Fig 1 - uploaded by Russell A Green
Content may be subject to copyright.
Source publication
The "equivalent number of cycles" concept is integral to cyclic liquefaction evaluations, whether applied directly in laboratory evaluations or via magnitude scaling factors in field evaluations. The premise of the concept is that the random motions of an earthquake can be represented by an equivalently damaging number of uniform stress cycles (n(e...
Contexts in source publication
Context 1
... of cycles N required to cause failure being noted. The test is repeated several times on similar compo- nents, subjecting each component to uniform cyclic stresses hav- ing different peak amplitudes. The S -N curve is a plot of the number of cycles to cause failure as a function of the peak am- plitude of the applied uniform cyclic stress e.g., Fig. 1. Several engineering materials e.g., ferrous alloys and titanium have clearly defined fatigue limits, which is the stress level below which an infinite number of cycles can be sustained without fail- ...
Context 2
... and alterna- tive implementations of the P -M hypothesis for a soil profile Table 1. Three plots to right of soil profile are median n eq values for different magnitude and distance bins, normalized by respective n eq at depth of 3.8 m. consisting of loose sand subjected to a ground motion recorded during the M7.3 Landers earthquake is shown in Fig. 10. The site response computer code SUMDES Li et al. 1992 was used to compute the results shown in this figure, wherein the soil was modeled using a reduced-order, bounding surface hypoplasticity model calibrated to the Ishibashi and Zhang's shear modulus deg- radation curves at the corresponding depths. As may be observed from this ...
Context 3
... eq was relatively constant with depth. However, Lee and Chan 1972 implemented the P -M hypothesis as outlined above for high cycle fatigue conditions. The results of a parametric study using the alternative imple- mentation of the P -M hypothesis show that the variation of n eq with depth is dependant on site-to-source distance, as illustrated in Fig. 11. Note: In this study, magnitude M was quantified by the moment magnitude scale M w and site-to-source distance R was quantified by the closest distance to the fault. For M 7, the variation of n eq with depth tends to increase as R increases up to R =75 km i.e., the average R for the bin, beyond which n eq is relatively independent of ...
Context 4
... by the closest distance to the fault. For M 7, the variation of n eq with depth tends to increase as R increases up to R =75 km i.e., the average R for the bin, beyond which n eq is relatively independent of R. For M 7, the decrease in n eq with depth is very pronounced, but the trend is relatively independent of R. The n eq curves shown in Fig. 11 represent the median n eq values resulting from a series of site response analyses. The ground motions used in the analyses were from 29 different earth- quakes, 163 stations i.e., 326 records, and were selected to be representative of western United States rock motions McGuire et al. 2001. Table 1 lists the earthquake motions used in ...
Context 5
... tendency for n eq to decrease with depth may explain, in part, why liquefaction is generally constrained to the upper 12 m 40 ft of a soil profile, as may be observed from Fig. 12. This figure is a histogram of the number of observed occurrences of liquefaction as a function of depth, with the 92 liquefaction cases used to construct the histogram coming from Cetin 2000. Below a depth of about 3.8 m, the trend in the decrease of the number of occurrences of liquefaction as a function of depth mimics closely the ...
Context 6
... Below a depth of about 3.8 m, the trend in the decrease of the number of occurrences of liquefaction as a function of depth mimics closely the decrease in the n eq as a function of depth for "far field" i.e., R 75 km earthquake motions. Above a depth of about 3.8 m, the decrease in the number of observed occurrences of liquefac- tion shown in Fig. 12 is likely due to upper strata being nonlique- fiable e.g., above the water table, cohesive material, etc.. Note: other factors may also contribute to the trend shown in the Fig. 12, such as increased "aging" effects with depth and/or deep liq- uefaction does not always manifest at the soil ...
