Development of acceleration time histories for Semarang, Indonesia, due to shallow crustal fault earthquakes Windu Partono, Masyhur Irsyam, and Sri Prabandiyani Retno Wardani Citation: AIP Conference Proceedings 1903, 090004 (2017); View online: https://doi.org/10.1063/1.5011607 View Table of Contents: http://aip.scitation.org/toc/apc/1903/1 Published by the American Institute of Physics Articles you may be interested in Sensitivity analysis of tall buildings in Semarang, Indonesia due to fault earthquakes with maximum 7 Mw AIP Conference Proceedings 1903, 020008 (2017); 10.1063/1.5011488 Preface: The 3rd International Conference on Construction and Building Engineering (ICONBUILD 2017) AIP Conference Proceedings 1903, 010001 (2017); 10.1063/1.5011479
Development of Acceleration Time Histories for Semarang, Indonesia, Due to Shallow Crustal Fault Earthquakes Windu Partono 1, a), Masyhur Irsyam 2, b) 1, c) and Sri Prabandiyani Retno Wardani 1 Civil Engineering Department, Diponegoro University, 50275 Semarang, Indonesia 2 Civil Engineering Department, Bandung Institute of Technology, 40132 Bandung, Indonesia. a) Corresponding author: windu_bapake_dila@yahoo.com b) masyhur.irsyam@yahoo.co.id c) wardani_spr@yahoo.com Abstract. Research on seismic, microzonation of Semarang is still ongoing. Following the research conducted by Team for Revision of Hazard Maps of Indonesia 2010, Lasem fault was the only fault that should be taken into account for seismic mitigation of Semarang. New research conducted by Team for Updating of Seismic Hazard Maps of Indonesia 2016 suggesting four new and closest shallow crustal fault sources (Rawapening, Weleri, Demak and Semarang Faults) which should be taken into account for seismic hazard mitigation of this city. Those four new seismic sources are typical reverse mechanism seismic sources. However Lasem fault is a typical strike slip mechanism seismic source. This paper presents the development of surface acceleration time histories due to three shallow crustal fault (Lasem, Semarang and Demak) earthquake sources with average magnitude 6.5 Mw. This research was performed by implementing deaggregation hazard analysis, response spectral matching and site response analysis to obtain modified acceleration time histories. The modified acceleration time histories were developed due to inadequate data caused by those three fault sources. Surface acceleration time histories were calculated at 288 boring locations and then separated into three different time histories based on site class soil conditions (hard, medium and soft soil classes). INTRODUCTION Research on seismic, microzonation of Semarang is still ongoing. Following the research conducted by Team for Revision of Seismic Hazard Maps of Indonesia 2010, Lasem Fault is one of the most interesting seismic sources which should be taken into account for seismic mitigation of this city. Based on final research conducting by Team for Updating of Seismic Hazard Maps of Indonesia 2016 there are another 4 (four) shallow crustal fault sources (Rawapening, Semarang, Demak and Welerei faults) which take the advantage for seismic mitigation for this city. Figure 1 shows the trace line of all faults located surrounding and crossing the city. Semarang, Demak and Lasem Faults are three closest shallow crustal which will take the important earthquake sources for this area. Seismic, microzonation for Semarang has already been performed on 2015 by conducting Lasem Fault as an active earthquake source for seismic hazard analysis [1]. Site response analysis for seismic microzonation was performed using 17 different time histories from five strike-slip mechanism shallow crustal faults earthquake with average magnitude 6.5 Mw. Site response analysis was performed at 190 soil boring locations with maximum 20 km distance from Lasem fault trace. This paper presents the development of surface acceleration time histories due to three shallow crustal faults Semarang, Lasem and Demak faults. Surface acceleration time histories were developed and calculated at 288 boring locations and conducting Semarang and Demak faults (reverse fault mechanism earthquake sources) and Lasem fault (strike slip mechanism earthquake source) as the main and dangerous seismic source for this study are. Figure 1, 1(c) and (1(d) shows distribution of all soil investigations and its position against Semarang fault trace, Lasem fault trace and Demak fault trace. Most of all soil boring locations (249 points) are distributed less than 5 Km, 32 locations are distributed in between 5 to 10 Km and the rest 7 locations are distributed more than 10 Km from Semarang Fault trace. Surface time histories, acceleration time histories, were Proceedings of the 3rd International Conference on Construction and Building Engineering (ICONBUILD) 2017 AIP Conf. Proc. 1903, 090004-1 090004-9; https://doi.org/10.1063/1.5011607 Published by AIP Publishing. 978-0-7354-1591-1/$30.00 090004-1
