Abstract

     The paper studies the performance of Offshore Wind Turbines (OWTs) founded on Suction Bucket Jackets (SBJs) in clay under combined wind and seismic loading. A detailed 3D FE model of the soil–foundation–structure (SFS) system is developed and used as a benchmark to assess the efficiency of an enhanced Winkler-based “Caisson-on-Winkler-Soil” (CWS) model, where the soil is replaced by nonlinear hysteretic elements. The proposed CWS model captures residual deformations and hysteresis and offers physical coupling between vertical and moment loading. It allows excellent prediction of the  failure envelope in the most relevant first quadrant of  space, where the loads act in the same direction. Despite its successful application for the simulation of inertial loading, the CWS model fails to reproduce the dual shearing mechanism that develops at the caisson shaft during shaking, stemming from the combination of kinematic soil shearing due to the vertically propagating shear waves, and shearing due to the superstructure inertial response, thus underpredicting the co-seismic caisson settlements. For the prediction of the latter, the research utilizes spectrum compatible input motions and 3D FE models of varying geometric and material properties to derive linear regression equations that correlate the co-seismic dimensionless settlement of caissons ) with characteristic dimensional variables of the problem under investigation and the Arias Intensity  of the surface ground motion. As a final step, the paper proposes a hybrid method for performance-based assessment of SBJ OWTs. The proposed method employs the simplified CWS model to calculate the VHM loads and approximately estimate horizontal displacements and rotations at the jacket legs, followed by a preliminary assessment of caisson settlements using the correlations of  with  on the basis of spectrum-compatible input motions.

Introduction

     Following the ambitious energy targets set by countries worldwide, the offshore wind sector has seen impressive growth over the last decade, transforming from niche technology to a global industry. As of 2019, Europe alone has 22 GW of installed offshore wind capacity – enough to cover 2.3% of its electricity consumption [1], while similar or even more potent growth is observed in many countries around the world, such as China, the USA, and India. Part of this energy transition is led by technology-driven innovation, with offshore wind turbines reaching new heights and larger capacities, while wind farms are moving deeper into the sea to harness the increased energy potential. Giventhis new “deep-water” environment, traditional OWT support structures are gradually being replaced by more cost-effective and agile foundation systems in a quest for reduction of investment costs.

Conclusions

     The paper has developed a simplified performance-based assessment technique for SBJ OWTs founded in clay under the combined action of wind () and earthquake () loading. Using an example of an 8 MW jacket-supported OWT, installed at 60 m depth in the Adriatic Sea, system performance was assessed employing a detailed 3D FE model of the soil–foundation–structure (SFS) system (global 3D FE model). After deriving insights on system performance, the global 3D FE model was used as a benchmark to assess the efficiency of an enhanced Winkler-based “Caisson-on-Winkler-Soil” (CWS) model. Soil–suction caisson interaction is represented by nonlinear hysteretic elements, capturing residual deformations and hysteresis. The proposed CWS model offers physical coupling between vertical and moment loading by introducing distributed vertical hysteretic elements along the caisson shaft, simultaneously contributing to vertical and moment shaft resistance.

Abstract

     An increasing number of offshore wind farms are being constructed in seismic regions over liquefaction susceptible soils. This paper presents a methodology for the analysis and design of monopiles in seismically liquefiable soils by extending the established “10-step methodology” with an additional 7 steps. These additional steps include assimilation of seismic data, site response analysis, stability check of the structure (ULS check through the concept of load-utilization ratio), input motion selection, prediction of permanent tilt/rotation, and ground settlement post liquefaction. A flow chart, which shows the interdependence of the different disciplines, is presented and can be extended to routine design. This proposed method is validated using the observed performance of an offshore and nearshore turbine from the Kamisu wind farm during the 2011 Great East Japan earthquake. Predicted results based on the proposed methodology compare well with the field observation and demarcate the (i) good overall performance of the offshore turbines and (ii) limit state exceedance of the nearshore turbine. It is envisaged that the proposed method will be useful towards the design of monopiles-supported wind turbines in seismic areas.

