1NCEE Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 1-5, 14 Anchorage, Alaska FEASIBILITY STUDY ON SEISMIC RESPONSE ESTIMATION OF DAMAGED R/C BUILDINGS BASED ON OBSERVATION DATA AND NUMERICAL ANALYSIS H. Choi 1, Y. Sanada, Y. Watanabe and Y. Nakano 4 ABSTRACT On September, 9, at 5:16 pm local time, a magnitude 7.6 earthquake (USGS) struck off the southern coast of West Sumatra, Indonesia (.7S, 99.86E, Depth: 81km), and Padang city which is the capital of West Sumatra province suffered moderate/heavy damage. In particular, a large number of reinforced concrete (R/C) structure buildings were damaged seriously. Authors conducted two earthquake damage investigations in Padang city. During these investigations, Finance and Development Audit Agency building (BPKP: Badan Pengawasan Keuangan Dan Pebangunan), which is a five story R/C public office building, was selected as a typical R/C building in surveyed areas, and the evaluation of the damage classes to each member based on the Japanese guideline and the ambient vibration measurement using micro tremor were carried out in detail. The objective of this research is to analytically estimate seismic responses, in particular maximum drifts of earthquake-damaged R/C buildings. In this paper, a predominant period obtained by ambient vibration measurements and a residual seismic capacity ratio index R based on damage classes of columns were applied to evaluate an experienced maximum drift of BPKP building. As a result, the proposed method based on the R -index approximately reproduced damage to columns in the most severely damaged story, while the other method by the predominant period overestimated. Characteristics and scope of both methods were also investigated through the application above. 1 Research Associate, Institute of Industrial Science, The University of Tokyo, Japan, Ph.D. Associate Professor, Graduate School of Engineering, Osaka University, Japan, Dr.Eng. Graduate Student, Graduate School of Engineering, Toyohashi University of Technology, Japan 4 Professor, Institute of Industrial Science, The University of Tokyo, Japan, Dr.Eng. Choi H., Sanada Y., Watanabe Y., Nakano Y. Feasibility study on seismic response estimation of damaged R/C buildings based on observation data and numerical analysis. Proceedings of the 1 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 14.
Feasibility Study On Seismic Response Estimation Of Damaged R/C Buildings Based On Observation Data And Numerical Analysis H. Choi 1, Y. Sanada, Y. Watanabe and Y. Nakano 4 ABSTRACT The objective of this research is to analytically estimate seismic responses, in particular maximum drifts of earthquake-damaged R/C buildings. In this paper, a predominant period obtained by ambient vibration measurements and a residual seismic capacity ratio index R based on damage classes of columns were applied to evaluate an experienced maximum drift of BPKP building. As a result, the proposed method based on the R -index approximately reproduced damage to columns in the most severely damaged story, while the other method by the predominant period overestimated. Characteristics and scope of both methods were also investigated through the application above. Introduction On September, 9, at 5:16 pm local time, a magnitude 7.6 earthquake (USGS) struck off the southern coast of West Sumatra, Indonesia (.7S, 99.86E, Depth: 81km), and Padang city which is the capital of West Sumatra province suffered moderate/heavy damage. In particular, a large number of reinforced concrete (R/C) structure buildings were damaged seriously. Authors conducted two earthquake damage investigations in Padang city. During these investigations, Finance and Development Audit Agency building (BPKP: Badan Pengawasan Keuangan Dan Pebangunan), which is a five story R/C public office building, was selected as a typical R/C building in surveyed areas, and the damage class evaluation to each member based on the Japanese guideline and the ambient vibration measurement were carried out in detail. The objective of this research is to analytically estimate seismic responses, in particular maximum drifts of earthquake-damaged R/C buildings. In this paper, a predominant period obtained by ambient vibration measurements and a residual seismic capacity ratio index R based on damage classes of columns were applied to evaluate an experienced maximum drift of BPKP building. 1 Research Associate, Institute of Industrial Science, The University of Tokyo, Japan, Ph.D. Associate Professor, Graduate School of Engineering, Osaka University, Japan, Dr.Eng. Graduate Student, Graduate School of Engineering, Toyohashi University of Technology, Japan 4 Professor, Institute of Industrial Science, The University of Tokyo, Japan, Dr.Eng. Choi H., Sanada Y., Watanabe Y., Nakano Y. Feasibility study on seismic response estimation of damaged R/C buildings based on observation data and numerical analysis. Proceedings of the 1 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 14.
