the roof decreases along the stiffness of in-plane stiffness of the timber member (I_z) decreases, and also the rotational stiffness of the connections along in-plane (K_θz) and out-of-plane (K_θy) direction decreases. Therefore, estimating the effects of these stiffness is essential for evaluating the buckling strength of the grid-shell roof.

 To derive the equations for estimating the effects of these stiffness, continuum shell analogy is applied. The classic shell buckling theories are developed including the effects of the in-plane member bending stiffness and in-plane / out-of-plane rotational stiffness of the connections. The derived equations are simplified and the reduction factor equations of buckling strength are proposed as the functions of I_z / I_y and K_θz / K_θy. The obtained conclusions are summarized as follows.

 (1) The buckling strength derived from the continuum shell analogy gives values agreeing well with FEM analyses in grid-shell roofs without diagonal bracing. While those with diagonal bracing gives higher values than FEM analyses. This is because the shear stiffness of the roof panels is underestimated for neglecting axial deformations of timber chords.

 (2) The proposed reduction factor equations are confirmed to give reasonable values for grid-shell roofs without diagonal bracing and low-rise grid-shell roofs with diagonal bracing.

 1. Introduction

 In recent years, researches aim at developing a new structural system which uses nonstructural walls as structural elements of RC buildings proactively have been conducted based on past experiences in earthquake disasters^1-3). On the other hand, the latest earthquakes such as the Kumamoto earthquakes in 2016 revealed that RC buildings suffered serious damage to stiff members like columns with wing walls resulting in restoration/demolition, while they survived the earthquakes⁴⁾. Therefore, this study presents and verifies a new rebar arrangement for columns with wing walls which omits wall vertical rebar anchorage to let the wing walls resist only compression, thus reduce damage (Fig. 1). The current paper discusses a series of static loading experiments using three columns with/without wing walls with different confining reinforcement arrangements.

 2. Test plans

 A prototype building was designed according to the following concept: 1) to satisfy the base shear coefficient of 0.55 when considering wing walls in which wall vertical rebar anchorage was omitted, and 2) to maintain that of 0.3 with high ductility even though the wing walls fail under unexpected high seismic loads (Figs. 2-3 and Tables 1-3). Then, three 1:2 scale column specimens with/without wing walls on both sides representing the prototype building were designed: Specimen C without wing walls, Specimen CWJ with wing walls having confining reinforcement satisfying requirements for FA (with high ductility) based on AIJ Standard⁹⁾, and Specimen CWA with wing walls having higher confining reinforcement for earthquake-resistant design based on ACI code⁷⁾ (Figs. 4-6 and Tables 4-9). Static cyclic loads were applied to the specimens (Figs. 7-8 and Tables 10-11).

 3. Test results

 Compared with Specimen C without wing walls, Specimens CWJ and CWA with wing walls increased the initial stiffness and the maximum strength (Fig. 9). On the other hand, the experimental behavior and performance were similar between Specimens CWJ and CWA with different confining reinforcement arrangements. This resulted from no compression failure of confined core concrete observed in both specimens, which indicated that the confining reinforcement satisfying FA⁹⁾ was sufficient for the present specimens. The structural damage to all specimens were limited: the column suffered slight damage¹⁰⁾ with residual crack widths less than 0.1 mm up to the loading cycle to 0.5% rad, while the wing walls showed larger opening at the bottom because of the omission of wall vertical rebar anchorage (Figs. 10-12). Furthermore, stress of confining reinforcement in Specimens CWJ and CWA was limited which was likely to attribute to the omission of anchorage resulting in no tensile yielding of vertical rebar (Figs. 13-14). However, the observed equivalent damping factors of Specimens CWJ and CWA were smaller than that commonly used for practical design under large drifts because they showed slippage behavior in the hysteresis loops (Figs. 15-16).

 4. Conclusions

 In this paper, the structural performance of the columns with wing walls which omitted wall vertical rebar anchorage was experimentally evaluated. It was found that the confining reinforcement according to FA based on AIJ standard⁹⁾ provided high ductility with the drift capacity of more than 4% rad for the present specimens. The proposed omission of wall vertical rebar anchorage successfully limited not only damage to the specimens, but also stress of the confining reinforcement.