Introduction

     Towards the end of 2020, the global wind energy generation capacity amounted to 733 GW, with new installations accounting for 111 GW, almost doubling that of 2019 [1]. Due to higher efficiencies and more stable wind conditions offshore than onshore sites, an increasing proportion of wind power is produced through large offshore wind farms. This demand for cost-efficient wind energy production has facilitated the industry to develop larger turbines with higher capacities. Fig. 1 presents the evolution of turbine capacity over the last 16 years, color-coded with rotor diameter. However, due to the high capital investments involved, profitability of these wind farms requires continuous operation and immediate functionality post-natural hazards. Therefore, the seismic resilience of OWTs is an important consideration in their engineering design and is, therefore, an area of active research. Readers are referred to recent publications [[2], [3], [4], [5], [6], [7], [8], [9]] on various aspects of seismic design.

Discussion and conclusions

     An existing framework for the foundation design of monopile-supported wind turbines was extended to include seismic load considerations in liquefiable and non-liquefiable soil. The framework was validated using a case study from Wind Power Kamisu, a near-shore farm that performed well during the 2011 Tohoku earthquake and subsequent tsunami.

The discussion and conclusions are presented succinctly below:

     1. A quantitative appraisal of appropriate damping, as recommended by Lombardi and Bhattacharya [58] and Adhikari and Bhattacharya [57], could vary the structural demands and therefore needs proper consideration. Back analysis of a case study from the Kamisu wind farm further indicated that structural demands could vary if suitable damping models are not selected.

      2. Definition of mechanism-specific p-y spring models for liquefiable and non-liquefiable soil is necessary as their selection could underpredict or overpredict structural response (deformation, tilt, RNA acceleration). As near-level ground conditions prevailed in this study, the authors employed hyperelastic (strain hardening) p-y curves for liquefiable soil to account for the increased resistance attributed to shear strain-induced soil dilation. During shear strain excursions, such dilative tendencies can transfer high acceleration pulses to the RNA and therefore need to be accounted for in the design stage.

Abstract

     The purpose of this paper is to study the behavior of pile-supported structures in liquefiable soils, specifically when the soil surrounding the pile transits from no-liquefaction to full-liquefaction. A series of shaking table tests were performed on four pile-supported structures subjected to different input motions and, as a result, with different times to reach full-liquefaction. The bending moment of the piles during the transient phase is compared with those predicted in pre- and post-liquefaction stages. The experimental results showed that the maximum bending moment may occur during the transient phase (i.e. during the development of excess pore pressure before the soil is fully liquefied). Arguably, the observed amplification in bending moment is caused by the the tuning effect between the predominant frequency of the input motion and the frequency of the pile-supported structure, which is progressively decreasing during the liquefaction process. Results are presented using a non-dimensional framework whose parameters are derived from the governing mechanics. A new parameter TAF (Transient Amplification Factor) is defined to predict the design bending moment during the transient phase. It is shown that the transient bending moment can be obtained from the newly introduced parameter TAF, and the values of maximum bending moments in the pre- and post-liquefaction stages. It is found that TAF is a function of two easily obtainable parameters: (a) time taken to reach full liquefaction, this can be obtained through site response analysis; (b) elongation of natural period of vibration, expressed as ratio of the time period of the structure at full liquefaction to the time period at zero-liquefaction. Finally, practical implications of the main findings are discussed.

Introduction

     The dynamic behaviour of pile-supported structures founded in liquefiable deposits is still an area of active research [[1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30]] due to the poor performance of piled foundations as observed in most of the recent earthquakes [9,31]. The current methods of pile design – see for Example codes of practices such as JRA [32,33] NEHRP [34], IS-1893 [35], focus on avoiding bending, buckling and settlement failures. The effect of dynamic soil-structure-interaction (SSI) effects is taken into account by means of empirical correlations.