developed following the same method used by [1 and 2]. Surface time histories developed form this study can be used for dynamic structure design and evaluation and also as part of the second seismic microzonation scenario for this city. (c) FIGURE 1. Fault traces surrounding Semarang, Soil boring locations and its distance against Semarang fault trace, Lasem fault trace (c) and Demak fault trace (d) Due to inadequate accelerations time histories data produced by Semarang, Lasem and Demak fault earthquakes, the modified acceleration time histories were developed using another similar seismicity reverse and strike slip mechanism fault earthquake data. The time histories for reverse and strike slip fault earthquake were collected from worldwide historical earthquake databases using three equal parameters such as magnitude, distance and seismic mechanism. The modified time histories, from another earthquake or past earthquakes, needs to match with earthquake scenario produced by those three fault earthquake sources. Response spectral matching analysis between specific time histories (from worldwide earthquake data) and spectral acceleration (predicted spectral target from Semarang, Lasem and Demak fault earthquake scenarios) should be performed to get modified time histories. The development of surface time histories was performed following the procedures: (d) Conducting geological, geophysical and soil investigation and analysis for developing soil profile, bedrock elevation and site class maps of Semarang. 090004-2
Conducting deterministic seismic hazard analysis and seismic hazard de-aggregation for estimating magnitude and distance for shallow crustal fault earthquake and developing spectral target acceleration. Collecting acceleration time histories from worldwide historical earthquake records due to shallow crustal fault sources with magnitude 6.5 Mw and maximum distance 30 km. Developing modified acceleration time histories by conducting response spectral matching analysis. Conducting shear wave propagation analysis using modified time histories calculated from response spectral matching analysis for obtaining surface acceleration time histories. GEOLOGICAL AND GEOTECHNICAL CONDITIONS Depth of engineering bedrock is one of the important parameters used to perform site response analysis. Identification of bedrock elevation for the study area was performed using single station feedback Microtremor and implementing three component ambient vibrations [2, 3 and 4]. Bedrock elevation measurement was performed at 218 locations. Using three component ambient vibrations, the depth of engineering bedrock can be predicted using two empirical formulas proposed by [5] and [6]. Figure 2 shows the distribution of bedrock depth of the study area. As can be seen in this figure the depth of engineering bedrock or the thickness of soil deposit is increase from Southern part to the Northern part of the study area. Site characterization (classification) of geotechnical data was carried out by interpreting the results of field investigations such as in-situ standard penetration test (SPT) and laboratory tests. To develop site response analysis, 288 boring investigations with a minimum 30 m depth was performed in all part of Semarang city. The shear wave velocity profile was estimated using SPT-N data and calculated using three empirical equations proposed by [7, 8 and 9]. Figure 2 shows the distribution of Vs30 (average shear wave velocity at top 30 meter soil layer) calculated from 288 boring locations. FIGURE 2. Depth of Engineering Bedrock Identified by Single Station Feedback Seismometer and Distribution of Vs30 DEVELOPMENT OF ACCELERATION TIME HISTORIES Acceleration time-history for shallow crustal fault source was developed through three steps de-aggregation seismic hazard analysis, deterministic