Geoscience Reference
In-Depth Information
waves arrivals induced by the basin edge, while
PGV
by the first S-waves. Furthermore,
although ground strains affecting buried pipelines are expected to be more pronounced
for medium to long period waves, the surface geology and site effects may play a very
important role as well, especially when they affect also the apparent velocity and the
ground motion incoherence.
Afurtherimportantremarkshouldbemadeonthestraincomponent,eitheraxialorshear,
to be adopted. It is well known that, while for shallow small-diameter buried pipelines
seismic design is mainly governed by axial loads, for large-diameter pipelines or tunnels
bending and cross-sectional distortions induced by shear strains may also be significant.
Since the ratio of axial to shear strain at a certain location strongly depends on the wave
type, incidence angle, earthquake magnitude and depth, as will be shown later in this
paper by some preliminary numerical results, we have distinguished throughout the text
peakaxialstrains,denotedas
PGSa
fromthepeakshearstrains,denotedas
PGSs
,unless
this distinction isnot relevant for presentation of results.
In the first part of this paper some of the previous limitations in the use of eq. (18.1)
will be discussed, addressing mainly (i) the experimental validations of the empirical
relationships for peak ground strain evaluation, highlighting the role of spatial inco-
herencyofgroundmotion,and(ii)theeffectofstronglateraldiscontinuitiesinsoilprop-
erties.
In the second part, we introduce some recent studies on the seismic risk assessment of
buriedlifelinesystems,mainlyrelatedtotheexperienceacquiredinsomerecentprojects
in Europe and Turkey in order to highlight the practical implications of ground strains
and
PGV
evaluation inthe vulnerability assessment of complex lifelinenetworks.
2. Earthquake-induced transient ground strains
2.1. EXPERIMENTALLY BASED
PGS-PGV
RELATIONS
FROM DENSE SEISMIC NETWORKS
Due to the lack of direct strain measurements during seismic ground motion, an experi-
mental basis should be provided to the evaluation of earthquake-induced ground strains,
typically in terms of a suitable measure of peak ground strain and its relationship with
other more frequently used ground motion parameters, such as peak ground velocity
(
PGV
)oracceleration(
PGA
).Apossiblesolutionistomakeuseofdatacollectedwithin
very dense seismic networks (Bodin et al., 1997; Gomberg et al., 1999). This approach
hasbeenrecentlyusedbySmerzinietal.(2006),whoconsideredtheweakmotionrecords
of a temporary network in Parkway Valley, New Zealand, in which the close spacing of
thestations,oftheorderoffewtensofmeters,allowedthemtoreconstruct,throughasuit-
able interpolation procedure, the three-component displacement field at ground surface,
from which the components of the strain tensor at ground surface could be calculated.
Subsequently, the same procedure has been applied by Paolucci and Smerzini (2007)