     Fig. 1 shows the two main stages of loading in a typical pile-supported structure during an earthquake. Fig. 1 (a) shows the stage at the start of the shaking, before any build-up of the excess pore water pressure takes place. In this stage, piles are mostly subjected to inertia loads induced by the ground shaking and oscillation of the superstructure. It is known that inertia loads tend to generate high bending moments at relatively shallow depths along the pile. Fig. 2 (a) shows the “Beam on Non-Linear Winkler Foundation” model, in which the effect due to soil-structure interaction is modelled by a set of springs distributed at discrete locations along the length of the pile. Each spring is defined by a relationship between the soil pressure (p) and pile deflection (y) referred to as p-y curve. This method is conventionally used to compute the internal forces (i.e. bending moment and shear force) and pile’s deflection. The typical shape of p-y curves for sand in non-liquefied condition is shown in Fig. 2 (b), and further details can be found in API [36]. However, in saturated loose to medium dense sand,the ground shaking induce a gradual increase in pore water pressure, resulting in the soil to progressively lose its strength and stiffness. When the excess pore pressure equalises the overburden pressure, the soil is said to be in a full-liquefaction condition; at full liquefaction the foundation loses the support from its sorrounding soil and the piles tend to act as an unsupported column over the liquefied layer, see Refs. [37,38]. Fig. 1c shows the time history of a real earthquake together with excess pore water pressure profile, which can be either measured in experiments or computed by means of site response analyses. From the figure it can be noted that the onset of liquefaction occurs in a finite amount of time, whose duration depends mostly on the input motion characteristics and the density of the soil; the latter is typically expressed in terms of relative density. Typical values of time to reach liquefaction range between 6 and 15s, however, the actual time can be estimated by means of nonlinear finite element analysis.

Experimental results

3.1. Frequency domain analysis

     To quantify the change in vibration characteristics of the models, the experimental data was first analysed in frequency domain. Before the start of each test, the natural frequencies of the models were estimated from free vibration tests performed by exciting the structure with an impact hammer. The frequency response functions FRFs obtained from the free vibration tests are depicted in darker lines (denoted by before shaking) in Fig. 4, Fig. 5, Fig. 6. In the same figures, the power spectral density (PSD) of the input motion and FRF estimated after liquefaction is shown for comparison. It may be observed that after liquefaction the natural frequency of the models reduced significantly, which may be attributed to the development of excess pore pressure, and in some cases formation of plastic hinges in the piles. From Table 3 it is interesting to note that the four models failed in tests CH1 to CH4. (see Fig. 4 for frequency response). Other tests listed in Table 3 were carried out to investigate the sole effect of liquefaction on the variation in natural frequency of the models and transient bending moments. In these tests, formation of plastic hinges was deliberately avoided with lower pile-head mass. The results shown in Fig. 5, Fig. 6 indicate that the natural frequency of the models reduced to about half due to subsurface liquefaction. As the model transited from higher to lower frequencies it was likely that the seismic response amplified due to a temporarily matching between natural frequency of the models and predominant frequency of the input motion. This aspect is further investigated in the next section by considering the bending moment time histories along the piles.

Microbially induced calcite precipitation (MICP) harnesses the natural metabolic action of bacteria to induce the precipitation of calcium carbonate and alter soil engineering properties. This paper presents the results of using MICP to improve the cyclic resistance of Fraser River sand specimens. The formation of calcite cementation among sand particles is confirmed using scanning electron microscopic images and X-ray compositional analysis of cemented sand clusters. The results show that the velocity of a shear wave (VS) traveling through the specimen starts to increase just as the calcium solution is introduced into each specimen. Liquefaction resistance of sand samples is subsequently measured in a series of cyclic direct simple shear tests. MICP-treated samples exhibit cyclic resistances of up to 67% higher than those of the untreated sand. Post-liquefaction volumetric strain and changes in cyclic resistance in a repeated cyclic loading are also assessed and compared for the original and the treated sand specimens.

Introduction

Despite being the least abundant element in the Earth’s crust [1], carbon is commonly found on the planet’s surface as large reservoirs of organic matter or as inorganic carbon in carbonate rocks such as limestone [2]. Many organisms mediate in what is known as the “carbon cycle”, by fixing inorganic carbon to form organic carbon and re-mineralizing organic carbon back to inorganic carbon. Particularly, bacteria facilitate the deposition of carbonate minerals on the Earth’s surface by precipitating calcium carbonate (CaCO3) extracellularly as a result of their metabolic process [3–5]. In an environment with a sufficient concentration of calcium ions (Ca2þ), CaCO3 precipitation can be stimulated by a microbial metabolism that increases the pH and the concentration of carbonate ions (CO3 2 ) on the cells’ surface. These cells in turn serve as nucleation sites for the precipitated mineral. This process is generally referred to as a “microbially induced calcite precipitation (MICP)” and can occur via different metabolic processes. MICP via urea hydrolysis involves the use of the microbial enzyme urease (urea amidohydrolase; EC 3.5.1.5) to hydrolyze or break down the organic compound urea. Urease positive bacteria, such as Sporosarcina pasteurii, use urea as a source of nitrogen and energy [6] and produce the enzyme urease at different levels depending on the bacterial strain [7]. Precipitation of CaCO3 typically begins with the formation of an amorphous form of CaCO3 with low stability and high solubility, followed by a transformation into a metastable and transitional phase known as vaterite, and ending in the subsequent transformation into a more thermodynamically stable state as calcite [8]. In a soil, the precipitated calcium carbonate can cement soil particles and fill void spaces. An alkaline environment with pH ¼ 8.3 to 9.5 [6,9] is critical to trigger the hydrolysis of urea. If the pH level becomes acidic (<7.0), the precipitated CaCO3 will begin to dissolve as opposed to fostering further precipitation in the above chain of reactions. A local rise in pH may also cause the microbes themselves to serve as nucleation sites for calcite formation on the surface of bacterial cells [10].