 In AIJ Standard¹⁾, it is required to check bond stress around each deformed bar in RC beams and columns. Bond strength formula is provided for three bond split modes: side split, corner split, and V-notch split modes (see Fig. 1). The bond strength of a tension reinforcing bar in multi-layers is influenced by bond stress transferred from adjacent bars. Therefore, the bond strength is reduced by 40% when the strengths of inner bars in multi-layers are calculated in accordance with the AIJ Standard. It is difficult to evaluate the bond strength with good accuracy because it is necessary to evaluate the bond stress of adjacent bars for calculating the bond strength of the bar in question. And therefore, in previous study²⁾³⁾, one of the authors have proved it is easier and more effective to evaluate bond capacity of entire tension bars, and proposed the bond capacity formula. This formula covers the side split mode (see Fig. 2), but it is still required to check the bond stress to avoid local bond failure, which some of bars in the tension reinforcement fail in bond.

 In this study, to evaluate strength on the local bond failure, pull-out tests of deformed bars straightly embedded in concrete were carried out. We prepared six specimens with double layers of reinforcement, and six with single layer. Parameters on the double layered specimens were the number of legs of transverse reinforcement, and placing or not placing bars inside of second layer (see Table1). The parameters on the single layered specimens were strengths and ratios of transverse reinforcement. To investigate influence of beam width, the width of specimens was extended to 350mm from 200mm of width of the previous specimens³⁾. As a result of the pull-out tests, the following conclusions were obtained.

 (1) The test results were classified into entire bond failure and local bond failure. All the bars reach the maximum bond stress at the same time in the entire bond failure, and some of the bars near the concrete surface failed ahead in the local bond failure. The effects of increasing transverse reinforcement did not be observed in the test results of single layered specimens due to the local bond failure.

 (2) It is showed that the bond capacity formula, provided in the previous paper³⁾, evaluates the effects of the beam width and the inner ties in the entire bond failure with good accuracy; only the factor for inner ties was modified.

 (3) The split modes in which more than one bar split concrete for upper side or lateral side were assumed, and methods to evaluate strengths on the local bond failure were proposed. The methods were developed from strength formulas for the corner split and the V-notch split modes.

 (4) The calculation by the proposed method showed good agreement with the test results. And the split modes obtained by calculation almost agreed with the modes in the tests; the calculated split modes were obtained from comparison of the capacities between the entire and the local bond failure modes.

 A compression member end-supported rotationally about its non-principal axis deflects in both directions of its principal axes. The elastic eigenvalue of this issue was theoretically solved by Rasmussen and Trahair, while neither elastic instability nor inelastic collapse of such struts with initial crookedness has been investigated.

 Thus, experiment was first performed as shown in Fig. 1. The strut is a steel flat bar, whose centroid and shear center coincide and then the flexural behavior is uncoupled with torsion. The flat bar is welded at both ends to a circular plate bolt-connected to a knife-edge block in order to alter the oblique angle between the knife-edge direction and the weak axis. Eight specimens in Table 1 were tested. Compressive stress-strain curves of the flat bar stubs are shown in Fig. 2, from which Young’s modulus and yield strength of the material are determined. Each specimen’s initial crookedness was measured as in Fig. 3, whose shapes are compared with cosine curves utilized in theoretical analysis.

 The experimental results are shown in the left-hand sides of Figs. 4 and 5 for elastic loading, and in Fig. 6 for collapse loading. From Fig. 4 showing the relationships between compressive load and mid-deflection due to weak-axis bending, the curves shift upward with an increase of the oblique angle. From Fig. 5 showing the relationships between mid-deflections about weak-axis and strong-axis, the deflection due to weak axis bending always follows the same direction of the initial crookedness, while the deflection due to strong axis bending does not.

