RESIDUAL SEISIC PERFORANCE OF WOODEN BUILDINGS BY LOW COST SENSOR RECORDING AXIU CONNECTION DEFORATION itsuhiro iyamoto 1, Noriko Takiyama 2, Haruki Takahashi 3, Yugo Ishizuka 4 and Yasuhiro Hayashi 5 ABSTRACT: It is important to grasp the residual seismic performance of damaged buildings as soon as possible after an earthquake for the early recovery of life. The objective of this paper is to construct the system to judge easily the safety of damaged wooden buildings after an earthquake by residents themselves. First, the low cost sensor, which is supposed to be attached by residents themselves and record the maximum deformation of the column-to-beam connection, is proposed. Next, the static lateral loading tests of wooden frames are performed to clarify the relationship between the connection deformation and the story drift angle. In addition, the formula for the estimation of the story drift angle from the connection deformation is constructed. With the proposed low cost sensor and the constructed formula, the maximum story drift angle of wooden buildings during an earthquake can be easily estimated. KEYWORDS: Wooden frame, Deformation of column-to-beam connection, Story drift angle, Sensor 1 INTRODUCTION 123 Recently, several large earthquakes, such as the 27 Noto Hanto earthquake or the 27 Niigata-ken Chuetsu-oki Earthquake, have been occurred and many wooden buildings were collapsed. After these large earthquakes, because of the demolition of buildings to be used continuously or the fear of the aftershock, the issue has been risen up to be forced the residents to live at the emergency place for a long time. When a large earthquake occurs in the future, it is supposed to take a lot of time to judge the safety of damaged buildings, such as the possibility of a continuous use or the 1 itsuhiro iyamoto, Graduate Student, Dept. of Architecture and Architectural Eng., Kyoto Univ.,.Eng., Kyotodaigaku- Katsura, Nishikyo-ku, Kyoto, Japan. Email: rp-miyamoto@archi.kyoto-u.ac.jp 2 Noriko Takiyama, Research Associate, Dept. of Architecture and Architectural Eng., Kyoto Univ., Dr. Eng., Kyotodaigaku- Katsura, Nishikyo-ku, Kyoto, Japan. Email: noriko@archi.kyoto-u.ac.jp 3 Haruki Takahashi, Graduate Student, Dept. of Architecture and Architectural Eng., Kyoto Univ., Kyotodaigaku-Katsura, Nishikyo-ku, Kyoto, Japan. Email: rp-takahashi@archi.kyoto-u.ac.jp 4 Yugo Ishizuka, Graduate Student, Dept. of Architecture and Architectural Eng., Kyoto Univ., Kyotodaigaku-Katsura, Nishikyo-ku, Kyoto, Japan. Email: rp-ishizuka@archi.kyoto-u.ac.jp 5 Yasuhiro Hayashi, Prof., Dept. of Architecture and Architectural Eng., Kyoto Univ., Dr. Eng., Kyotodaigaku- Katsura, Nishikyo-ku, Kyoto, Japan. Email: hayashi@archi.kyoto-u.ac.jp necessity of repairs for damaged buildings, because of the wide damaged area or the shortage of experts. It is important to grasp the residual seismic performance of damaged buildings as soon as possible after an earthquake for the early recovery of life. eanwhile, the quick inspection of damaged buildings is conducted after an earthquake immediately, but the judgement of the safety is partly inaccurate and the severe damage can be missed since the degree of damage is decided only by watching. It is expected that the reliable system to judge easily by residents themselves the safety of damaged buildings after an earthquake immediately is constructed. The objective of this paper is to construct the method to evaluate the residual seismic performance of damaged wooden buildings by residents themselves. First, to estimate easily the maximum story drift angle of wooden buildings during an earthquake, the low cost sensor is proposed, supposed to be attached before an earthquake and to judge the safety of damaged buildings after an earthquake by residents themselves. Next, the static lateral loading tests are performed to grasp the relationship between the deformation of the column-tobeam connection and the story drift angle of wooden frames. In addition, the formula for the estimation of the story drift angle from the connection deformation is constructed. With the relationship between the damage degree and the maximum story drift angle during an earthquake, the residual seismic performance of damaged wooden buildings can be easily evaluated.