Liquefaction potential analysis is a required task in the foundation design process of offshore wind turbine in Taiwan. The standard penetration test (SPT) is mostly used in the preliminary soil investigation of the pilot offshore wind farm in Chang-Bin, Taiwan. Due to the different experimental conditions and operating conditions, the N value (SPT-N) of SPT varies greatly. This study applies statistical methods in conjunction with the New Japan Road Association simplified-empirical method (NJRA method) to incorporate the uncertainty of SPT-N values into the offshore liquefaction potential assessment to quantify the risk of seabed liquefaction. The study statistics the field experimental geotechnical parameters SPT-N and determines the probability density function of the SPT-N distribution of each layer of soil in the offshore wind farm. In order to quantify the risk of seabed soil liquefaction potential, the Monte Carlo random sampling method is used, and by the NJRA method to carry out the seabed liquefaction potential in Changhua, Taiwan. After comparing, the results obtained by the current SPT-based soil liquefaction potential assessment by deterministic approach are conservative. In the ground investigation of offshore wind farm development, the method proposed in this paper can access the thickness of liquefiable soil layers under any given probability for optimizing offshore wind turbine foundation design.

Introduction

The Chang-Bin offshore wind farm of Taiwan Power Company is located in the offshore area of Changhua. The preliminary geotechnical investigation results show that the seabed of Chang-Bin offshore wind farm in Taiwan is mainly composed of silty sand (SM) and low plasticity clay (CL), it contains a small number of low plasticity silt (ML) and silty poorly graded sand (SP-SM) [CSC (2017), Hai-Shia Offshore Wind Energy Co., LtD. (2017a, 2017b), Chien at al. (2014), TORI (2012), TPC (2012, 2018)].. The soil within 80 meters below the seabed can be roughly divided into three layers. The uppermost layer is mainly loose to medium dense sand (SPT-N<30) and the depth is about 25 meters. The soil conditions of Taiwan’s offshore wind farm are very different from the offshore wind farm in the North Sea [Le at al. (2014)]. For a sandy soil within 20 meters of the surface depth, it is highly likely that soil liquefaction will occur when an earthquake occurs [Seed and Idriss (1971)]. The soil liquefaction potential can assess by deterministic approach in Taiwan, which the SPT-N based soil liquefaction potential semi-empirical analysis method is the common adopted in the engineering design in Taiwan [JRA (1996), CPA (2011)], However, soil liquefaction can be affected by soil layer distribution, soil properties, seismic wave transmission, stress conditions, etc. Therefore, quantifying the risk of soil liquefaction in highly seismic region is essential for the development of offshore wind farms.

1- Introduction

2- Brief review of existing methods

3- Methodological approach

4- Case study: liquefaction potential assessment

5- Conclusions

References

Abstract

This paper presents a new evolutionary neural network (ENN) algorithm coupled with the dimensionality reduction technique ‘t-distributed stochastic neighbour embedding’ (t-SNE). The ENN model features the crossbreeding of a differential evolution method and a stochastic gradient optimisation algorithm. The t-SNE is used to visualise the training and testing datasets and the ENN model performance. The proposed ENN model is applied to a relatively large soil liquefaction database. The good convergence and generalisation ability of the proposed model and the negligible misclassification results demonstrate that the proposed ENN model can provide accurate, efficient, and flexible results. The prominent and practical abilities of t-SNE to recover the structure of the initial conditions and to demonstrate the ENN model performance are discussed. This coupled approach simplifies the analysis and/or prediction of hazards for which large quantities of data are required.

