Synthetic Near-Field Rock Motions in the New Madrid Seismic Zone

Transcription

1 Synthetic Near-Field Rock Motions in the New Madrid Seismic Zone Genda Chen*, Ph.D., P.E., and Mostafa El-Engebawy Engebawy,, Ph.D. *Associate Professor of Civil Engineering Department of Civil, Architecture and Environmental Engineering University of Missouri-Rolla Geotechnical and Bridge Seismic Design Workshop New Madrid Seismic Zone Experience October 28-29, 29, 2004 Rock Motions - 1 Participants Genda Chen, Ph.D., P.E. (team leader) Mostafa El-Engebawy Engebawy,, Ph.D. Yuehua Zeng,, Ph.D. (seismologist) David Hoffman (geologist) David Rogers, Ph.D., P.E. (geologist) Robert Hermann, Ph.D. (seismologist) Rock Motions - 2 1

2 Outline of Presentation Objectives Overview of Study Area Characteristics of Near-Field Motions Generation of Synthetic Near-Field Rock Motions Discussion of Results Simulated vs. AASHTO & NCHRP Spectra Comparison with other Simulation Methods Concluding Remarks Rock Motions - 3 Objectives To provide rock motion time histories at three bridge sites within the NMSZ for various combinations of moment magnitude and fault mechanism To evaluate near-field characteristics in the NMSZ To compare the spectra of the simulated motions with those of the AASHTO and the NCHRP project To compare the results of the composite-source source method with those of the finite-fault fault and the point- source models Rock Motions - 4 2

3 Overview of Study Area 39 N -92 W Longitude -91 W -90 W -89 W -88 W Latitude 38 N 37 N Southwestern segment Reelfoot fault Bootheel lineament L472 bridge site A1466 bridge site IS55 bridge site events Memphis & St. Louis 36 N 35 N Rock Motions - 5 Characteristics of Near-Field Motions Forward Directivity: rupture towards the site and is characterized by a two-sided velocity pulse(s) in the fault- normal direction Fling Step: characterized by one-sided velocity pulse in the same direction as the slip on the fault Rock Motions - 6 3

4 1992 Landers earthquake in Southern California (Strike-Slip Slip Earthquake) Rock Motions Landers earthquake - Lucerne Records Fault-normal: double-sided velocity pulse; small permanent displacement Fault-parallel: single-sided sided velocity pulse; large permanent displacements Rock Motions - 8 4

5 Effects of Forward Rupture Directivity Increase the amplitude of intermediate and long period ground motion Fault-normal component is larger than fault- parallel component at intermediate and long periods after Somerville et al. (1997) Rock Motions - 9 Parameters of Forward Rupture Directivity after Somerville et al. (1997) Directivity parameters: Rock Motions - 10 X cos θ for strike-slip faults Y cos Ф for dip-slip faults 5

6 Generation of Synthetic Near-Field Ground Motions in the NMSZ Earthquake Processes & Key Parameters Bridge 1) Earthquake Source Mechanism and Generation of Seismic Waves Fault mechanism (strike, dip & rake) Rupture area / Moment Magnitude M W Depth to top of fault Hypocenter location Rupture velocity Slip distribution Stress drop Fault Plane High slip zones Hypocenter 3) Site Response Soil depth Soil linear/nonlinear properties Shear wave velocity & damping 2) Wave Propagation Velocity and density profile of the earth crust Information on geometrical spreading, inelastic energy absorption, reflection, refraction & wave scattering through the use of analytical Green s function Rock Motions - 11 Seismic Source Parameters Best-estimates estimates & uncertainties Rupture Area log A = M W s = 22% (26%) s is the standard deviation for strike-slip slip (reverse) faults after Wells & Coppersmith (1994) Fault Southwestern segment (strike-slip slip fault) Reelfoot fault (reverse fault) Best-estimate estimate mechanism Strike = dip = 90 rake = 180 Strike = dip = 32 rake = 90 Best-estimate estimate rupture area L = 120 km, W = 18 km L = 56 km, W = 13.6 km L = 27 km, W = 10 km L = 82 km, W = 28 km L = 44 km, W = 18 km L = 22 km, W = 11 km for M 7.5, W for M 7.0, W for M 6.5 W for M 7.5, W for M 7.0, W for M 6.5 W Rock Motions

