PROBABILISTIC SITE IDENTIFICATION ANALYSIS USING NUPEC RECORDED FREE FIELD MOTIONS. Page: 2 of 7
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
8.9 gal to 174: gal, representing a reasonable range
of energy transmissions through soils that can be
used to calibrate:.both low strain .free-field soil
profiles and thee' strain, dependent modulus
degradations. The objective of the BNL free-field .
analysis is to identify the test site soil properties in
terms. of shear wave velocity using the
seismometer recordings from these earthquakes.
The methodology -employed in the BNL
probabilistic :, analysis foray site identification
includes the Monte Carlo process in conjunction
with equivalent linear convolution; analyses. The
random variable selected to characterize, the site
profile is the shear wave velocities in each: soil
layer of the site profile. A lognormal distribution is
assumed with the standard deviation determined .
from available site data and applicable regulatory
requirements. The':median values, are determined
through the Fourier ratio methodwith least square
fitting (Xu; 2001). The convolution analyses were
performed using the CARES program (Xu, 1993)
with the appropriate: soil degradation model. and.
the outcrop input motions generated .from the...
recorded in-column rock' motions. The BNL
analysis produced results . in terms of the mean,'
median and various fractiles' of: free-field.: soil;
properties. at the test site, and the corresponding.
surface response spectra. In this paper, the method
and :procedure-. employed to perform :, the
probabilistic site analysis are,: described first,
followed by the results of the BNL analysis with,
thirty. randomly sampled profiles generated from
one earthquake motion recorded at the .NUPEC.
ANALYSIS APPROACH' ,
Conventional 'probabilistic analysis consists of
steps, including:.. .1) identification of random
variables and' their associated' distributions; 2)
modeling. of probabilistic :process for, developing '.
various statistical data which ' characterize, the,
random variables. under " consideration;'. 3)
assessment 'of uncertainties associated with ;the:
quantification of the random: variables. In the site
identification analysis,: soil properties such as unit
weight, damping, shear wave velocity (V,) and the "
thickness of soil .layers all have inherent
randomness in their measurements, and therefore
could all be selected as random variables.- The
probabilistic -process established based on all soil
properties and the associated uncertainties .could..
be rather complicated and requires significant ,
effort to simulate. However, since the SSI effect is
primarily influenced by the shear. wave velocity of '.
soils, BNL selected V, to be the random variable
in order to simply :the analysis. The measurement
of the other properties could be improved through.
the application 'of better techniques and
instruments. The soil: damping is characterized
through the GEI modulus;degradation model (GEI,
1983), which has been established to be 'suitable..
for the NUPEC test site...
The BNL analysis assumes . a lognormal
distributions. for V. The probability density
function . (PDF) of V,, therefore, assumes , the
f = 1 2n v -n v,,,,d
v, (v, exp L
where vsmed represents .the median estimate of VS
and: p is : the. lognormal standard' deviation of .Vs
characterizing:, the uncertainty" for 'the vsmed
estimate. The..low-strain vsmed is determined using
the Fourier ratio .method with. a non-linear. least
square fitting. technique. for 'correlating the
analytical site amplification with the amplification
computed using the recorded . seismometer, data.
The BNL' estimate (Xu, 2001) of the low strain
vsmed is shown inFigure 2. .The lognormal standard
deviation is determined using the guidelines of the
current version '(NUREG-0800) of the USNRC
Standard ' Review Plan. (SRP). The SRP
recommends that the 1.5G and G/1.5 variations :be-
used.in SSI'analyses to account for the free field
soil uncertainties. Where,. G represents the best
estimate of the shear modules of soils. The range
of uncertainties determined by 1.5G and G/1.5 is:
about the 15th to 85th, percentiles, corresponding
to the mean+3 and mean-p estimates. With the aid
of the relation::V, = (G/p)'2, and the following
,# =.In v-. ,
the lognormal standard deviation p for -V. is.
readily calculated to be 0.2.
In order to apply-the Monte Carlo method,'random
sampling of the low strain Vs of the free field
profile needs be performed.. This is done by using
the. uniform, distribution for. sampling and the
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
XU,J.; COSTANTINO,C.; HOFMAYER,C.; MURPHY,A. & KITADA,Y. PROBABILISTIC SITE IDENTIFICATION ANALYSIS USING NUPEC RECORDED FREE FIELD MOTIONS., article, August 4, 2002; [Upton, New York]. (digital.library.unt.edu/ark:/67531/metadc874926/m1/2/: accessed May 28, 2018), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT Libraries Government Documents Department.