m1 c m2 D D m m θ R Figure 1: Deformation of wooden frame Fluorocarbon Loading direction Fixed δ max Fluorocarbon δmax δ max Graduation Guide Before an earthquake During an earthquake After an earthquake Figure 3: Behavior of sensor Bead (Graduation) Nail(Guide) Nail(Fixed on beam) Figure 2: Proposed sensor 2 PROPOSE OF SENSOR 2.1 DEFORATION OF COLUN-TO-BEA CONNECTION OF WOODEN FRAE When an earthquake occurs, the deformation of the column-to-beam connection of wooden frames is expressed as shown in Figure 1. The connection rotational angle θ and the separation displacement from the beam at the middle of the column c are expressed in Equation (1) and (2): θ = ( m1 m2)/2dm (1) c = ( m1+ m2)/2 (2) Where, mi is separation displacement at opposite positions D m distant from middle of column. 2.2 OUTLINE OF SENSOR The sensor proposed in this paper is supposed to be attached on the column-to-beam connection by residents themselves so it is necessary to get easily the material of the sensor. In the test, the sensor is made of the fluorocarbon, which is used for fishing lines, nails, wires and beads. The following procedure is the method to make the sensor: The edge of the fluorocarbon is fixed on a beam with a nail. The nail as the guide is tied with a wire so that the fluorocarbon can move only to the vertical direction along a column. The bead as the graduation is fixed on the fluorocarbon with an adhesive. Figure 2 shows the sensor before it is attached on the column-tobeam connection. The sensor operates only when a column is separated from a beam and can record the maximum deformation of the column-to-beam connection during an earthquake. Fixed Fluorocarbon Guide Graduation Guide Figure 4: Attachment of sensor δ 1 δ 2 δ 3 D m Sensor The behavior of one sensor as shown in Figure 3 is following: [Before an earthquake] The sensor is attached on the column-to-beam connection of wooden frames before an earthquake. The bead as the graduation between two guides is located at the original position. [During an earthquake] When the column-to-beam connection deforms during an earthquake, the bead as the graduation moves to the upper direction along the fluorocarbon by displacement δ max, where δ max is the maximum displacement of a column separated from a beam during an earthquake. [After an earthquake] After an earthquake, the fluorocarbon bends by displacement δ max because it is irreversible that the fluorocarbon returns to the original position. By reading the movement of the bead as the graduation, the maximum separation displacement δ max during an earthquake is estimated. 2.3 USE OF SENSOR Figure 4 shows the situation of the attached sensor. Three sensors are attached on one column-to-beam connection. One is attached on the middle of the column and others are attached on the opposite positions: same displacement distant from the middle of the column. D m Figure 5: Displacement of sensor
Loading direction Plus inus Actuator (15 24) À( 15 ~ 18) (15 24) À(15 ~18) 91 2798 À(15 ~18) Vertical load Specimen 2798 (15 15) Short tenon with metal connector Ground sill (15 15) 2798 1888 (15 15) Dry mud-panel Short tenon with metal connector Ground sill (15 15) Front column Rear column y ä i 15 ~15 j Fixed on frame 182 Figure 8: Two types of specimens 182 182 Figure 6: Loading instrument W U1 U 2 H Story drift angle[rad].3 1/5.2 1/6.1 1/1 1/8 1/2 1/15 1/12 1/1 1/75 1/5 1/3 -.1 Figure 7: Loading schedule Before an earthquake, residents themselves attach the sensor on the column-to-beam connection of their wooden building. After an earthquake, they read the displacement of the graduation. The maximum connection rotational angle θ max during an earthquake can be estimated from Equation (3): max ( δ1, δ3) δ2 θ = (3) max D m 1/3.5 1/4 Where, δ i is displacement of each sensor as shown in Figure 5. Furthermore, the maximum story drift angle R max of the column-to-beam connection during an earthquake can be estimated, by using the formula shown in chapter 5 for the estimation of the story drift angle