Introduction

One of the focuses of civil engineers in the past three decades has been to tackle the soil liquefaction phenomenon efficiently, given the random nature of its occurrence. Soil liquefaction occurs when a saturated or partially saturated soil substantially loses strength and stiffness in response to an applied stress (such as shaking during an earthquake or other sudden changes in its stress condition), in which a material that is ordinarily a solid behaves like a liquid [1,2]. Soil liquefaction occurs when the effective stress (shear strength) of the soil is reduced to essentially zero and imposes a real risk on the surrounding structures [3–8]. The dedication of engineers and researchers has thus been triggered and further reinforced by a plethora of casualties [9] and the severity of structural damages [10–13] caused by this phenomenon. However, the prediction of soil liquefaction has remained a formidable task owing to the nonlinearity of soil behaviour [14–22] and the characteristics of the seismic power dissipation. The traditional methods for predicting soil liquefaction [23–26] still have numerous limitations, which have initiated the proposals of more powerful (efficient, flexible, and accurate) strategies. Among them, data mining (DM) has attracted particular attention recently. Data mining (DM) is a multidisciplinary subfield of computer science that includes statistics, database technology, and machine learning for the analysis of previously unknown or unsuspected relationships buried in large datasets. The DM algorithms are increasingly accepted in geotechnical engineering and have generally shown an excellent ability to solve multi-dimensional and complex problems [27].

Detailed geotechnical characterization and in-depth liquefaction assessment using seismic effective stress analyses are presented for 55 liquefaction case histories (sites) from Christchurch. Fifteen of the sites manifested liquefaction in the two major earthquakes during the 2010–2011 Canterbury earthquakes (YY-sites), while 17 sites did not manifest liquefaction in either event (NN-sites). The YY- and NN-sites are shown to have practically identical critical layer characteristics, with low CPT tip resistance (qc1Ncs = 80–86), and shallow location of the critica.l layer at approximately 2 m depth. However, there are significant differences between the YY- and NNsites with regard to their deposit characteristics including the thickness and vertical continuity of their critical zones and liquefiable materials. Effective stress analyses are used to demonstrate key mechanisms of system-response of liquefying deposits that either intensify (for the YY-sites) or mitigate (for the NN-sites) liquefaction manifestation at the ground surface. The study illustrates the need to consider system-response of liquefying soils in the assessment of liquefaction manifestation and severity of liquefaction-induced damage.

Introduction

Simplified liquefaction triggering procedures used in current engineering practice have been developed based on case histories in which liquefaction was manifested at the ground surface during past earthquakes. Such liquefaction case histories essentially reflect the overall response of soil deposits during earthquakes, and their key characteristics with regard to the severity of liquefaction manifestation. However, despite the intent to capture the overall performance of the deposit at a given site, in the simplified liquefaction evaluation procedures each layer is considered in isolation, and a factor of safety against liquefaction triggering, maximum shear and volumetric strains are estimated separately, and independently, for each layer. In these calculations, interactions between different layers in the dynamic response, and through excess pore water pressures and water flow are ignored. Hence, principal mechanisms of interaction or system-response effects of liquefying deposits that potentially contribute to the severity of liquefaction manifestation are not accounted for in the simplified procedures. Liquefaction damage indices, such as LSN [22] and LPI [14,15] use specific weighting functions to quantify the damage potential of liquefying layers depending on their proximity to the ground surface, but still they do not account for cross-interactions between different layers during the development of liquefaction and post-liquefaction triggering.

Earthquake induced foundation failure has occurred in a number of events worldwide, such as the 1964 Niigata earthquake and 1995 Kobe earthquake. Fabric anisotropy has been considered as one of the factors that affect liquefaction characteristics. In this study, dynamic centrifuge tests were conducted for studying the influence of sand fabric anisotropy on seismic response of circular foundations. For this purpose, sub-angular Toyoura sand was used as the most anisotropic material, while sub-rounded Nevada sand was used for the comparison tests. Each group of tests were carried out in both dry and saturated conditions. These tests are considered complementary to similar centrifuge tests that the authors performed using rectangular foundations. The tests show that sand fabric anisotropy has significant impact on the seismic response of circular foundations as well, especially in saturated conditions. Specifically, as the deposition angle of the sand layer increased, the data show a remarkable decrease in acceleration amplitude concurrently with a considerable decrease of pore pressure ratio and foundation settlement. These effects were more pronounced in the sub-angular Toyoura sand. These effects of increased deposition angle are consistent with the corresponding decrease of bearing capacity and increase of footing settlements in static centrifuge tests from the literature.