Context 7
... of depth for "far field" i.e., R 75 km earthquake motions. Above a depth of about 3.8 m, the decrease in the number of observed occurrences of liquefac- tion shown in Fig. 12 is likely due to upper strata being nonlique- fiable e.g., above the water table, cohesive material, etc.. Note: other factors may also contribute to the trend shown in the Fig. 12, such as increased "aging" effects with depth and/or deep liq- uefaction does not always manifest at the soil ...
Context 8
... addition to the depth dependency, the results of the para- metric study showed n eq varies as a function of site-to-source distance, as illustrated in Fig. 13. As may be observed from this figure, for a given magnitude earthquake n eq increases with site- to-source distance R up to R =75 km i.e., the average R for the bin, after which n eq is relatively independent of R. This trend is a result of setting the amplitude of the equivalent cycle ref pro- portional to a fixed percentage of the ...
Context 9
... therefore, a fixed fraction of max is representative of the amplitude of the induced peak stresses at other times. In con- trast, in the near field the stress time histories are much more chaotic, and often max is associated with a high frequency spike which is not representative of the entire time history. Also influ- encing the trend shown in Fig. 13 are the competing effects of increased duration of the ground motion with distance e.g., Abra- hamson and Silva 1996 and decreased motion amplitude with distance. The increased duration of motion with distance would result in an increase in n eq , while the decrease in amplitude would result in a decrease in n eq ...
Context 10
... finding that n eq is a function of site-to-source distance was also made by Liu et al. 2001, although the exact natures of the trends differ between the two studies, as may be observed from Fig. 13. In contrast to the trends resulting from this study, Liu et al. 2001 show that n eq continually increases with site-to-source distance. However, as with Chan 1972, Liu et al. 2001 implemented the P -M hypothesis for high cycle fatigue condi- tions, which may account for the differences in the resulting trends. Finally, also shown in ...
Context 11
... from Fig. 13. In contrast to the trends resulting from this study, Liu et al. 2001 show that n eq continually increases with site-to-source distance. However, as with Chan 1972, Liu et al. 2001 implemented the P -M hypothesis for high cycle fatigue condi- tions, which may account for the differences in the resulting trends. Finally, also shown in Fig. 13 are the n eq relations pro- posed by Seed et al. 1975 which are independent of ...
Similar publications
Cyclic loading of saturated soils is often accompanied by pore pressure accumulation, the magnitude of which, has a tendency to significantly affect the post-cyclic characteristics. This study discusses the post-cyclic behaviour of fine sand and low plastic silt in laboratory testing. Different levels of excess pore pressure ratios were first allow...
Citations
... Such conversion depends on the technique used to choose and count the applied stress cycles and the adopted conversion curve. Thus, the necessary procedures are frequently complex and unreliable [43][44][45]. For this reason, the adopted methodology generally consists of relating earthquake magnitude to the equivalent number of loading cycles [22,[46][47][48] even if these correlations often result in inaccurate estimates of the number of equivalent loading cycles [44]. ...
During an earthquake, excess pore water pressure generation in saturated silty sands causes a reduction in shear strength and even liquefaction of the soil. A comprehensive experimental program consisting of undrained cyclic simple-shear tests was undertaken to explore the key factors affecting the energy-based excess pore water pressure generation models for non-plastic silty sands. The examined influencing factors were non-plastic fines content (less than and greater than the threshold value ≈ 25%), packing density, vertical effective stress, applied cyclic stress ratio, and soil fabric. The relationship between excess pore water pressure ratio and dissipated energy per unit volume was found to be mainly dependent on the relative density and fines content of soil, whereas the cyclic stress ratio, initial vertical effective stress, and soil fabric (i.e. the reconstitution method) appeared to have a minor impact. A revision of the original energy-based model developed for clean sand by Berrill and Davis was proposed to improve prediction accuracy in terms of residual excess pore water pressures versus normalised cumulative dissipated energy. Nonlinear multivariable regression analyses were performed to develop correlations for the calibration parameters of the revised model. Lastly, these correlations were validated through additional cyclic simple-shear tests performed on different silty sands recovered at a site where liquefaction occurred after the 2012 Emilia Romagna (Italy) earthquake.