Damage Overview of an Investigated Typical R/C Building In this study, Finance and Development Audit Agency building (BPKP: Badan Pengawasan Keuangan Dan Pembangunan), which is a five story R/C public office building built in located in the central part of Padang city, is selected to analytically assume the seismic response after the earthquake. Photo 1 shows the whole view before and after earthquake of BPKP building, and Fig. 1 shows the third floor plans together with the damage classes of R/C columns and unreinforced masonry (URM) infills evaluated by the Japanese guideline [1]. The basic concept of this guideline is briefly described in Appendix. Since some structural plan was not able to be obtained, non-destructive tests using Schmidt hammer and re-bar locator were carried out during the investigations. Column cross-sections are shown in Table 1. In the third floor suffered the most serious damage, the damages to west side of the building are relatively severe as shown in Fig. 1, and the buckling and the rupture of reinforcing bars are found in some columns of rows 4 and 5 as shown in Fig. 1 and Photo (a). In rows and, exposed reinforcing bars are also observed in many columns as shown in Fig. 1 and Photo (b). The residual seismic capacity ratio index R values calculated by Japanese guideline are shown in Table. The R value for the transverse (EW) direction of the third floor is evaluated to be 54.1%, and the damage rating is judged as Heavy. (a) Before earthquake damage (b) After earthquake damage Photo 1. Whole view before and after earthquake damage of BPKP building - V: NS dir., - V: EW dir., - V: URM wall Figure 1. Floor plan and damage class of rd floor (a) Damage class V (b) Damage class IV Photo. Damage classes of R/C columns
Table 1. Column cross-section Table. Damage rating based on R index Longitudinal (NS) direction Transverse (EW) direction 1F & F F to 5F Floor R value Damage rating R value Damage rating Size (mm) 55 55 45 45 1 98.8 Slight 94.5 Light Longitudinal 85.9 Light 89.4 Light 16-D19 1-D19 reinforcement Shear reinforcement Unknown 85. Light 54.1 Heavy 4 1. No damage 96.7 Slight 5 98.6 Slight 99. Slight Vibration Characteristic of BPKP Building In this chapter, predominant periods and mode shapes of BPKP building after the earthquake are computed from the ambient vibration measurement data using micro tremor. The measurement cases and computed results are described in the following sections. Measurement Case Micro tremor data-logger (GEODAS-1-4DS by Buttan-Service Corporation in Japan) and pick-up equipment (velocimeter CR4.5- by Buttan-Service Corporation) are used for measuring ambient vibrations. Sampling frequency is set for 1Hz and record duration is seconds. The measurement positions of two cases used for calculating periods and mode shapes among all the 7 cases are shown in Fig.. In Case 1 and Case, the velocimeters are installed in the vertical directions of row D-5 and row D-, respectively, as shown in Fig.. H/H spectrum ratio 1 4 5 6 7 8 Frequency (Hz) 5 1st period : 1.14sec. 5 nd period :.sec. nd period :.sec. 1 4 5 6 7 8 Frequency (Hz) Figure. Two measurement cases Figure. H/H spectrum ratios of 5th floor to 1st floor H/H spectrum ratio 15 1 5 1 75 Longitudinal dir. Case 1 1st period :.89sec. Case 1st period :.6sec. nd period :.4sec. Transverse dir. 1st period : 1.14sec. nd period :.17sec. Case 1 Case
Measurement Results Predominant Periods of BPKP Building after the Earthquake Fourier spectrum ratios (H/H Spectra) of fifth floor to first floor in both directions of Case 1 and Case are shown in Fig.. As can be found in the figure, the predominant periods of the longitudinal (NS) direction of the building after the earthquake are about.6 second in Case 1 and.89 second in Case, respectively, while those of the transverse (EW) direction have approximately the same value of 1.14 second in both cases. The predominant periods of the transverse direction is larger than those of the longitudinal direction, and this result corresponds well with the actual damage condition of the building. Mode shapes in Vertical Plane after the Earthquake The first and the second mode shapes in vertical plane of the building after the earthquake are shown in Fig. 4. As shown in the figure, this building has almost the same mode shapes for each predominant period regardless of measurement positions and directions. In particular, second mode shapes are consistent well with the actual damage condition that the damage to column capital of the third story is the largest. Floor 5 4 1 Longitudinal dir. Transverse dir. Longitudinal dir. Transverse dir. First mode Second mode (a) Case 1 (row D-5) (b) Case (row D-) Figure 4. First and second mode shapes in