Introduction

In 2015 alone, magnitude seven and greater earthquakes happened 19 times in the world, including two major ones in Nepal. The April 2015 Nepal earthquake (Gorkha earthquake) killed over 8000 people and injured more than 21,000. Liquefaction is a phenomenon that occurs both in natural and man-made events, in which the strength and stiffness of a soil are significantly reduced. It is typically induced by earthquakes or other rapid loading processes. Foundation failure due to soil liquefaction during earthquakes has been and continues to be a major type of damages. Soil anisotropy is widely observed in the natural deposited sand. It is determined by particle shape, contact, particle roughness, deposition history and many other more. It affects a few important parameters in soil properties including shear modulus, friction angle and bearing capacity. In previous studies, researchers found that fabric anisotropy of soil grains is an important factor in dynamic soil response as well. Brewer [1] introduced the definition of anisotropy as the property of being directionally dependent, and in most cases, the result of weathering of hard rock or sedimentary solids. Several researches were conducted using true triaxial tests, nonlinear models and their applications in 1970s. Saada and his colleges at Case Western Reserve University investigated the mechanical behavior of anisotropic materials. They used one-dimensional consolidation of the test specimens in a tri-axial cell to simulate field condition [2]. Oda stated that sands deposited in air or in water show anisotropic shear strength in their triaxial compression tests [3].

The influence of polymeric structure of polycarboxylate-ether (PCE) based superplasticizers on the intercalation behavior in sodium montmorillonite clay (Na-MNT) is investigated by performing in-situ X-ray diffraction (XRPD) on fresh, unaltered clay pastes. The use of this technique reveals the real influence of the PCE structure and of the PCE/clay dosage ratio on the expansion profile of the clay. This is not observed with the traditional XRPD methodology performed on dried clay pastes, which shows the same values of d-spacing despite using polymers of diverse structures. It is observed that PCE polymers with long side chains and high side chain density result in larger expansion. Additionally, polymers with a high anionic charge saturate the interlaminar space of montmorillonite at a lower dosage. The experimental results also indicate that clay exfoliation is critical in the intercalation process and the exfoliation tendency of the clay is influenced by the structure of PCE polymers.

Introduction

The dispersing capacity and water reduction efficiency of polycarboxylate-ether (PCE) based superplasticizers is severally affected by the presence of montmorillonite clays (MNT) in the sands used for concrete production. This type of clays has the ability to absorb large quantities of PCE polymers through an intercalation mechanism of the PEG/PEO (poly-ethylene glycol/polyethylene oxide) side chains of the polymer [1,2], which is responsible for the partial or total loss of their dispersing capacity. The absorption behaviour of PCE-based superplasticizers on MNT clay has already been investigated by different authors using XRPD analysis on centrifuged and dried clay pastes [1,2]. These studies show almost no influence on the d-spacing of MNT clay when different structures of PCE polymers are used since the maximum clay expansion value always remains in the range of 18–21 Å and, in all cases, the average differences of d001 values within PCE polymers of diverse structures does not exceed 3 Å (the distance equivalent to one water molecule arranged in the interlayer of MNT clays).

In risk assessment of spatially distributed infrastructure, the probability of demand exceeding capacity is evaluated across the system. We describe and compare two levee system reliability analysis frameworks for seismic and high-water demands. The first approach is general, but computationally intensive, and uses Monte Carlo simulations to model capacity and demand for “segments” (i.e., elemental levee lengths) as spatially correlated random variables. We apply a capacity correlation model derived from seismic case histories in Japan. The seismic demand correlation model is based on global ground motion data, whereas the high-water correlation is taken as unity. The second approach achieves computational efficiency by grouping segments into physics-based “reaches” (i.e., length of levee having uniform statistical distributions of capacity and demand). Statistics and spatial correlation of the limit state function are computed using a procedure based on the firstorder reliability method. The probability of failure of the reach is then computed using level-crossing statistics. The application of level crossing statistics required an adjustment, introduced here, to previously utilized capacity correlation functions. We apply both methods for a levee system subjected to realistic demand and capacity distributions and show that characteristic lengths (defined as lengths of levee that can be considered as statistically independent) are comparable for high-water and seismic demands. This outcome is specific to the considered failure mechanisms and is driven by use of similar capacity correlation models, whereas differences in demand correlation models have limited impact.