R from the connection deformation θ. 3 OUTLINE OF STATIC LATERAL LOADING TEST 3.1 LOADING INSTRUENT Figure 6 shows the loading instrument in the test. The actuator capacity of loading is 1kN and that of displacement is 1mm. Specimens are fixed on the ground sill with anchor bolts. The plus and minus of the loading direction are decided and columns are named as shown Figure 6. Specimens are subjected to quasi-static cyclic lateral loading, gradually increasing the maximum story drift angle symmetrically from 1/12, 1/1, and so on, until the restoring force of specimens is kn. But, in the range of more than 1/15, they are increasing only plus l l m Q1 Q2 R C D1 D 2 Strain gage Figure 9: Force acted on wooden frame Displacement sensor 35 35 Figure 1: easurement of connection displacement because of the actuator capacity. Figure 7 shows the loading schedule of the test. 3.2 OUTLINE OF SPECIEN Figure 8 shows two types of wooden frame specimens in the test. The effect of a hanging wall on the deformation of column-to-beam connections has been investigated in the test to found where to attach the sensor. In addition, the relationship between the connection deformation and the story drift angle is clarified. The dimension of specimens is 182mm 2798mm. s and ground sills are made of cedar, and beams of Douglas fir. The short tenon with metal connector is used at the column-to-beam connection. The dimension of short tenon is 84mm in width, 3mm in thickness and 52.5mm in height. As a hanging wall, a dry mud-panel is adopted [1]. The vertical load is decided considering the separation between a column and a ground sill [2],[3]. The vertical load of wooden frame specimen is 9.3kN, and that of wooden frame with a hanging wall specimen is 23.7kN. Young s modulus of columns is calculated from Equation (4)-(6), considering the balance of force acted
Lateral loading H [kn] 2 1.5rad Nail is pulled out -1.125rad Bending fracture is occurred at tenon -2 -.1.1.2 on the specimen at the small story drift angle as shown in Figure 9 [4]: H = Qi cos RC W / l (4) Q = ( + )/ l (5) i Ui Di m.1rad Tension fracture is occurred at ground sill Figure 11: Relationship between story drift angle and lateral loading with damage situation (Wooden frame specimen) Lateral loading H [kn] 6 4 2.3rad Corner of hanging wall is cracked.1rad Bending fracture is occurred at column -2.5rad -4 Tension fracture is occurred at ground sill -6 -.1.1.2 Figure 12: Relationship between story drift angle and lateral loading with damage situation (Wooden frame with hanging wall specimen) Ui, Di = ( εai εbi ) E Z /2 (6) Where, H is lateral load, Q i is shearing force, R C is rotational angle of member, W is vertical load, is lateral displacement at top of column, l is length of column, ui,di is bending moment at top and bottom of column, l m is distance between strain gages, ε ai,bi is value of strain gage, E is average young s modulus of both columns and Z is section modulus of column. The average young s modulus of wooden frame specimen is 8.74N/mm 2, and that of wooden frame with a hanging wall specimen is 7.94N/mm 2. 3.3 EASURING INSTRUENT OF COLUN- TO-BEA CONNECTION The displacement sensors are attached on the column-tobeam connection at 35mm distant from the middle of a column as shown in Figure 1. The vertical displacement is measured to calculate the connection rotational angle. In addition, the displacement sensor and the proposed sensor are attached on the same position to investigate the accuracy of the sensor. The graduation of the sensor is read by watching with.1mm accuracy. Another displacement sensor is attached on the specimen to measure the story drift angle and the rotational angle of members. Strain gages are pasted to investigate the stress distribution. Damage situation is observed by watching. 