... In other words, this indicates that the E.P.W.P is not highly dependent on the duration of GMs. Instead, it is widely recognized that the development of E.P.W.P depends on the equivalent number of earthquake event cycles [69]. ...
In earthquake-resistant design, amplitude and frequency content of ground motions (GMs) have been considered using spectral matching techniques; however, duration effects remain insufficiently explored in designing buildings in liquefiable soils. This study investigates the influence of ground-motion duration on seismic response of shallow-founded buildings under strong earthquakes. Buildings in liquefiable soils are analyzed using nonlinear dynamic analysis with coupled up formulations. The numerical code and calibrated constitutive parameters of Toyoura sand are validated through dynamic centrifuge testing. Two ground-motion suites, including 30 pairs of long and short-duration events, are scaled to the target PGA of 0.3 g, and then selected to be spectrally equivalent to isolate duration measures from the others. Comparative results show that longer duration events result in greater settlements and tilt compared to shorter events. Therefore, this study emphasizes the importance of considering duration measures in assessing seismic responses of buildings. Furthermore, correlation between settlements and peak transient tilt, and intensity measures (IMs) of GMs are comprehensively analyzed. It is found that employing compound IMs can lead to notable improvements in predictive accuracy for settlement and peak transient tilt compared to single common IMs. The compound IMs, namely CAV 2/3 × Ds 5-95 1/3 and SMV × Ds 5-95 1/3 , are newly proposed for use in order to achieve best correlation with the shear-induced settlements and peak transient tilt, respectively.
... The equivalent number of stress cycles, N eq , associated with each acceleration time history and a range in exponent b was computed following the methodology presented by Boulanger and Idriss (2015) based on the cycle counting procedure developed by Seed et al. (1975). In this approach, the ratio of the CSR (i.e., τ cyc =σ 0 v0 ) and the absolute maximum CSR, CSR max is assumed equal to the ratio of acceleration and absolute maximum acceleration or PGA (Green and Terri 2005;Verma et al. 2019). The computation of N eq requires selection of a reference CSR, CSR ref , set equal to 0.65 CSR max (Seed and Idriss 1971;Seed et al. 1975) herein; additionally, any absolute CSR associated with a positive or negative half-cycle less than 0.1 CSR max was neglected, following Boulanger and Idriss (2004) and Verma et al. (2019). ...
... As PI increases, b reduces, and N eq increases at an increasing rate. The N eq derived in this study for subduction zone events (i.e., interface and intraslab) is generally similar to the results of previous studies (e.g., Boulanger and Idriss 2004;Verma et al. 2019); however, slight differences may be noted and are attributed to the differences in the number of ground motions evaluated and their characteristics (Green and Terri 2005;Kishida and Tsai 2014). ...
Regression models are presented to estimate the cyclic resistance of transitional silts as an alternative to semi-empirical case history- and stress history and normalized engineering parameters- (SHANSEP-) based models using a database of cyclic laboratory tests on intact specimens. Expressions for the plasticity index-, PI-, dependent semi-logarithmic slope, exponent b, of the curve describing the variation of cyclic resistance ratio, CRR, and number of loading cycles, N, and the variation of number of equivalent loading cycles, Neq, with b are developed for subduction zone earthquakes alongside corresponding magnitude scaling factors for use in cyclic failure assessments (i.e., liquefaction triggering and cyclic softening) within the Simplified Method framework. A PI-, overconsolidation ratio-, OCR-, and shear strain-dependent model for the variation of CRR with N is presented to estimate cyclic resistance as function of cyclic shear strain failure criteria ranging from 1% to 10%. The suite of models can be used together or separately to facilitate assessments of cyclic failure of, or plan cyclic laboratory test programs and/or calibrate constitutive models for transitional silts in the absence of site-specific laboratory tests.