Introduction

Levees are defined as man-made or natural embankments along rivers or water bodies. Their primary purpose is to provide protection against high-water events. The performance of levees when subjected to high-water or earthquakes is essential for the resilience of surrounding communities. Despite their critical function, many levees were not engineered at the time of their construction and are often founded on soft and weak soils. As a result, levees are frequently damaged during high-water events (water level rise in the river channel; e.g., [1–3]) and following major earthquakes (e.g., [4–8]). For levees that continuously impound water, a single failure anywhere along their length will produce flooding, and hence comprises system failure. For levees that intermittently impound water, the seismic failure probability is related to the combination of seismic deformation potential and probability of high-water during or shortly following the event, whereas the high-water failure probability is simply the single-segment failure probability during a high-water event. In either case (continuously or intermittently loaded), levees constitute spatially distributed series systems, which present particular challenges for reliability assessment. This paper describes two conceptually similar approaches for analysis of levee reliability, with an emphasis on the system probability of failure given knowledge of capacity and demand on a more local level. We defer to other documents for recommended analysis procedures for computing capacity at the segment, or crosssection level (Zimmaro et al. [9] for seismic, URS Corporation, Jack R. Benjamin & Associates Inc. [10] for high-water).

Correlations between the liquefaction resistance of sands in spreader dumps of lignite opencast mines on one side and the CPT tip resistance or the shear wave velocity on the other side have been developed based on an extensive experimental study on 10 different sands. The cyclic resistance ratio CRR N( 10) f = for a failure in 10 cycles, defined as the liquefaction resistance herein, was determined from undrained cyclic triaxial tests for all materials. The CPT tip resistance qc was measured in soundings performed in a calibration chamber. The shear wave velocity vS was obtained from measurements in a triaxial cell by means of bender elements. All three quantities CRR N( 10) f = , qc and vS were determined for different relative densities Dr. In case of qc and vS also the pressure-dependence was examined in order to derive suitable equations for a pressure-normalization being necessary for the interpretation of field test data, giving values qc1 and vS1 for a mean effective stress of p = 100 kPa. The procedure for the development of the CRR N( 10) f = -qc1 and CRR N( 10) f = -vS1 correlations based on the experimental data is explained. The correlations derived for the various sands in the present study are analyzed with respect to the influence of the grain size distribution curve and compared to respective relationships for natural soils or uncompacted recent artificial fills from the literature.

Introduction

The liquefaction resistance of a ground is often evaluated based on correlations with the CPT or SPT penetration resistance [5,6,10,19,23,24,27,28,30–33] or with the shear wave velocity [4,10,11,20,29,35]. Examples for respective correlation diagrams are shown in Figs. 1–3. They are showing a cyclic resistance ratio, defined as the shear stress amplitude ampl causing a liquefaction in a certain number of cycles divided by the vertical effective stress v on the ordinate and the penetration resistance or shear wave velocity on the abscissa. Such diagrams have been developed by collecting field data from CPT or SPT (tip resistance qc in case of CPT, blow count NSPT in case of SPT) or shear wave velocity measurements (vS). Furthermore, one needs information regarding the cyclic stress ratio CSR induced in the ground by an earthquake and whether this action has lead to a liquefaction or not. Combinations of CSR with qc, NSPT or vS having lead to liquefaction are usually marked by filled symbols (see Figs. 1–3), while the symbols for combinations without any visible signs of liquefaction remain empty. Finally, a bounding curve is drawn, separating the “liquefaction” from the “no liquefaction” cases. These bounding curves sometimes consider further parameters like fines content or mean grain size. They are usually established for a certain earthquake magnitude MW . Curves for other magnitudes can be obtained by multiplication with a correction factor MSF . However, almost all such correlations proposed in the literature are valid for natural soils only (see Figs. 1 and 2). A first CSR-vS correlation diagram for uncompacted recent artificial fills has been recently proposed by Dobry et al. [11] (Fig. 3). No correlations for an estimation of the liquefaction resistance of such fills based on CPT or SPT field data exist so far.