4 TEST RESULTS 4.1 RELATIONSHIP BETWEEN STORY DRIFT ANGLE AND LATERAL LOADING Figure 11 shows the relationship between the story drift angle and the lateral loading with damage situation of the wooden frame specimen. The nails of the metal connector are pulled out at.5rad, the tension fracture of the ground sill is occurred at.1rad and the bending fracture of tenon is occurred at.125rad. Although the damage is occurred at the part of the specimen, the lateral loading is gradually dropped by the influence of P effect. Figure 12 shows the relationship between the story drift angle and the lateral loading with damage situation of the wooden frame with a hanging wall specimen. The corner of the hanging wall is cracked at.3rad, the tension fracture of the ground sill is occurred at.5rad and the bending fracture of both columns is occurred at.1rad. When both columns are broken and the lintel is separated from the column, the lateral loading is rapidly dropped. 4.2 RELATIONSHIP BETWEEN STORY DRIFT ANGLE AND CONNECTION ROTATIONAL ANGLE Figure 13 shows the relationship between the story drift angle and the ratio of the story drift angle to the connection rotational angle at each column-to-beam connection of the wooden frame specimen respectively. The near the ratio θ/r is 1, the smaller the ratio of bending deformation of the column to the story drift is. The ratio θ/r is similar to each other regardless of the position and is recorded in the range from.6 to.9. In addition, the ratio θ/r grows gradually as the story drift angle R grows. Figure 14 shows the relationship between the story drift angle and the ratio of the story drift angle to the connection rotational angle at each column-to-beam connection of the wooden frame with a hanging wall specimen respectively. The ratio θ/r is recorded in the range from to.3 at the top of a column and from 1. to 1.3 at the bottom of a column. This is because the rotation at the top of a column is restrained by a hanging wall so the bending rotation of a column is concentrated at the bottom of a column. Figure 15 shows the examples of the damage situation. The ground sill of the wooden frame specimen is damaged at.1rad but the ratio θ/r is almost equal. However, the ratio θ/r of the wooden frame with a
1.5 Top of front column Bottom of front column Top of rear column Bottom of rear column 1 θ 1 1 3 H R l θ/r 1 R R EI,.5 θ 2 2 2 4.1.2 Figure 13: Relationship between story drift angle and connection rotational angle (Wooden frame specimen) Figure 16: Force acted on wooden frame Loading direction Loading direction Loading direction 1.5 1 Top of front column Bottom of front column Top of rear column Bottom of rear column y 1 N 1 L L/2 µ 1 N 1 σ N w σ w N y µ 2 N 2 2 y 2 (a) Embedment and (b) Embedment frictional resistance of resistance of column tenon and beam column and beam r i sp i (c) Shear resistance of nail θ/r.5 Figure 17: Force acted on column-to-beam connection.1.2 Figure 14: Relationship between story drift angle and connection rotational angle (Wooden frame with hanging wall specimen) Tension fracture of ground sill (Wooden frame specimen) Figure 15: Example of damage Bending fracture of column (Wooden frame with hanging wall specimen) hanging wall specimen changes rapidly at.1rad when the bending fracture of both columns is occurred. It is found that the isolated columns are adequate to attach the sensor on the column-to-beam connection. 5 ESTIATION OF STORY DRIFT ANGLE FRO COLUN-TO-BEA CONNECTION DEFORATION In this chapter, focused on the connection between an isolated column and a beam, the method for the estimation of the story drift angle from the connection deformation is shown. 5.1 RELATIONSHIP BETWEEN STORY DRIFT ANGLE AND COLUN-TO-BEA CONNECTION DEFORATION The force acted on a wooden frame is shown in Figure 16: where H is lateral loading, i is moment of column edge and θ i is connection rotational angle. When the moment and the rotational angle of each connection are equal, the relationship between a story drift angle R and a connection rotational angle θ is expressed in Equation (7) from the slope deflection method: l R = θ + (7) 6EI Where, E is young s modulus of column, I is geometrical moment of inertia of column and l is length of column.