... The value of n eq;M is estimated using a relationship developed by Lasley et al. (2017). The relationship was developed using a lowcycle implementation of the Palmgren-Miner fatigue theory (Green and Terri 2005) and accounts for plastic response of the soil and for multidirectional shaking. The recommended form of the equation used in this study is ...
... Additionally, if required inputs are available, equivalent linear site response analyses can be performed to estimate ΔW directly. Toward this end, both the shear stress and shear strain time histories at the depth of interest should be output from site response analyses, and ΔW is calculated as the cumulative area enclosed by the stress-strain hysteresis loops (Green and Terri 2005). This area can be estimated using the trapezoidal rule ...
The most commonly used approach for evaluating liquefaction triggering is via stress-based simplified models. Proposed herein is a model for evaluating liquefaction triggering where the imposed loading and ability of the soil to resist liquefaction are quantified in terms of normalized dissipated energy per unit volume of soil (ΔW/σ'vo), computed within a total stress framework. The proposed model overcomes limitations of many previously proposed energy-based triggering models. Additionally, the proposed energy-based model unites concepts from both stress-based and strain-based procedures, overcoming some of their limitations, and in its simplified form is implemented similarly to the simplified stress-based models. An updated field case history database is used to develop probabilistic limit-state curves. These limit-state curves express ΔW/σ'vo required to trigger liquefaction as a function of corrected cone penetration test tip resistance (q c1Ncs) for different probabilities of liquefaction (PL) and have comparable predictive abilities to stress-based limit-state curves in terms of number of correct predictions for the cases analyzed. However, because dissipated energy is a scalar quantity, multidirectional shaking and other effects such as soil-structure interaction, nonvertical wave fields, and topographic site effects can readily be accounted for. Additionally, the applicability of the proposed triggering curve is not limited to earthquake loading but, rather, can be used in relation to other sources of vibrations (e.g., construction vibrations and explosive loading, among others).
... To obtain N eq , one should first choose the reference peak amplitude of equivalent uniform cyclic loading (i.e., τ e ) and the corresponding number of loading cycles to trigger liquefaction (i.e., N e ), which damage effect is regarded equal to that of another combination of uniform loading τ i with the corresponding number of loading cycles N if as shown in Fig. 2. It means that the damage effect of each cycle of shear stress with amplitude τ i is equivalent to that uniform shear stress τ e with number of loading cycle N e /N if. Based on Fig. 2, a series of irregular loading could be converted into an equivalent uniform cyclic loading of amplitude (τ e ), and then calculating N eq according to the following equation [31]: ...
... Empirical models for the evaluation of the equivalent number of cycles have been proposed in various studies in the literature [34,35,[39][40][41][42][43][44][45]. Initially, the empirical relations employed only the magnitude as a dependent parameter, while the more recent models consider other parameters as well (source-site distance, soil conditions, faulting mechanism, other ground motion parameters, etc.). ...
This paper is focused on the evaluation of the liquefaction hazard for different sites in Romania. To this aim, a database of 139 ground motions recorded during Vrancea intermediate-depth earthquakes having moment magnitudes MW ≥ 6.0 is employed for the evaluation of the equivalent number of cycles for this seismic source. Several functional forms for the empirical evaluation of the equivalent number of cycles considering various seismological or engineering parameters are tested and evaluated. The regression analysis shows smaller uncertainties for the empirical models based on ground motion engineering parameters. Considering the lack of information in terms of engineering parameters, a simpler empirical model which accounts for the earthquake magnitude, source–site distance and soil conditions is selected for the liquefaction hazard analysis. Based on the proposed empirical model, specific magnitude scaling factors for Vrancea intermediate-depth earthquakes are proposed for the first time as well. The liquefaction hazard analysis is performed for sites whose seismic hazard is generated by either the Vrancea intermediate-depth seismic source or by local shallow crustal seismic sources. In the case of some of the selected sites, liquefaction phenomena were observed during past large-magnitude earthquakes. Unlike previous studies dealing with liquefaction analyses for sites in Romania, in this research, the hazard assessment is performed for various ground motion levels evaluated based on probabilistic seismic hazard assessment. Liquefaction hazard curves are constructed for each analyzed site. The results of the liquefaction hazard analysis show that this phenomenon is more likely to occur in the areas exposed to Vrancea intermediate-depth earthquakes, compared to the areas affected by local shallow earthquakes. In the case of the analyzed soil profiles from Bucharest, Craiova and Ianca, the minimum liquefaction safety factors less than one even for seismic hazard levels having mean return periods of 100 years and less.