This paper summarizes the results from a comprehensive laboratory experimental research program conducted at the University of British Columbia, Canada to study the mechanical behavior of natural fine-grained soils. Constant-volume direct simple shear tests were performed on natural silt samples collected from various locations in British Columbia to investigate the monotonic and cyclic shear loading response of these materials. Influencing factors such as effective confining stress, over-consolidation ratio, initial static shear bias, soil plasticity, and soil fabric/micro-structure on the mechanical behavior of silts were systematically investigated using a series of controlled laboratory tests. The key observations arising from these examinations are presented and discussed to serve as input for the development and/or refinement of design practice approaches for seismic geotechnical design. New directions to be considered to advance the current state of understanding and knowledge on the mechanical behavior of natural fine-grained soils are identified.

Introduction

Liquefaction of soils and associated ground movements could cause damage to buildings and infrastructure during earthquakes. For example, it has been reported that the damage costs from a major earthquake in the Metro Vancouver Region of British Columbia (BC), Canada could be as high as $75 billion [1], with significant damage expected to arise from soil liquefaction. Southern Ontario and the St. Lawrence Valley are the other regions in Canada where earthquakeinduced soil liquefaction is considered an engineering concern. The assessment of liquefaction susceptibility of natural fine-grained soils, particularly those with low plasticity, poses significant challenges to the current design practice. This is mainly due to the lack of understanding of the complex stress-strain-strength behavior of finegrained soils such as low-plastic silt; this knowledge gap, in turn, has led to the use of liquefaction assessment criteria that are primarily based on soil index properties such as plastic limit (PL), liquid limit (LL), plasticity index (PI), and water content. The consideration of silt with PI < 7 as “sand-like” and that with PI ≥ 7 as “clay-like” in behavior for the purpose of liquefaction assessment by Idriss and Boulanger [2] is one such example in this regard; the approach proposed by Bray and Sancio [3], again based on index parameters, serves as another example. Fine-grained silty soils with high levels of saturation are commonly found in natural river deposits, and due to many reasons, most population centers, along with critical infrastructure, industrial plants, etc., are located near river banks on such soil deposits.

The undrained shear strength and liquefaction cyclic resistance of silty sands are examined based on a large number of laboratory triaxial test results. The influence of saturation on the liquefaction triggering and occurrence of liquefaction-induced flow slides is highlighted. The laboratory triaxial tests are conducted separately in the three phases of full saturation, partial saturation and unsaturation. The past studies of the authors’ are first reviewed and some new data are added where appropriate. The different responses of silty sands in those three phases of saturation are discussed in detail.

Introduction

The liquefaction resistance and undrained shear strength of soils are the key parameters to determine the liquefaction triggering and liquefaction-induced flow slides of earth structures during earthquakes. This paper is aimed at highlighting the influence of saturation on the liquefaction resistance and undrained shear strength of silty sands, based on laboratory triaxial test results. In this paper, the phases of saturation in soils are separated into three categories, full saturation, partial saturation and unsaturation. It is known from the observation of field velocity logging tests that a velocity of propagation of primary wave of approximately 1600 m/s is observed in soil layers far below a groundwater level, indicating that these soil layers are fully saturated. However, it is also known that there is a soil layer of 3–5 m deep immediately below a groundwater level that typically exhibits a velocity of propagation of primary wave of 500–1000 m/s, implying that this soil layer is partially saturated with pore water containing some minute air bubbles. On the other hand, there is a soil layer immediately above a groundwater level, where capillary water rises up through soil aggregates and a negative pore water pressure develops relative to an atmospheric air pressure, due to surface tension of pore water, leading to development of capillary suction within soil aggregates. Fig. 1 illustrates conceptually the phase transformation of saturation in soils in a diagram of pore air and pore water pressures, ua & uw, against confining stress σ, based on the work of Fredlund and Rahardjo [1]. When soils are unsaturated, though nearly saturated under an atmospheric pressure, there would be some continuous air phases within soil aggregates, where the matric suction of su = ua – uw would be present due to surface tension of pore water within soil aggregates. The surface tension tends to interact with soil aggregates and to produce bonds of soil structures, and therefore is likely to mobilise shear strength of soils.