seismic hazard analysis and response spectral matching. De-aggregation seismic hazard analysis was performed to obtain controlling earthquake in terms of magnitude and distance which probably can gives maximum spectral acceleration for the whole area of the city. De-aggregation seismic hazard for Semarang was performed for shallow crustal fault with three different periods PGA, T=s and T=1s and following the procedure propose by [10 and 11]. Figure 3 shows one example de-aggregation results for PGA with return period 2500 years. It seems in this figure that earthquake with magnitude 6.5 Mw until 6.7 Mw and maximum distance 30 km controlling the earthquake of this area. Based on de-aggregation seismic hazard analysis results the scenario for shallow crustal fault earthquake was implemented using average magnitude 6.5 Mw and maximum distance 30 Km. Due to the limited earthquake 090004-3
records of Semarang, Lasem and Demak fault earthquake with magnitude 6.5 Mw, historical earthquake records with magnitude ranging from 6 to 7 Mw and maximum distance 30 km were then collected for shallow crustal fault. Table 1 shows 19 acceleration time histories from 5 different reverse mechanism earthquake events and Table 2 shows 16 acceleration time histories from 5 different strike slip mechanism earthquake events used in this study. All acceleration time histories data were collected from Pacific Earthquake Engineering Research (PEER) NGA-West 2 Databases. FIGURE 3. De-Aggregation PGA for Shallow Crustal Fault The target spectrum used for spectral matching analysis was predicted at bedrock elevation using deterministic hazard analysis by conducting three different attenuation functions proposed by [12, 13 and 14]. All target spectrums were calculated for reverse and strike slip mechanism shallow crustal fault sources. Figure 5 shows an example of 4 (four) target spectrums for Northridge-01 earthquake events (reverse mechanism) with magnitude 6.69 Mw. Figure 5 shows an example of 3 (three) target spectrum for Superstation Hills earthquake events (strike slip mechanism) with magnitude 6.54 Mw. The matched (modified) acceleration time histories were predicted using response spectral matching analysis and following the same method proposed by [15]. The modified time histories was predicted using all acceleration time histories and should be matched with target spectrum. Initial time histories must have similar shape with matched time histories. Care must be taken into account to evaluate both time histories. Figure 6 shows an example result of spectral matching calculated for Northridge-01 earthquake and Figure 6 shows an example result of spectral matching for superstation Hills earthquake. Figure 7 shows initial and matched time histories for Northridge-01 earthquake and. Figure 7 shows initial and matched time histories results for Superstation Hills earthquake calculated from spectral matching analysis. 090004-4
TABLE 1. Reverse Mechanism Earthquake Sources Used in this Study Event Station M (Mw) R (km) Arleta - Nordhoff Fire Sta 6.05 1.48 Northridge-02 (1994) Newhall - Fire Sta 6.05 7.36 LA - Century City CC North 6.05 18.34 TCU084 6.2 3.68 Chi-Chi, Taiwan-03 (1999) TCU089 6.2 5.93 TCU076 6.2 13.04 TCU073 6.2 19.06 Arleta - Nordhoff Fire Sta 6.69 3.3 Northridge-01 (1994) Beverly Hills - 14145 Mulhol 6.69 9.44 LA - Brentwood VA Hospital 6.69 12.92 LA - Hollywood Stor FF 6.69 19.73 Nagaoka 6.8 3.97 Chuetsu-oki, Japan (2007) Kashiwazaki City Takayanagicho 6.8 10.38 Yan Sakuramachi City watershed 6.8 12.98 Yahiko Village Yahagi 6.8 19.73 IWTH24 6.9 3.1 Iwate, Japan (2008) IWT011 6.9 8.41 Kurihara City 6.9 12.83 AKTH06 6.9 19.15 TABLE 2. Strike Slip Mechanism Earthquake Sources Used in this Study Event Station M (Mw) R (km) El Centro Array #8 6.53 3.86 Imperial Valley (1979) Chihuahua 6.53 7.29 El Centro Array #11 6.53 12.56 CHY074 6.2 6.02 Chi-Chi, Taiwan-03 (1999) CHY080 6.2 12.44 CHY028 6.2 17.63 Parachute Test Site 6.54 0.95 Superstition Hills (1987) Superstition Mtn Camera 6.54 5.61 Westmorland Fire Sta 6.54 13.03 Kobe University 6.9 0.9 Kobe Japan (1995) Nishi-Akashi 6.9 7.08 Amagasaki 6.9 11.34 Fukushima 6.9 17.85 Victoria Hospital Sotano 6.33 6.07 Victoria Mexico (1980) Cerro Prieto 6.33 13.8 Chihuahua 6.33 18.53 090004-5
