Near-field and directionality issues for ground motion selection Jack W. Baker 35 minutes

J. Baker
1
Near-field and directionality issues for
ground motion selection
Jack W. Baker
35 minutes
J. Baker
2
Language from ASCE 7-10
16.1.3.2
At sites within 3 miles (5 km) of the active fault that controls the hazard,
each pair of components shall be rotated to the fault-normal and faultparallel directions of the causative fault and shall be scaled so that the
average of the fault-normal components is not less than the MCER
response spectrum for the period range from 0.2T to 1.5T.
J. Baker
3
Language from the BSSC proposal
16.2.3.3:
When the MCER ground motion level is controlled by events for which
near-fault effects are expected, the site shall be identified as a nearfault site and a suitable number of the ground motions shall include
near-fault and directivity effects including direction of fault rupture and
velocity pulses as appropriate.
16.2.5.1:
For sites identified as near-fault in Section 16.2.3.3, each pair of
horizontal ground motion components shall be rotated to the faultnormal and fault-parallel directions of the causative fault and applied to
the building in such orientation.
At all other sites, each pair of horizontal ground motion components
shall be applied to the building at arbitrary orientation angles.
J. Baker
4
When are you at a “near-fault” site?
• Two phenomena to consider when selecting time series:
– Directivity pulses
– Polarization of ground motions
J. Baker
• REDi:
• For sites located 20 km or closer to a fault capable of producing a
M6.5, the site- specific seismic hazard calculations should account
for near-fault directivity effects.
5
J. Baker
6
(slide with velocity pulses, NGA W-2 predictions?)
J. Baker
7
Observations from the 1979 Imperial Valley earthquake
The algorithm identifies ground motions with clear pulses, and the identified
motions are generally from locations where directivity is expected
J. Baker
8
Example classifications from past earthquakes
1979 Imperial Valley
1992 Landers
The algorithm identifies ground motions with clear pulses, and the identified motions are
generally from locations where directivity is expected. The algorithm has also been
extended to 3D.
J. Baker
9
How many pulses should you pick?
Shahi, S. K., and Baker, J. W. (2011). “An Empirically Calibrated Framework for
Including the Effects of Near-Fault Directivity in Probabilistic Seismic Hazard
Analysis.” Bulletin of the Seismological Society of America, 101(2), 742–755.
P( pulse | R, s ) =
Predictive model
1
1 + e0.642+ 0.167 R −0.075 s
Model predictions
J. Baker
10
How many pulses should you pick?
• Hayden, C., Bray, J., and Abrahamson, N. (2014). “Selection of
Near-Fault Pulse Motions.” Journal of Geotechnical and
Geoenvironmental Engineering, 140(7), 04014030.
• For example, if R=10 km for an Mw = 7 governing event and ɛ=1:0
for the 5% damped spectral acceleration at a period of 1s, the
proportion of ground motions that should be pulse motions is 0.40.
J. Baker
11
Orientation of ground motions
J. Baker
Orientation of SaRotD100 (using α as angle to strike parallel)
T* = 1.5 s
Site
α
Strike parallel
orientation
Fault rupture
•
Dependence of α on various
parameters was studied
•
A parametric model to predict the
distribution of α is proposed
T’ = 3 s
Oscillator responses to 1979 Imperial
Valley-06,
El Centro Differential Array recording
12
J. Baker
13
Dependence of α on M and Rrup
Rrup bins
J. Baker
Distribution of α for varying T, with Rrup between 0 and 5
km
Apparent division at 0.5 or 1 second
14
J. Baker
Why maintain fault-normal and fault-parallel orientations?
15
J. Baker
16
Conclusions
J. Baker
17
J. Baker
18
Extra
J. Baker
19
Measuring single-orientation (pseudo) response spectra
x(t )
J. Baker
20
Measuring direction-dependent response spectra
z (t )
=
z (t ) x(t )cosθ + y (t )sin θ
J. Baker
21
Measuring direction-dependent response spectra
Sa(θ,1s) / SaRotD100(1s)
Displacement / Maximum
displacement
SaRotD100 orientation
(T )
SaGeoMean
=
SaRotD100
= 2
SaRotD 50
Sa X (T ) 2 + SaY (T ) 2
SaRotD0 orientation
SaGMRotI 50 (T ) = ...
SaRotD50 orientation
Gilroy Array #6, 1984 Morgan
Hill
J. Baker
22
Example 1-second oscillator responses
to multi-component motions
Sa(θ,1s) / SaRotD100(1s)
Displacement / Maximum
displacement
SaRotD100
= 2
SaRotD 50
SaRotD100
≅1
SaRotD 50
HWA031, 1999 Chi-Chi-04
Gilroy Array #6, 1984 Morgan
Hill
J. Baker
23
Ground motion model for SaRotD100 at a specified period
SaRotD100 =
SaRotD100
SaRotD 50
SaRotD 50
ln SaRotD100 ln( SaRotD100 / SaRotD 50 ) + ln( SaRotD 50 )
=
“Max direction factor”
ln( SaRotD100 / SaRotD 50 ) = a + η '+ ε '
Simple prediction (constant?)
Independent of primary GMPE
Primary GMPE
ln( SaRotD 50 ) f ( M , R,VS 30 ,...) + η + ε
=
Complex prediction
`