vertical plane Seismic Response Estimation of BPKP Building In this chapter, the experienced maximum drift of the transverse (EW) direction in BPKP building which earthquake damage is relatively large is analytically estimated by the predominant period and the residual seismic capacity ration index R, and the estimated damage classes of each column obtained by two procedures are compared with the actually surveyed values to evaluate the compatibility of their procedures. Pushover Analysis The pushover analysis is carried out for the transverse (EW) direction of BPKP building. The assumptions on the analysis are as follows. (1) The lateral force distribution is assumed as the elastic first mode shape. () The foundation, the slab, and the beam-column joint are considered as fix, rigid diaphragm,
and rigid zone, respectively. () R/C columns are modelled as having the nonlinear flexural springs to their both ends, and having the elastic shear spring and the elastic axial spring to their center, respectively. The Takeda model [] is employed for the nonlinear flexural spring. (4) Since the damages to beams are slight, they are assumed to be elastic. (5) The URM walls located in frames are disregarded. The relation of story shear force and inter-story drift angle obtained from the pushover analysis based on the assumptions above is shown in Fig. 5 together with the distribution of inter-story drift angle of each floor. As shown in the figure, the building drifts concentrate on the third floor. This result is consistent well with the damage states shown in Fig. 1 and Table. Story shear force (*1 ) 9 8 7 6 5 4 1..5 1. 1.5..5. Inter-story drift angle (%)..5 1. 1.5..5. Inter-story drift angle (%) Figure 5. Shear force vs. drift angle and distribution of inter-story drift angle of each floor Seismic Response Estimation Method based on Predominant Period 1F 4F 5F F Floor 5 4 1 In this section, the experienced maximum drift is estimated by the predominant period obtained by the ambient vibration measurements, and the estimated damage classes of each column are compared with the actually surveyed values to evaluate the compatibility of this procedure. Proposed Procedure In order to estimate the maximum drift experienced during the earthquake, the following procedure shown in Fig. 6 is applied. (1) Pushover analysis for overall frames of BPKP building is carried out for each step j. () Unloading is performed at each step j, and each story stiffness K u i (i: the number of story) at unloaded state is calculated. () Overall frames are replaced by multi-degrees-of-freedom model having K u i obtained in (), and the first natural period T j is then calculated using modal analysis. (4) Iterative calculation of (1) through () is carried out until T j is equal to T T which is the target natural period (In this paper, T T =1.14 roughly for the transverse direction of the building, refer to Fig. ). (5) The experienced maximum drift of this building is determined at T j =T T. F
Figure 6. Experienced maximum drift estimation procedure Procedure Results The relationship between story shear coefficient and inter-story drift angle of the third story is shown in Fig. 7 together with the development of predominant period with respect to the its inter-story drift. As shown in the figure, the drift angle of the third story when becoming the same as predominant period obtained by ambient vibration measurements (1.14sec., refer to Fig. ) is approximately 1.%, and the shear coefficient value of the third story at its drift angle is about.. Shear coefficient Predominant period (sec.).4....1..8 1. 1. 1.14sec. 1.% 1.4..5 1. 1.5..5. Drift angle of rd story (%) Figure 7. Drift angle and shear coefficient of rd story at predominant period In this paper, the relationship between the load-deformation curve and the damage class for ductile member is defined as shown in Fig. 8 by varying axial forces [], [4]. Fig. 9 shows the damage classes of each column estimated by the proposed procedure together with the surveyed damage classes shown in parentheses. As shown in the figure, the estimated damage classes are relatively larger than the surveyed. As shown in Fig.7, the increase of predominant period to drift angle tends to become dull rapidly bordering on near 1.% of drift angle. Therefore, this procedure is suitable for brittle failure building rather than ductile failure one. Seismic Response Estimation Method based on Residual Seismic Capacity Ratio Index R In this section, the experienced maximum drift is estimated by residual seismic capacity ratio index R calculated by damage classes of columns (refer to Appendix and [5]), and the estimated damage classes are compared with the actually surveyed values.
