Centrifuge tests to assess seismic site response of partially saturated sand layers
Introduction
Seismic waves generated by earthquakes often travel through soils with different mechanical and hydraulic characteristics where they can be dramatically altered in terms of intensity, frequency content, and duration. This transition is commonly evaluated using “Site Response Analysis”, which is a crucial step toward seismic design of soil-structure systems. Applications of site response analysis include development of design response spectra for surface structures, estimating seismically induced stresses, strains, and settlements, and liquefaction hazard assessment. Local site conditions such as soil density, plasticity index, stiffness, and damping can significantly affect seismic site response [1], [2], [3], [4], [5], [6], [7], [8]. Thus, ignoring the effects of changes in the site conditions may lead to inaccurate assessment of the site response.
The role of local site condition and the intensity of rock motion in the site response have been highlighted using fully monitored and instrumented sites during past earthquakes [1], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. In general, lower site amplification factors were observed during earthquakes with higher bedrock motion intensities [8], [11], [12], [13], [14], [15]. This could be attributed to the nonlinear stress-strain behavior of soils and higher damping values as a result of higher induced strain levels. The motion amplification was, also, found inversely proportional to the square root of shear wave velocity as a representative measure of local site conditions [1], [4], [5], [6], [7], [16]. Data obtained from instrumented sites under strong ground motions (e.g. [17], [18]) as well as physical modeling experiments (e.g. [19], [20]) can be used to develop guidelines for site response assessment. Traditionally, different methods have been employed to consider the effects of local site conditions and motion intensity in the surface motion evaluation, ranging from simplified procedures regulated by seismic provisions [21], [22], [23] to more complex site-specific ground response analysis for sensitive seismic designs using available software [24], [25], [26], [27], [28], [29]. In current seismic provisions the local site condition is reflected through site classification system using an average shear wave velocity of the top 30 m of the soil profile () (Table 1).
Degree of water saturation is among the parameters that influence the seismic response of soil layers [30]. Inter-particle suction in partially saturated soils increases the effective stresses on the grain skeleton [31]. This, in turn, yields to different soil dynamic properties including small-strain and strain-dependent shear modulus and damping [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43]. As a result, seismic wave propagation mechanisms may vary in partially saturated soil layers [44] that would lead to a different seismic site response [30], [45], [46], [47], [48], [49]. Soils in either dry or fully saturated conditions have been believed to result in more conservative solutions because matric suction in unsaturated soils increases the ground stiffness. Therefore, partial saturation has not been directly considered in the state-of-the-practice site response analysis. However, recent investigations on the site response in unsaturated soils showed that this assumption might not be always reasonable [47], [48], [49]. Further, the influence of the degree of saturation on seismic response analysis is often considered by incorporating the in-situ measured shear wave velocity of shallow unsaturated soil layers. However, the extent of this influence might be beyond the suction-dependency of the dynamic soil properties where the wave propagation mechanisms may vary [45]. In addition, soil properties may differ between the time of the construction and prior to the earthquake due to the seasonal fluctuation of water table. Thus, recognizing this difference would be essential in assessing the uncertainty in projected site response.
Yang [45] analytically studied the frequency-dependent amplification of inclined vertically propagated shear waves (SV waves) in soil layers overlying bedrock. The results indicated that a slight decrease in the degree of saturation of fully-saturated soil layers causes a dramatic difference in vertical amplification of the SV waves. Specifically, for regular earthquake frequencies, unsaturated soils may lead to a higher vertical amplification than saturated soil layers. D’Onza et al. [46] implemented the small-strain shear modulus and damping obtained from suction-controlled resonant column tests in a series of numerical site response analyses. Suction was found to significantly affect the natural frequency and Peak Ground Acceleration (PGA) amplification factor in clayey silt and silty sand layers. According to their numerical study, the natural frequency of the soil layers increased in higher suction values whereas PGA amplification factor was reduced. Ghayoomi et al. [30] studied seismically induced settlements in partially saturated sand by applying sinusoidal cyclic loads to sand layers in a set of suction-controlled centrifuge tests using steady-state infiltration technique [50]. The least amount of surface settlement occurred in middle range degrees of saturation due to the increase in shear modulus. Moreover, they observed a maximum 20% increase in PGA amplification factor in unsaturated sand with respect to the one in dry condition [47].