... During the last few decades, seismologists and civil engineers have studied the characteristics of seismological events more precisely and have developed methods to identify several features [3]. Among the seismological features studied, the pulse-type near-fault ground motions (NFGMs) were extensively employed to examine the structural responses of high dams [4], bridges [5], and liquefaction events [6]. In particular, after several as-recorded extreme earthquake events, contemporary studies have revealed that the pulsetype NFGMs at short site-to-source distances were more destructive than those GMs at long distances [7][8][9][10][11]. ...
Current research trends in significant hydraulic engineering projects focus on investigating seismological properties of intensity and frequency content of pulse-type near-field earthquakes on structural response. Conversely, the duration impact is not expressly addressed in the seismic design code for underground buildings. Simultaneously, current various duration indicators of as-recorded strong ground motions mainly consider the effective duration of the initial accelera-tion component record. In contrast, the duration indicators for the effective velocity duration (EVD) of the original velocity time-history component record have rarely been addressed. That is to say, there is a gap between the effective velocity duration and the structural response. To il-lustrate the impact on the structural response, an EVD of pulse-type NFGM duration was used. This EVD can be calculated for seismic excitations with set threshold values that enable a quan-titative examination of the duration effects. A fluid-hydraulic tunnel-rock interaction system was built and used to estimate the seismic response characteristics induced by different duration NFGMs. The investigation’s findings highlight that the inelastic dynamic response and damage degree are strongly affected by the EVD. Meanwhile, the fixed threshold value of 5%-95% shows an excellent correlation coefficient with the structural response. The significant duration was also found to be the most suitable alternative indicator to replace the EVD index. In addition, the reduced time-history methodology of near-fault earthquake records is presented and validated, using this method to improve the efficiency of the dynamic time-history analysis of hydraulic arched tunnels.
... The equivalent number of cycles is evaluated and the number of cycles for the related current load bin is added ( Figure B.1 (a)). There are also other approaches which use, for instance, a normalized CSR curve (Seed, 1975;Green and Terri, 2005;Chang, 1981;Allotey and Naggar, 2008). A detailed overview of the calculation procedures can be found in Glasenapp (2016) and . ...
In particular during storm events an accumulation of excess pore pressures may occur in
the soil around cyclically loaded offshore foundations. The excess pore pressure build-up
reduces the effective stresses in the soil and, hence, may negatively affect the structural
integrity by influencing the soil-structure interaction. Besides a loss in bearing capacity,
large plastic deformations may occur to the structure. Especially for offshore wind turbines
an accurate estimation of such deformations is of great importance. Even though the
consideration of this degradation effect on the bearing capacity is commonly demanded by
the involved certification or approval bodies, no general applicable and accepted method
for the calculative verification currently exists. Over the past decades several researchers
investigated the excess pore pressure build-up around offshore foundations due to environmental
cyclic loads. They tried to capture the loss of bearing capacity, the accumulation
of plastic rotation and the essential influence on the serviceability limit state and fatigue
design. However, even if there are some sophisticated concepts, none of them is seen as
the simple general applicable choice.