7 Seismic Source Parameters Best-estimates estimates & uncertainties Depth to top of the fault 1 km or 5 km Rake angle of slip on fault 150, 180 or -150 º Stress drop 100, 150 or 200 bars Rupture velocity 80% or 85% of V S Rock Motions - 13 Wave Propagation Parameters Best-estimates estimates & uncertainties 0 Velocity (km/sec) Velocity model of the earth crust Chiu et al. (1992) 20% variation Depth from ground (km) Rock Motions - 14 P-wave S-wave 7

8 Logic Tree of Uncertain Parameters Increase fault length by σ Best-estimate rupture area Increase fault width by σ Depth to top of fault 1 km Weight 1/2 Depth to top of fault 5 km Weight 1/2 Hypocenter location along strike and dip Equally-Distributed Rupture velocity = 80% shear wave velocity Weight 1/2 Rupture velocity = 85% shear wave velocity Weight 1/2 Reference rake angle 30º Reference rake angle Reference rake angle + 30º 20% decrease in velocity USGS velocity model 20% increase in velocity Stress drop = 100 bars Stress drop = 150 bars Stress drop = 200 bars Rock Motions - 15 The Composite Source Model Earthquake Source The source of a strong earthquake is taken as a superposition of the radiation from a significant number of circular subevents with a constant stress drop. Rupture initiates at the presumed hypocenter and propagates radially at a constant rupture velocity. Each subevent is triggered when the rupture front reaches the center of the subevent. The subevent then initiates the radiation of a displacement pulse. Rock Motions

9 The Composite Source Model Wave Propagation & Wave Scattering The generated displacement pulse propagates through a flat multi-layered layered earth crust. The wave propagation process is modeled with synthetic (analytical) Green s s functions in both short- and long-period ranges. The short-period component is modified to account for the effects of random lateral heterogeneity of the earth by adding scattered waves. Rock Motions - 17 The Composite Source Model Validation re 5 Obser ed (red) s S nthetic (bl e) Gro nd Accelerations Velocit Observed (red) vs. synthetic (blue) ground motions at station SKR R (east horizontal component) during 1999 Kocaeli earthquake (strike-slip) slip) Rock Motions

10 Observed (red) vs. synthetic (blue) ground motions at station T052 (NS component) during 1999 Chi-Chi earthquake (reverse) Rock Motions - 19 Discussion of Results of the Maximum Considered Earthquake (MCE) or M W 7.5 Rock Motions

11 Total Uncertainty Southwestern segment 1 +1 L472 site - FN 1 +1 A1466 site - FN Average spectrum Average spectrum Average + 1 σ 10-2 Average + 1 σ Rock Motions - 21 Total Uncertainty Reelfoot fault 1 +1 IS55 site - FN Average spectrum Average + 1 σ A1466 site - FN Average spectrum Average + 1 σ Rock Motions

12 Average Response Spectra Southwestern segment L472 site SW - Mw 7.5 FP FN V A1466 site SW - Mw 7.5 FP FN V Rock Motions - 23 Average Response Spectra Reelfoot fault 4.0 IS55 site RF - Mw 7.5 FP FN V A1466 site RF - Mw 7.5 FP FN V Rock Motions

13 Influence of Fault Mechanism on the Fling Step at L472 site Displacement (m) Basic analysis Fault strike 231.5º Fault strike 236.5º Rock Motions - 25 Fault strike Fault strike Influence of Fault Mechanism on the Fling Step at L472 site Displacement (m) Basic analysis Fault dip 70º Displacement (m) Basic analysis Rake angle 150º Rake angle -150º For basic analysis: Fault dip 90 Rake angle 180 Rock Motions

14 Influence of Depth to top of Fault and Stress Drop on the Fling Step at L472 site Displacement (m) Basic analysis Depth 3km Depth 5km Displacement (m) Basic analysis Stress drop 100 bars Stress drop 175 bars For basic analysis: Depth 1km Stress drop 150 bars Rock Motions - 27 Influence of Hypocenter Location on Peak Rock Velocity at L472 Site Location of L472 Site Directivity Parameter (X COS θ) at that subfault Peak rock velocity if the hypocenter is located at the center of this subfault Rock Motions

15 Influence of Rupture Velocity on Velocity Pulses at L472 Site Velocity (m/sec) Vr = 75% Vs Vr = 85% Vs Rock Motions Validation of Synthetic Rock Motions Comparison with Attenuation Relations Southwestern segment L472 site - Mw 7.5 This study Campbell 2002 Toro 1997 Frankel 1996 A1466 site - Mw 7.5 This study Campbell 2002 Toro 1997 Frankel Rock Motions