Spectra Target for Northridge-01 Earthquake 6.69 Mw Spectra Target for Superstation Hills Earthquake 6.54 Mw 1.4 1.2 1.2 1 0.8 0.6 3.3 Km 9.44 Km 12.92 19.73 1 0.8 0.6 0.95 Km 5.61 Km 13.03 0 0 0.6 0.8 1 Period (sec) 0 0 0.6 0.8 1 Period (sec) FIGURE 4. Target spectrum used for spectral matching analysis, Spectral matching result for Northridege-01 earthquake FIGURE 5. Matching spectra output for northridge-01 NS earthquake with 3.3 Km epicenter distance and matching spectra output for superstation hills NS earthquake with 0.95 Km epicenter distance 090004-6
SITE RESPONSE ANALYSIS Site response analysis was performed at 288 positions of soil boring for obtaining surface time histories. Site response analysis was performed using matched time histories calculated from spectral matching analysis. Due to inadequate data related with soil properties from end boring level until bedrock elevation, site response analysis was performed using soil deposit model proposed by [2]. Figure 9 shows the corresponding soil deposit model for site response analysis. Site response analysis was implemented based on the assumption that all boundaries are horizontal and the response of a soil layer is predominantly caused by shear wave propagating vertically from the underlying bedrock. In this study the general response analysis was performed using equivalent linear approach and performed using the model proposed by [16 and 17] and utilizing free software NERA [18]. Acceleration Time Histories Northridge-01 NS 6.69 Mw 0.3 Initial Matched 0.1 0-0.1 - -0.3-0 5 10 15 20 25 30 35 40 Time (sec) Acceleration Time Histories Superstation Hills NS 6.54 Mw 0.0 - - -0.6 Initial Matched -0.8 0 5 10 15 20 25 Time (sec) FIGURE 6. Initial and Matched NS Time Histories for Northridge-01 Earthquake with Magnitude 6.69 Mw and Distance 3.3 Km ; Initial and Matched NS Time Histories for Superstation Hills 6.54 Mw and Distance 0.95 Km Figure 10 shows two time histories, initial time histories at bedrock elevation and surface time histories, calculated from site response analysis at Lok53 BH1. Site response analysis was performed using modified time histories from Northridge-01 earthquake with magnitude 6.69 Mw and 3.3 Km distance from Semarang Fault. Site response analysis was performed at medium (SC) soil class [19]. Surface peak ground acceleration (PGA), spectral s (T=s) and spectral 1s (T=1s) at point Lok53BH1 are 0.54g, 1.08g and 0.56g respectively. Figure 10 shows two time histories calculated from site response analysis at Lok97 BH1. Site response analysis was performed using modified time histories from Superstation Hills earthquake with magnitude 6.54 Mw and 0.95 Km distance from Lasem Fault. Site response analysis was performed at medium (SD) soil class [19]. Surface peak ground acceleration (PGA), spectral s (T=s) and spectral 1s (T=1s) at point Lok97BH1 are 4g, 1.64g and 0.39g respectively. Ground Surface Depth of borehole VS1 VS2 VSn Layer 1 Layer 2 Layer n End of Borehole elevation Soil Deposit below boring level Shear Wave Velocity Profile Bedrock Elevation with VS = 760m/s Bedrock (SB) FIGURE 7. Soil Profile Model for Site Response Analysis 090004-7
RESULT AND DISCUSSION Based on de-aggregation analysis for shallow crustal fault source conducted in this study identify controlling magnitude and distance of the earthquake sources for PGA, T=s and T=1s. Earthquake with magnitude inbetween 6-7 Mw and maximum distance 20 Km are controlling the seismic hazard for the study area. Spectral acceleration calculated at 288 boring locations using 19 time histories from 5 earthquake events (reverse fault mechanism) and 16 time histories from 5 earthquake events (strike slip mechanism) seems that the maximum value were identified within the closest area to Semarang Fault and Lasem fault trace. Due to the position of Semarang fault and Demak fault (two equal seismic mechanism sources) to the whole area of this city, Semarang fault earthquake will give a significant spectral acceleration to hazard mitigation to this city. Time Histories for Lok53 BH1 Northridge 01 6.69 Mw 0.6 0.5 Surface Bedrock 0.3 0.1 0-0.1 - -0.3-0 5 10 15 20 25 30 35 40 Time (sec) Time Histories for Lok97BH1 Superstation Hills Earthquake 6.54 Mw 0.5 Bedrock 0.3 Surface 0.1 0-0.1 - -0.3 - -0.5 0 5 10 15 20 25 Time (sec) FIGURE 8. Bedrock and surface time histories calculated at Lok53BH1 due to Northridge-01 earthquake 6.69Mw, Bedrock and surface time histories calculated at Lok97BH1 due to Superstation Hills earthquake 6.54Mw. Table 