(a) Row 5 (b) Row 4 (c) Row 1 Figure 8. Definition of damage class of column 6, 6, 6, 6, Figure 9. Estimation of damage class of each column in the third floor Proposed Procedure In order to estimate the maximum drift experienced during the earthquake, the following procedure is applied. (1) Pushover analysis for overall frames of BPKP building is carried out for each step j. () Index R j at each step j based on damage class is calculated. () Iterative calculation of (1) and () is conducted until R j is equal to R T (54.1%, refer to Table ). (4) The experienced maximum drift of this building is determined at R j =R T. Procedure Results The relationship between story shear coefficient and inter-story drift angle of the third story is shown in Fig. 1 together with the development of index R with respect to the its inter-story drift. As shown in the figure, the drift angle of the third story at the target index R (54.1%, refer to Table ) is approximately 1.5%, and the shear coefficient value of the third story at its drift angle is about.95. Fig. 11 shows the damage classes of each column estimated by the proposed procedure together with the surveyed damage classes shown in parentheses. As shown in Fig. 9 and 11, the results by this procedure are better than those by former procedure. The index R sensitively decreases
to overall drift angle as shown in Fig. 1, and this result cause that factor η tends to gradually decrease to increment of drift angle, although factor η for evaluating R value is discontinuous shown in Table 5 of Appendix. Therefore, this procedure can both use brittle and ductile failure building. Shear coefficient R-index (%).4...1. 8 6 4.95 54.1% 1.5%..5 1. 1.5..5. Drift angle of rd story (%) Figure 1. Drift angle and shear coefficient of rd story at target index R 6, 6, 6, 6, Figure 11. Estimation of damage class of each column in the third floor Conclusions In this paper, the seismic responses represented by experienced maximum drift of BPKP building during the earthquake are analytically estimated with the predominant period and the residual seismic capacity ratio index R. The results can be summarized as follows. (1) In BPKP building, the residual seismic capacity ratio index R value for the transverse (EW) direction of the third floor which suffered the most serious damage is evaluated to be 54.1%, and the damage rating is judged as Heavy. () The predominant periods of the longitudinal (NS) direction are about.6 and.89 seconds in Case 1 and Case, respectively, while those of the transverse direction are approximately 1.14second in both cases. The predominant period of the transverse direction which the damage is relatively severe is larger than that of the longitudinal direction, and this result
corresponds well with the actual damage condition. () From the pushover analysis, the building drifts concentrate on the third floor, and this result is consistent well with the damage states. (4) The experienced maximum drift is estimated by the predominant period obtained by the ambient vibration measurements. The drift angle of the third story at target predominant period is about 1.%, and the estimated damage classes at this drift angle are relatively larger than the surveyed. Since the increase of predominant period to drift angle tends to become dull rapidly bordering on near 1.% of drift angle, this procedure is suitable for brittle failure building. (5) The experienced maximum drift is estimated by residual seismic capacity ratio index R calculated by damage classes of columns. The drift angle of the third story at the target index R is roughly 1.5%, and index R sensitively decreases to overall drift angle. Therefore, this procedure can both use brittle and ductile failure building. Appendix Simplified Residual Seismic Capacity Ratio Index R [5] A damage classification of R/C columns and walls is performed based on the damage definition shown in Table. As defined in the