oment [kn.m] 3 2 1 Top of front column Bottom of front column Top of rear column Bottom of rear column Estimation.1.2 Connection rotational angle θ [rad] Figure 18: Relationship between connection rotational angle and moment.3.2.1 Test Estimation.1.2 Connection rotational angle θ [rad] Figure 19: Estimation of story drift angle from connection rotational angle 5.2 RELATIONSHIP BETWEEN COLUN-TO- BEA CONNECTION DEFORATION AND OENT The bending strength of the column-to-beam connection is calculated to estimate the relationship between connection rotational angle θ and moment. The factors of the bending strength consist of embedment resistance and frictional resistance of a column tenon and a beam (a), embedment resistance of a column and a beam (b) and shear resistance of nails (c) as shown in Figure 17. When the ultimate stress distribution of a column tenon is equal to the vertical direction [5],[6], the bending strength of a column tenon and a beam A is expressed in Equation (8): Ny 1 1+ Ny 2 2 x A = + ( µ 1N1 + µ 2N2) 2 2 (8) N = byσ, N = by σ, y + y = y 1 1 W 2 2 W 1 2 Where, σ W is embedment strength of tenon, y is tenon height, b is tenon width of short direction, x is tenon width of long direction and µ i is friction coefficient of column tenon and beam. The embedment strength of tenon σ W is twice that of cedar standard [5]. When the inflection point is located at the middle of a tenon and friction coefficient is.5 [7], Equation (8) becomes as follows: by( x + y) σw A = (9) 4 When the normal force acts equally on the edge of each column, the bending strength of a column and a beam B is expressed in Equation (1) [8]: WL B = (1) 4 Where, W is vertical load and L is column width. The bending strength of the shear resistance of nails C is expressed in Equation (11) [5]: = Pr (11) C s i i Where, S P i is shear resistance of each nail and r i is distance from column-to-beam connection to nails [9]. The bending strength of the column-to-beam connection all is expressed in Equation (12): all = A + B + C (12) The relationship between the connection rotational angle θ and the moment is idealized as the elastic perfectly plastic model. When the yield deformation angle θ y is.1rad and the ultimate moment u is all, the relationship between the connection rotational angle and the moment, compared with the test results, is shown in Figure 18. The estimation from Equation (12) is similar to the test results. 5.3 ESTIATION OF STORY DRIFT ANGLE FRO COLUN-TO-BEA CONNECTION ROTATION ANGLE The story drift angle R is estimated from the connection rotational angle θ using Equation (7)-(12). The estimation result is similar to the test results as shown in Figure 19. 6 ACCURACY OF SENSOR In this chapter, focused on the wooden frame specimen composed of isolated columns and beams, the accuracy of sensor is examined from the results of the tests. 6.1 ESTIATION OF AXIU SEPARATION DISPLACEENT Figure 2 shows the comparison of the estimated maximum separation displacement at each column-tobeam connection from the sensor and the test results. The maximum separation displacement can be estimated by the sensor though it is a little uneven. This is because of the play in the initial state and the error when the displacement of the graduation is read. 6.2 ESTIATION OF AXIU CONNECTION ROTATIONAL ANGLE Figure 21 shows the comparison of the estimated maximum connection rotational angle at each column-tobeam connection from the sensor and the test results.
aximum separation displacement δ [mm] max (sensor) The maximum connection rotational angle of each loading cycle is estimated from the sensor by Equation (13): θ max 15 1 5 5 1 15 aximum separate displacement δ max [mm] (test) Figure 2: Estimation of maximum separation displacement aximum connection rotational angle θ [rad] max (sensor).3.2.1.1.2.3 aximum connection rotational angle θ max [rad] (test) Figure 21: Estimation of maximum connection rotational angle aximum story drift angle R [rad] max (sensor).3.2.1.1.2.3 aximum story drift angle R max [rad] (test) Figure 22: Estimation of maximum story drift angle ( δ1 δ2)/ D = ( δ3 δ2)/ D m m (Plus direction) (inus direction) (13) Where, δ i is displacement of each sensor as shown in Figure 5 and D m is distance from middle of column to sensor. The maximum connection rotational angle can be estimated by the sensor within the range of more than.2rad. 6.3 ESTIATION OF AXIU STORY DRIFT ANGLE Figure 22 shows the comparison of the estimated maximum story drift angle each the column-to-beam connection from the sensor and the test results. The maximum story drift angle is estimated