Recently, Ghayoomi and Mirshekari [48] and Mirshekari and Ghayoomi [49] numerically studied the seismic response of partially saturated sand and silt layers using site response software DeepSoil [24]. In the absence of any available numerical procedure to account for partial saturation, this influence was investigated by adjusting the soil unit weight and effective stress for any given degree of saturation. Changes of the effective stress in unsaturated soils, in turn, altered soil dynamic properties including small-strain and strain-dependent shear modulus and damping. Accordingly, partial saturation in the soil layers appeared to considerably influence the site response, where the extent of this effect was a function of soil type as well as induced motion characteristics. For example, partial saturation in sandy soils with low-range suction level (e.g. 10 kPa) resulted in higher amplifications and lateral deformations whereas in silty soils with high suction range (e.g. 70 kPa) led to lower amplifications and lateral deformations in comparison with those of dry soil layers.
Despite the proven influence of the degree of saturation on the dynamic soil properties and the site response, well-documented field or laboratory seismic site response data in partially saturated soils are still needed. Centrifuge physical modeling of free-field seismic ground response using a “degree of saturation-controlled” system is of a great value to validate this effect and to calibrate future numerical and analytical predictive models. This paper describes the adaptation and modification of an experimental setup to control the degree of saturation in a geotechnical centrifuge and its application to study seismic site response of partially-saturated soil layers. Furthermore, the effect of partial saturation on the site response of a sand layer is investigated and discussed in terms of different motion characteristics including PGA amplification factor (FPGA), low-period and mid-period amplification factors (Fa and Fv, respectively), 5% damped spectral acceleration, Arias intensity (Ia), lateral deformation, and seismically induced settlements.
Section snippets
Suction control in geotechnical centrifuge
Modeling unsaturated soils under high gravitational acceleration is a challenging task where controlling suction or the degree of saturation is the key to any systematic investigation involving partially saturated soils. Centrifuge modeling of unsaturated fine-grained soils could be accomplished by using methods such as compacting soils with a target degree of saturation [51] or in-flight free drainage of an initially saturated specimen [52]. For sand layers, however, centrifugation along with
Centrifuge, shaking table, and container
The recently renovated 5 g-ton centrifuge at the University of New Hampshire was used in this investigation [62]. The arm radius is 1 m from the center of rotation to the platform in its fully extended position. The centrifuge is equipped with an in-flight shaking table that is capable of imposing harmonic or seismic displacement time histories. A hydraulic servo valve controlled by National Instruments system operates the table. Due to servo valve limitations, the current system has an analog
Results and discussion
The centrifuge experiments were performed on three dry specimens (D1, D2, and D3), eleven unsaturated specimens (specified with the letter U followed by the degree of saturation) with degrees of saturation varying between 32% and 68%, and one saturated (S) specimen. The captured acceleration and displacement time histories were post-processed to obtain various motion characteristics. Then, the influence of the degree of saturation on the seismic response was investigated in terms of different
Summary and conclusion
The implementation process of a steady state infiltration system in a laminar container inside a geotechnical centrifuge was presented in this paper, which was successfully deployed to control degree of saturation and, consequently, matric suction in scaled physical models. Seismic excitation simulating scaled Northridge earthquake motion was applied to 11.43 m-thick dry, saturated, and unsaturated specimens of loose Ottawa sand with uniform degrees of saturation in depth. The recorded data were
Acknowledgment
The authors would acknowledge funding of this project by the National Science Foundation through the NSF CMMI grant No. 1333810.
References (77)
- et al.
Comparison of 1D linear, equivalent-linear, and nonlinear site response models at six KiK-net validation sites
Soil Dyn Earthq Eng
(2015) - et al.
Evaluation of 1-D seismic site response modeling of sand using centrifuge experiments
Soil Dyn Earthq Eng
(2015) - et al.
Stiffness of intermediate unsaturated soil from simultaneous suction-controlled resonant column and bender element testing
Eng Geol
(2015) - et al.
The effect of quaternary alluvium on strong ground motion in the Coyote Lake, California, earthquake of 1979
Bull Seism Soc Am
(1981) - et al.
Effect of soil plasticity on cyclic response
ASCE J Geotech Eng
(1991) - M. Vucetic, Soil properties and seismic response, In: Proceedings of the tenth world conference on earthquake...
- R.D. Borcherdt, Simplified site classes and empirical amplification factors for site-dependent code provisions. In:...
Estimates of site-dependent response spectra for design (methodology and justification)
Earthq Spectra
(1994)- R.D. Borcherdt and G. Glassmoyer, Influences of local geology on strong and weakground motions in the san francisco bay...