Within this thesis a new numerical method – termed Excess Pore Pressure Estimation
method (EPPE) – is presented in great detail. This method allows for the transfer of
the soil behaviour obtained in cyclic simple shear tests to the bearing behaviour of the
entire foundation. Herein, the numerical model accounts for the cyclic excess pore pressure
accumulation by respecting the element-based mean stress and stress amplitude as
well as an equivalent number of load cycles. The simulation of the excess pore pressure
build-up due to certain cyclic loading is based on undrained conditions, i.e. the excess
pore pressure build-up due to cyclic loading is derived by disregarding the simultaneous
consolidation process. The respected transfer method, in the form of contour plots, enables
the consideration of site-specific cyclic direct simple shear and triaxial test results
from laboratory devices to elements within the finite element model. Each integration
point is evaluated individually. Based on the derived excess pore pressure field, a consolidation
analysis takes place in the second step. The actual accumulated excess pore
pressure in each element at the end of the storm (or cyclic loading event) is then found
by analytically superposing the excess pore pressure decay curves from the consolidation
analysis.
For a deeper understanding of cyclic soil behaviour, the cyclic response in different laboratory
devices with different densities and under varying stress states was investigated by
the author. A contour approach based on cyclic load- and displacement-controlled test
results is derived to study the element response from the numerical point of view and
use these for the calibration of an implicit model. Moreover, different explicit approaches
are presented and compared in terms of their estimation behaviour of cyclic excess pore pressure generation, their predicted foundation capacity and their model assumptions.
The intention is hence to examine existing approaches and their applicability by means of
an elaborate comprehensive study. A simple modular explicit model is presented which
can be easily assessed with engineering judgment. If needed, the different individual
calculation steps can be exchanged with more sophisticated ones.
For a reference sandy soil, results of cyclic laboratory tests are presented and used on a
reference monopile foundation for a predefined storm event. The EPPE approach helps
to quantify the risk of capacity degradation as well as to evaluate an appropriate safety
margin. It is possible with the current methodology to evaluate the degradation potential
for different sites quite easily and fast.
... For historical reasons, MSFs quantify the influence of the duration of shaking on liquefaction relative to the duration of a moment magnitude 7.5 (M7.5) event, where the durations of events having moment magnitudes M and 7.5 are expressed in terms of their corresponding numbers of equivalent stress cycles, N eq and N eq M7.5 , respectively. As noted previously, procedures for computing N eq may also rely on b-values (e.g., Liu et al. 2001;Green and Terri 2005;Hancock and Bommer 2005;Stafford and Bommer 2009;Lasley et al. 2017). For example, if a CSR-N L curve plots as a straight line in log-log space, then the Seed et al. (1975) procedure for computing N eq can be expressed as ...
... Energy is primarily dissipated in sands due to friction developed from relative movement between sand grains as the soil skeleton breaks down under cyclic loading. In addition, dissipated energy underlies the commonly used approaches to compute N eq for earthquake motions, whether explicitly (e.g., Green and Terri 2005;Lasley et al. 2017) or implicitly (e.g., Seed et al. 1975;Boulanger and Idriss 2015). These fundamental relationships make a strong case for using an energy-based criterion to identify liquefaction initiation in laboratory tests and thereby define b-values. ...
... The energy-based criterion proposed herein is consistent with how N eq is determined using fatigue theories (Green and Terri 2005) and correlates well with pore pressure generation (Green et al. 2000;Polito et al. 2008). In this study, the laboratory testing results showed that b-values derived from using ΔW total =σ 0 vo ¼ 0.001 as the liquefaction triggering criterion were relatively insensitive to changes in D r and σ 0 vo , which justifies the use of a constant b-value rather than a b-value that is a function of D r and σ 0 vo . ...