from the sensor as follows: The maximum connection rotational angle θ max is estimated by Equation (13). The relationship between the connection rotational angle θ and the moment is estimated by Equation (8)-(12). The maximum story drift angle is estimated by Equation (7). It is shown that the maximum story drift angle can be estimated by the sensor within the range from.2 to.2rad. 7 CONCLUSIONS In this paper, the low cost sensor, recording the maximum deformation of column-to-beam connections during an earthquake, has been proposed. The static lateral loading tests of wooden frames have been conducted, so that the maximum story drift angle of wooden buildings during an earthquake can be estimated easily. First, the effect on the deformation of column-to-beam connections, according to the existence of a hanging wall, has been investigated to found where to attach the sensor. Next, the formula for the estimation of the story drift angle from the connection deformation has been constructed. Finally, the accuracy of the low cost sensor proposed in this paper has been examined in the tests. The major conclusions of this paper are summarized as follows. [1] The rotation of the column with a hanging wall is different at the top and the bottom of a column because of the restrain by a hanging wall. [2] There is clear correlation between the connection rotational angle and the story drift angle of the isolated column. So the formula for the estimation of the story drift angle from the connection rotational angle has been constructed. [3] The low cost sensor recording the maximum deformation of column-to-beam connections has been proposed. From the results of the tests, it is found that the maximum story drift angle can be estimated by the sensor within the range from.2 to.2 rad. This sensor is supposed to be attached by residents themselves so there is a little difference in the accuracy of the sensor, which is the problem to be improved in the future. ACKNOWLEDGEENT A part of this research was supported by TOSTE Foundation for Construction aterials Industry Promotion. We would like to show our appreciation to Dr. T.orii of former Kyoto Univ. for his great contribution to the tests.
REFERENCES [1] Sugiyama R., Suzuki Y., Gotou. and urakami H., Development of bearing wall for timber frame by using dry mud-panels, AIJ Journal of Technology and Design, AIJ, No. 24, 26, pp.125-13. (in Japanese) [2] Editorial committee for manual of seismic design for wooden frame structures, anual of seismic design for wooden structures taking advantage of traditional structural techniques - ethods for seismic design and seismic reinforcement design based on response-limit capacity analysis -, Gakugei Shuppan Sha Co. Ltd, 24. (in Japanese) [3] Yamada., Suzuki Y., Gotou. and Shimizu H., Dynamic and Static Tests of Wooden Frames for Evaluation of Seismic Performance, Journal of Structural and Construction Engineering, AIJ, No.582, 24, pp.95-12. (in Japanese) [4] orii T., Takahashi H., iyamoto. and Hayashi Y., Experimental Study on Seismic Performance Evaluation of Wooden Frame Structure Part1 Effect of P Effect on Deformation Capacity of Wooden Frame Structure, Summaries of Technical Papers of Annual eeting, AIJ, C-1, 29, pp.525-526. (in Japanese) [5] Hirasaka T., Experimental Studies on the Ultimate Strength of the Wooden Construction Connections, Journal of Structural and Construction Engineering, AIJ, No.548, 21, pp.89-94. (in Japanese) [6] Haruyama S., Gotou., Nishimura T. and Suzuki Y., Experimental study on joints of wooden frame and proposal of an analysis model, Summaries of Technical Papers of Annual eeting, AIJ, C-1, 27, pp.141-142. (in Japanese) [7] Iwamoto I., Shimizu H. and Suzuki Y., Rotational Compression and Friction at Fundamental Connection odels of Timber Structure, Proceeding of the 12th Japan Earthquake Engineering Symposium, 26, pp.786-789. (in Japanese) [8] Kudo W., orii T. and Hayashi Y., Evaluation of Wooden House Deformation Capacity Based on Static Loading Test Part2 Estimated Horizontal Load of Wooden House, Summaries of Technical Papers of Annual eeting, AIJ, C-1, 28, pp.477-478. (in Japanese) [9] Architectural Institute of Japan (edited), Standard for Structural Design of Timber Structures, aruzen, 26. (in Japanese) [1] iyamoto., Takahashi H., orii T., Takiyama N. and Hayashi Y., Estimation of Story Drift Angle from Connection Deformation and Proposal of Sensor Recording aximum Deformation of -to-beam Connection of Wooden Frame Structure, Journal of Structural and Construction Engineering, AIJ, Vol.75 No.649, 21, pp.69-616. (in Japanese)