2 shows distribution of spectra accelerations (PGA, T=s and T=1s) for site class SC, SD and SE respectively calculated at 288 locations by implementing Semarang fault source earthquake scenario with magnitude form 6.05 Mw to 6.9 Mw. Site class SC, SD and SE was predicted using Vs30 values from 288 locations and following the same method proposed by [19]. Surface PGA for site SC is distributed between 7g until 0.77g ( g is represents gravitational acceleration). Spectra s for site class SC are distributed between 0.75g until 1.71g. However spectra 1s for site class SC are distributed in between 0.12g until 0.66g. PGA for site SD is distributed between 0.19g until 1.04g. Spectra s are distributed between 0.75g until 2.07g. Spectra 1s are distributed between 0.17g until 0.71g. Table 3 shows the distribution of surface PGA, spectra s and spectral 1s due to Lasem fault earthquake with magnitude from 6.2 Mw to 6.9 Mw. TABLE 3. Distribution spectral acceleration for site class SC, SD and SE due to Semarang fault earthquake PGA(g) T02s(g) T1s (g) SC SD SE SC SD SE SC SD SE Min 7 0.19 0.32 0.75 0.38 0.16 0.12 0.17 8 Max 0.77 1.04 1.09 1.71 2.07 8 0.66 0.71 0.76 Average 0.55 0.58 0.66 1.49 1.60 0.37 3 2 5 TABLE 4. Distribution spectral acceleration for site class SC, SD and SE due to Lasem fault earthquake PGA(g) T02s(g) T1s (g) SC SD SE SC SD SE SC SD SE Min 0.12 0.06 0.06 0.35 0.10 0.07 0.10 0.11 0.09 Max 0.33 0.34 7 0.81 0.87 0.58 0.13 6 0.35 Average 4 0.19 0.12 0.59 0.53 0.32 0.11 0.15 5 CONCLUSIONS Surface time histories for Semarang for structural design has already been implemented based on two major earthquake scenarios, Semarang fault earthquake and Lasem fault earthquake. Nineteen acceleration time histories were already implemented for Semarang fault earthquake sources and sixteen acceleration time histories for Lasem fault earthquake. All acceleration time histories were developed based on three major steps de-aggregation, response spectral matching and site response analysis. Based on average site response analysis results calculated at 288 boring 090004-8
locations Semarang fault earthquake source caused PGA, spectral Ts and spectral T1s greater than Lasem fault earthquake. For seismic mitigation of this city a total of 19 surface time histories must be taken into account for structural design and evaluation. REFERENCES 1. W. Partono, S.P.R. Wardani, and M. Irsyam, Jurnal Teknologi, 77:11, pp. 99-107, (2015). 2. W. Partono, S.P.R. Wardani and M. Irsyam (Jurnal Teknologi 78: 8-5, pp. 31-38, (2016). 3. Y. Nakamura (Quarterly Report of Railway Technical Research Institute 30:25-33, Tokyo Japan, 1989). 4. S.B. Claudet, S. Baize, L.F.Bonilla, C.B. Thierry, C. Pasten, J. Campos, Volant P. Volant and R. Verdugo Geophys. J. Int., 176, pp. 925-937, (2009). 5. I. Seht and J. Wohlenberg, Bulletin of the Seismological Society of America, 89, pp. 250-259, (1999). 6. S. Parolai, S., P. Bormann and C. Milkert, New Relationships Between VS, Bulletin of the Seismological Society of America 92, pp. 2521-2527, (2002). 7. Y. Osaki and R. Iwasaki, Soil and Foundations, JSSMFE, 13, pp. 59-73, (1973). 8. Y. Ohta and N. Goto, Earthquake Engineering and Structural Dynamics, 6, pp. 167-187, (1978). 9. T. Imai, and K. Tonouchi, Proc. of Second European Symposium on Penetration Testing, Amsterdam, The Netherlands, pp.67-72, 1982. 10. R.K. McGuire, Bulletin of the Seismological Society of America, 85, No. 5, pp 1275-1284, (1985). 11. S.L. Kramer, S. L., Prentice Hall, Upper Saddle River, New Jersey 07458, pp. xviii+653, (1996). 12. D.M. Boore and G.M. Atkinson, Earthquake Spectra, EERI, 24(1), pp. 99-138, (2008). 13. K.W. Campbell and Y. Bozorgnia, Earthquake Spectra, EERI, 24(1), pp. 139-171, (2008). 14. B.S.J. Chiou, and R.R. Youngs, Pacific Engineering Research Center, College of Engineering. University of California Berkeley, pp. xv + 95, (2008). 15. N.A. Abrahamson, Internal Report, 1998. 16. W.D. Iwan, Journal of Applied Mechanics, ASME, 34, pp. 612-617, (1967). 17. Z. Mroz, Journal of Mechanics and Physics of Solids, 15, pp. 163-175, (1967). 18. J.P. Bardet and T. Tobita, Department of Civil Engineering University of Southern California, pp. 44, (2001). 19. Institution of National Standardization, Procedure of Planning of Building Structure of Building and Non Building SNI 1726:2012, (BSN, ICS 91.125:91.080.01, Jakarta, 2012), pp. viii+138. 090004-9