16 4.0 Validation of Synthetic Rock Motions Comparison with Attenuation Relations IS55 site - Mw 7.5 This study Campbell 2002 Toro 1997 Frankel 1996 Reelfoot fault A1466 site - Mw 7.5 This study Campbell 2002 Toro 1997 Frankel Rock Motions - 31 Validation of Synthetic Rock Motions Comparison with NCHRP & AASHTO Guidelines Southwestern segment L472 site - Mw 7.5 Fault-parallel Fault-normal ATC/MCEER AASHTO A1466 site - Mw 7.5 Fault-parallel Fault-normal ATC/MCEER AASHTO Rock Motions

17 Validation of Synthetic Rock Motions Comparison with NCHRP & AASHTO Guidelines 4.0 IS55 site - Mw 7.5 Fault-parallel Fault-normal ATC/MCEER AASHTO Reelfoot fault A1466 site - Mw 7.5 Fault-parallel Fault-normal ATC/MCEER AASHTO Rock Motions - 33 Validation of Synthetic Rock Motions Comparison with Finite-Fault Fault & Point-Source Models Southwestern segment L472 site - Mw 7.5 This study Finite-fault Point-source A1466 site - Mw 7.5 This study Finite-fault Point-source Rock Motions

18 Validation of Synthetic Rock Motions Comparison with Finite-Fault Fault & Point-Source Models Reelfoot fault 4.0 IS55 site - Mw 7.5 This study Finite-fault Point-source A1466 site - Mw 7.5 This study Finite-fault Point-source Rock Motions - 35 Near-Field Characteristics of the Selected Motions Selection criteria of rock motions 1) Fit the average response spectra 2) Fling step in the direction of the slip on the fault 3) Velocity pulse in the fault-normal direction 4) Realistic peak rock accelerations (within 75%-125% of Toro et al., 1997) Rock Motions

19 Near-Field Characteristics of the Selected Motions Fling step from the southwestern segment Displacement (m) Mw 6.5 Mw 7.0 Mw 7.5 L472 site - FP Displacement (m) Mw 6.5 Mw 7.0 Mw 7.5 A1466 site - FP Rock Motions - 37 Near-Field Characteristics of the Selected Motions Fling step from the Reelfoot fault Displacement (m) Mw 6.5 Mw 7.0 Mw 7.5 IS55 site - FN Displacement (m) Mw 6.5 Mw 7.0 Mw 7.5 IS55 site - V Rock Motions

20 Near-Field Characteristics of the Selected Motions Velocity pulse from the southwestern segment Velocity (m/sec) - - L472 site - FN Mw 6.5 Mw 7.0 Mw Velocity (m/sec) - - A1466 site - FN Mw 6.5 Mw 7.0 Mw Rock Motions - 39 Near-Field Characteristics of the Selected Motions Velocity pulse from the Reelfoot fault Velocity (m/sec) IS55 site - FN Mw Velocity (m/sec) IS55 site - FN Mw 6.5 Mw Rock Motions

21 St. Francis River Site (Far-Field) Field) 39 N -92 W Longitude -91 W -90 W -89 W -88 W A3708 site is about 50km from the Reelfoot fault and 87km from the southwestern segment 38 N 37 N Latitude 36 N A N Rock Motions - 41 Far-Field Field Rock Motions Comparison with NCHRP & AASHTO Guidelines FP component FN component V component ATC/MCEER AASHTO FP component FN component V component ATC/MCEER AASHTO Southwestern segment Rock Motions - 42 Reelfoot fault Average of 20 simulations 21

22 Concluding Remarks The uncertainty of near-fault motions increases with moment magnitude and decreases with distance to fault The southwestern segment (strike-slip) slip) contributes more to the total uncertainty than the Reelfoot fault (reverse) due to its forward rupture directivity effects The vertical component associated with the Reelfoot fault is stronger than that of the southwestern segment Fling step is dependent on the fault mechanism (strike, dip and rake), depth to top of the fault and stress drop Rock Motions - 43 Concluding Remarks Velocity pulses are dependent on the hypocenter location along the strike and rupture velocity The simulated spectral accelerations are higher than those of the attenuation relations, point-source or finite- fault models due to forward rupture directivity effects, particularly for M W 7.5 for strike-slip slip faults Velocity pulses associated with M W 7.5 are very large as compared to M W 7.0 or 6.5 that may impose special seismic demands for structures very close to active faults Rock Motions

23 Concluding Remarks In comparison with ATC/MCEER spectra, the near- field motions in the proximity of the faults (<5 km) are generally higher, and those around 10km are similar in long period components but smaller in short period components. The far-field field rock motion is on the average less than what ATC/MCEER specified in their recommended guidelines. Rock Motions