 The load vs. lateral deflection curves of the flat bar struts are analytically derived by solving Eq.(1) so as to satisfy the boundary conditions of Eq.(6) considering the rigid-end offsets as well as the cosine crookedness and referring to the section forces and lateral displacement in Fig. 7. The solution is detailed in the appendix. The analytical curves are shown in the right-hand sides of Figs. 4 and 5 showing an acceptable level of resemblance with the experimental curves in the left hand sides. The major cause of the deviation is attributed to the difference of actual crookedness and assumed one. Buckling load as an eigenvalue can be obtained either by finding the load which makes the lateral deflection diverge or by solving the eigenequation of Eq.(a28) in the appendix. The elastic buckling loads P_cr of the tested specimens are calculated as in Table 2.

 The strength curve is influenced by the oblique angle as observed in Fig. 4, and also by the bending stiffness ratio of weak-axis to strong-axis. This is analytically investigated as exemplified in Fig. 8, where r is defined by Eq.(7). For a small r–value, i.e., for strong-axis rigidity much higher than weak-axis rigidity, the oblique angle is very influential for the strength curve. This is accounted for by high end-fixity of the strut as demonstrated in Fig. 9, where the end-fixity is represented by the minus sign ratio of end curvature to mid-length curvature about weak-axis deflection.

 The relation of maximum compressive stress to slenderness ratio is simply called the column curve, in which the slenderness ratio is the effective one regarding the end-fixity. The experimental data plotted in Fig. 10 are found well coincident with the column curves in current design specifications.

 A small sensor for the dynamic measurement of local moisture content within a wooden member exposed to fire heating is developed. Time variation of local moisture content of a wooden member during fire tests has not been grasped for the lack of non-destructive measurement system suitable for high temperature. This work intends to develop a heat resistive small sensor applicable to wooden specimen for fire tests to make it possible to monitor the distribution and time variation of moisture content within a wooden specimen just as thermocouples commonly used for the measurement of the distribution and time variation of temperature of a specimen.

 Importance of the measurement of local moisture content of a wooden specimen during fire tests comes from the recent advance in the recognition of the importance of moisture content in the high temperature performance of wood, e.g. significant influence of moisture content on the mechanical properties especially at high temperature and on the thermal decomposition as well as for the possible significant transfer of water within heated porous material such as wood due to the vaporization and re-condensation during fire. The possible influence of moisture content of wood may affect not only the charring rate of load bearing members but ignition and flame spread above wood based interior surface.

 The proposed measurement system is based on the unique dependence of electric resistance of wood on the moisture content and temperature. The present work consists of the design of a small heat resistant electrical resistance sensor and the development of calibration curves by experiments for the correlation of electrical resistance against temperature and moisture content for most typical tree species for building construction: Cryptomeria japonica, Larix leptolepis, Pseudotsuga menziesii and Zelkova serrata are chosen. Main results as follows.

 First, the specifications of the heat resistant sensor are shown in Fig. 1, which essentially monitor the electric resistance between the couples of the wires. Copper wire reported in a previous work for the measurement at room temperature ( Fredriksson et al, 2013 ) is replaced by electrically insulated nickel wire for the improvement of heat resistance. The upper temperature limit of electrodes the proposed sensor is 150℃; electrically insulated wires need to be wired not to be exposed to temperature higher than 150℃ during measurement.

 Second, the proposed sensors are applied to wooden plate specimens with the moisture content gradient from 6.9% to 232.8% conditioned at test temperatures from 20℃ to 90℃, for the determination of the relationship between electrical resistance and moisture content at each test temperature. The specimen was cut after the test to take samples around each sensor couples to confirm the moisture content by comparing the weight before and after drying. Relations between electric resistance and moisture content are found to differ for tree species.

 Finally, the measurement is found to be most effective within the moisture content 10% - 30%: below the lower limit, about 10%, electric resistance rises rapidly with decrease of moisture content, while above the upper limit electric resistance becomes insensitive to moisture content.

 In the future, time variation of local moisture content of wooden specimens under steady heating should be measured with the proposed sensors and measurement results should be compared with actual distributions of moisture content calculated from bone-dry weight to verify the validity of the measurement system and improve the prediction of heat and water transfer. Time variation of local moisture content of wooden members during fire furnaces test need to be measured with the sensor to evaluate influence of moisture content on mechanical properties.