- S. Midorikawa, M. Matsuoka and K. Sakugawa, Site effects on strong-motion records observed during the 1987...
Nonlinear site amplification as function of 30 m shear wave velocity
Earthq Spectra
Relationships between maximum acceleration, maximum velocity, distance from source and local site conditions for moderately strong earthquakes
Bull Seism Soc Am
Site dependent spectra for earthquake-resistant design
Bull Seism Soc Am
Amplification factors for spectral acceleration in tectonically active regions
Bull Seism Soc Am
Empirical evidence for site coefficients in building-code provisions
Earthq Spectra
Estimation of response spectra and peak accelerations from Western NorthAmerican earthquakes: an interim Report, Part 2
US Geol Surv Open-File Report
Centrifuge modeling studies of site response in soft clay over wide strain range
J Geotech Geoenviron Eng
DEEPSOIL 6.1, user manual
User's manual for SHAKE91
DESRA-2: dynamic effective stress response analysis of soil deposit with energy transmitting boundary including assessment of liquefaction potential
Soil Mechanics Series
Seismic response of composite horizontally-layered soil deposits
The opensees command language manual
Empirical methodology to estimate seismically induced settlement of partially saturated sand
ASCE J Geotech Geoenviron Eng
Suction stress characteristic curve for unsaturated soils
J Geotech Geoenviron Eng
Capillary effects on dynamic modulus of sands and silts
ASCE J Geotech Eng
Resonant column tests on partially saturated sands
Geotech Test J
Small strain behavior of a silty sand in controlled-suction resonant column – torsional shear tests
Can Geotech J
Stiffness of an unsaturated compacted clayey soil at very small strains. Advanced experimental unsaturated soil mechanics
Effects of imbibition-drainage and stress ratio on anisotropic stiffness of an unsaturated soil at very small strains
Can Geotech J
Cited by (35)
Effectiveness of rolling dynamic compaction with a three-sided compactor on unsaturated sand
2023, Transportation GeotechnicsFailure of partially saturated frozen soils: A micromechanical analysis
2023, Cold Regions Science and TechnologyNatural rubber latex treatment of sand: A novel remediation technique for soil liquefaction
2023, Soil Dynamics and Earthquake EngineeringCitation Excerpt :Several experimental studies have corroborated the effectiveness of these desaturation techniques in preventing sand liquefaction. Even so, a few studies reported an increased acceleration of superstructure founded on desaturated soil mass [29–32]. Microbially induced calcite precipitation (MICP) is a fast-developing technique for liquefaction remediation, and its possibilities are being explored by researchers from different parts of the world [33–37].
Site response analysis of induced seismic events in the CENA region
2022, Soil Dynamics and Earthquake EngineeringAnalysis of unsaturated soil columns with application to bulk cargo liquefaction in ships
2021, Computers and GeotechnicsCitation Excerpt :These studies have shown that partial saturation significantly increases the resistance to liquefaction (e.g. Yoshimi et al, 1989, Unno et al, 2008, Tsukamoto et al, 2014) and it has been suggested that the initial degree of saturation needs to be greater than 0.7 combined with a low relative density for liquefaction to occur (Tsukamoto, 2019). Several model 1-g shaking table and centrifuge tests of partially saturated soils subject to cyclic loading have also been performed to investigate the potential for liquefaction, settlement and slope instability (e.g. Ghayoomi, 2011, Higo et al, 2015, Mirshekari and Ghayoomi, 2017) but these have been focused on earthquake loading with small numbers of cycles and mostly sandy soils. Centrifuge tests investigating the rocking motion on partially saturated cargo materials (Atkinson and Taylor, 1994), have shown that saturated conditions can develop at the base of the cargo and cyclic loading can cause the generation of pore pressure that results in a loss of strength and a loss in resistance to the shifting of cargo.
Seismic behavior of pile-supported systems in unsaturated sand
2018, Soil Dynamics and Earthquake EngineeringCitation Excerpt :The objective of this paper is to present a physical modeling program using geotechnical centrifuge, aimed to investigate the effect of the degree of saturation of a fine sand on the seismic behavior of pile-supported structures. The steady state infiltration technique [32,38] was used in this study to develop a relatively uniform degree of saturation profile along the soil layer. The experimental program, and the observed seismic response (in terms of acceleration, lateral displacement, and frequency content) of two distinct pile-supported systems under various degrees of saturation are explained.