Table, columns and walls are classified in one of five categories I through V. Table. Damage class definition of RC columns and walls Damage class I II III IV V Description of damage - Visible narrow cracks on concrete surface (Crack width is less than.mm) - Visible clear cracks on concrete surface (Crack width is about. -1.mm) - Local crush of cover conc. - Remarkable wide cracks (Crack width is 1. -.mm) - Remarkable crush of concrete with exposed reinforcing bars - Spalling off of concrete cover (Crack width is more than.mm) - Buckling of reinforcing bars - Cracks in core concrete - Visible vertical and/or lateral deformation in columns and/or walls - Visible settlement and/or leaning of the building Considering the normalized strength index C shown in Table 4 and the seismic capacity reduction factor η listed in Table 5 for a damaged member, the simplified residual seismic capacity index R can be expressed as shown in Eq. 1. 5 Aj j= R = 1 (%) (1) Aorg where, A = S + M + W + CW + 6CWC, A 1 =.95S1 +.95M1 +.95W 1 + 1.9CW1 + 5. 7CWC1 A =.6S +.75M +.6W + 1.CW +. 6CWC, A =.S +.5M +.W +.6CW + 1. 8CWC A 4 =. 1M 4, A = 5, Aorg = Ssum + M sum + Wsum + CWsum + 6CWCsum A, A 1, A, A, A 4, A 5, A org : Sum of normalized residual seismic capacity of members having damage class through V and normalized seismic capacity of a building in pre-damaged condition, respectively S, S 1, S, S, S 4, S 5, S sum : Number of brittle columns having damage class through V and their total number, respectively M, M 1, M, M, M 4, M 5, M sum : Number of ductile columns having damage class through V and their total number, respectively
W, W 1, W, W, W 4, W 5, W sum : Number of walls without boundary columns having damage class through V and their total numbers, respectively CW, CW 1, CW, CW, CW 4, CW 5, CW sum : Number of columns with wing wall(s) having damage class through V and their total number, respectively CWC, CWC 1, CWC, CWC, CWC 4, CWC 5, CWC sum : Number of walls with boundary columns having damage class through V and their total number, respectively Section Table 4. Normalized strength index C for simplified procedure Ductile / Brittle column Wall without boundary columns Column with wing wall(s) Wall with boundary columns τ u (N/mm ) 1 1 C 1 1 6 Damage Class Brittle column 6cm 6cm Table 5. Seismic capacity reduction factor η Ductile column 4cm 15cm Wall without boundary columns Colum n with wing wall(s) Wall with boundary columns I.95.95.95.95.95 II.6.75.6.6.6 III..5... IV.1 V The residual seismic capacity ratio index R defined above can be considered to represent damage sustained by a building. For example, it may represent no damage when R = 1 % (1 % capacity is preserved), more serious damage with decrease in R, and total collapse when R = % (no residual capacity). The guideline defines the damage rating criteria shown below. [Slight] 95(%) R [Light] 8(%) R <95(%) [Moderate] 6(%) R <8(%) [Heavy] R <6(%) [Collapse] Building which is deemed to have R due to collapse References 1. The Japan Building Disaster Prevention Association (JBDPA). (1991, revised in 1). Guideline for Postearthquake Damage Evaluation and Rehabilitation.. Takeda, T., Sozen, A. and Nielsen, N.M. (197). Reinforced Concrete Response to Simulated Earthquakes. Journal of Structural Division, ASCE. Vol.96:No.ST1, 557-57.. Architectural Institute of Japan (AIJ). (1999). Design Guidulines for Earthquake Resistant Reinforced Concrete Buildings Based on Inelastic Displacement Concept. 4. Uchida, T., Ohara M. and Maeda, M. (5). Post-earthquake Capacity Evaluation of R/C Buildings Composed of Columns with Different Ductility. Summries of Technical Paper of Annual Meeting, Architectural Institute of Japan. Vol.C-, 7-74. 5. Nakano, Y., Maeda, M., Kuramoto, H. and Murakami, M. (4). Guideline for Post-Earthquake Damage Evaluation and Rehabilitation of RC Buildings in Japan. 1th World Conference on Earthquake Engineering, Paper No.14. 4cm 15cm 48cm 15cm