Magnitude scaling factors (MSFs) account for the influence of ground motion duration on liquefaction triggering in simplified stress-based models in which the duration of the motion is quantified in terms of number of equivalent stress cycles (N eq). Central to computing N eq and MSF is the relationship relating the amplitude of applied loading and the corresponding number of cycles to trigger liquefaction, that is, cyclic stress ratio ðCSRÞ-N L curves. Based on empirical evidence (and mathematical convenience), CSR-N L curves are commonly assumed to plot as straight lines on log-log scales, with the line having a slope of −b. As such, the b-value is central to computing N eq and MSF and has a significant influence on computed normalized seismic demand in simplified liquefaction evaluations. It is widely assumed that the b-value varies significantly as a function of soil density. However, in this study a review of published laboratory data and analysis of constant-volume cyclic direct simple shear tests performed as part of this study were used to assess the dependency of the b-value on soil density and other factors. We show that the criterion used to define liquefaction triggering in laboratory tests and the nonlinearity of the CSR-N L curves can result in the apparent dependency of the b-value on soil density. However, using a consistent liquefaction criterion based on the cumulative dissipated energy in a unit volume of soil yields b-values that are relatively insensitive to changes in soil density. Published modulus reduction and damping (MRD) curves can be used to compute b-values using an energy-based framework; this yields more generalized and less test-and soil-specific b-values. As a result of these efforts, a b-value of 0.28 is recommended for computing N eq and MSF, independent of soil density.
... The fourth method includes strain energy-based methods developed by applying seismic energy dissipated in the soil [21][22][23][24][25][26][27][28][29][30]. This method has been applied in three main procedures by researchers which are using histories of site exploration liquefied [29][30][31], and laboratory test results [23,27,[29][30][31][32][33][34][35][36][37][38][39][40] and Arias intensity-based models [32,41]. To evaluate the potential of liquefaction in energy concept method, the capacity strain energy (W ) value of the soil is required to be estimated to compare with the energy transferred to the soil by the earthquake loads. ...
... Based on laboratory test results six input parameters including effective confining pressure (σ ' c ) kPa, initial relative density (D r )%, FC%, coefficient of uniformity (C u ), mean grain size (D 50 ) (mm) and coefficient of curvature (C c ), have been identified and confirmed as the most influential factors in modeling liquefaction to estimate liquefaction resistance of sandy soil based on capacity strain energy concept [23,27,[29][30][31][32][33][34][35][36][37][38][39][40]. Clearly, permeability of the soil is considered implicitly in soil properties parameters of D r , C u , D 50 and C c . ...
... In this study, two different databases were arranged to train two ANN models to investigate the complex influence of FC on liquefaction resistance. According to previous research [23,27,[29][30][31][32][33][34][36][37][38][39]41], six parameters of σ c (kPa), D r (%), FC(%), C u , D 50 (mm), and C c were assigned as the inputs to create ANN models to calculate Log (W ) as a target. The first dataset includes 284 experiments created by Baziar et al. [45], including 217 cyclic triaxial laboratory test results [87], six laboratory cyclic simple shear experiments [87] and 61 cyclic torsional laboratory tests [39,86] in addition to 22 samples added from Verification of Liquefaction Analyses by Centrifuge Studies (VELACS) [44,79,88], 48 cyclic trixial laboratory test results [89], 20 laboratory test results from Dief [81] and 27 cyclic torsional laboratory test results [44]. ...
Liquefaction is one of the most destructive phenomena caused by earthquakes, which has been studied in the
issues of potential, triggering and hazard analysis. The strain energy approach is a common method to investigate
liquefaction potential. In this study, two Artificial Neural Network (ANN) models were developed to estimate the
liquefaction resistance of sandy soil based on the capacity strain energy concept (W) by using laboratory test
data. A large database was collected from the literature. One group of the dataset was utilized for validating the
process in order to prevent overtraining the presented model. To investigate the complex influence of fine content
(FC) on liquefaction resistance, according to previous studies, the second database was arranged by samples with
FC of less than 28% and was used to train the second ANN model. Then, two presented ANN models in this
study, in addition to four extra available models, were applied to an additional 20 new samples for comparing their
results to show the capability and accuracy of the presented models herein. Furthermore, a parametric sensitivity
analysis was performed through Monte Carlo Simulation (MCS) to evaluate the effects of parameters and their
uncertainties on the liquefaction resistance of soils. According to the results, the developed models provide a higher
accuracy prediction performance than the previously published models. The sensitivity analysis illustrated that the
uncertainties of grading parameters significantly affect the liquefaction resistance of soils.
