Developments in Earthquake Geotechnics

The paper gives an overview of performance-based approach for designing geotechnical structures against earthquakes, including remediation of liquefiable soils. Case histories of implementation and performance of remediation measures during past earthquakes are reviewed. The paper discusses the applicability and limitations of the conventional simplified approach for designing remediation of liquefiable soils and how these limitations can be overcome in the performance-based approach that explicitly considers residual displacements and structural strains beyond elastic limit. The paper also presents recent developments in performance evaluation against extremely long duration earthquake motions such as those during the 2011 East Japan earthquake (M = 9.0).

The evolution of dynamic response analysis and its role in geotechnical earthquake engineering primarily in North American practice is traced from 1964–2015: from the elementary visco-elastic analyses of site response of the 1960s to the more complex methods of analysis in use today that are based on sophisticated plasticity constitutive models. This review is limited to North American practice because the author is most familiar with this practice, having served it as researcher, professor and professional engineer over the past 50 years. A selection of case histories is presented below that demonstrate the wide-ranging capability of geotechnical engineers today of solving critical and important types of problems in practice since the widespread availability of reliable nonlinear methods of analysis.

Liquefaction of soils has been widely recognized as an important cause of damage in many past earthquakes. Although recognized and named in the early 1950s (Mogami T, Kubo K, The behavior of soil during vibration. In: Proceedings of the 3rd international conference on soil mechanics and foundation engineering, vol 1, Zurich, pp 152–155, 1953; Terzaghi K, Variety of submarine slope failures. In: Proceedings of the 8th Texas conference on soils and foundation engineering, University of Texas, Austin, pp 1–41, 1956), liquefaction sprang to the attention of the geotechnical engineering profession in 1964 following large earthquakes in Niigata, Japan and Alaska, USA. Since that time, a great deal of research on soil liquefaction has been performed, particularly in Japan and the United states but also in other seismically active countries such as Canada, Chile, New Zealand, Taiwan, and Turkey. This research has led to breakthroughs in understanding of the basic mechanics of liquefiable soils, the development of practical, empirical procedures for evaluation of liquefaction potential, and the development of numerical procedures for site-specific analysis of liquefaction and its effects. This paper presents a brief and incomplete review of the history of liquefaction hazard evaluation, assesses its current status, and discusses future developments in this area.

Within the limited scope of the author’s experiences on countermeasures against liquefaction, lattice walls by in-situ cement-mixing, cut-off sheet pile walls and dewatering are picked up. Technical issues on the effectiveness of these countermeasures and their evaluation procedures are discussed based on results from relevant 1-g and centrifuge model tests, numerical analyses and triaxial tests.

Frequently, deep foundations extend through potentially liquefiable sand layers near the ground surface and bear on more competent layers at depth. When liquefaction occurs, the skin friction in the liquefied layer would be expected to decrease to some negligible value, but as the liquefiable layer settles, negative skin friction could potentially develop around the pile in this layer as effective stress increases. To investigate the loss of skin friction and the development of negative skin friction, axial load tests were performed on instrumented full-scale piles before and after blast-induced liquefaction at sites in Vancouver, Canada and Christchurch New Zealand. Following blasting, liquefaction developed within sand layers resulting in significant settlement. Skin friction in the liquefied layer initially dropped to essentially zero. However, as the liquefied sand reconsolidated, negative skin friction became equal to about 50% of the pre-blast positive skin friction. Despite significant ground settlement, pile settlement was relatively small. A neutral plane approach for computing pile settlement resulting from negative skin friction provided reasonable agreement with observed behavior.

One of the greatest challenges facing civil engineers is the stewardship of ageing infrastructure. At present, it is difficult to inspect the condition of infrastructure while it is being used. The fragility of old infrastructure presents a challenge for new construction in congested urban environments. Little is also known of the long-term performance of such infrastructure. We see many cases around the world of continuous retrofit/renovation of infrastructure during its lifetime. The vision discussed in this chapter is to significantly improve our capability in predicting and managing the life expectancy of large infrastructure. For future proofing, active monitoring of construction and operational processes of civil engineering infrastructure is essential. This implies that structures are instrumented to assess their performance against engineering design parameters or predictive models. In recent years, sensor and communications research has been undergoing a revolution. There are possibilities to use emerging sensor technologies (distributed fiber optics sensing, wireless sensor network, low power miniature sensors, energy harvesting for continuous monitoring, robotic inspections, satellite images, crowd source data, etc.) to address the particular needs to look after our infrastructure.

This paper presents the recent efforts in Indonesia to mitigate the impacts of earthquake hazards. The actions includes as follows: updating of the seismic hazard maps of Indonesia 2010 and 2016; revision and continuous updating of building and infrastructure design codes; development of microzonation maps for big cities in Indonesia; development of academic draft of Indonesian Earthquake Master Plan; development of design guidelines for tsunami vertical evacuation; development of a national design code for geotechnics and earthquake; and establishment of the national center for earthquake studies. Revision of seismic hazard maps for Indonesia 2010 has been developed based upon updated: seismotectonic data, fault models, and GMPEs up to 2010. The updating of infrastructure design codes related to earthquakes activities are being performed for buildings, bridges, dams, harbors and others special structures. The development of microzonation map of seismic risk has been initiated for Jakarta city. The academic draft for Indonesian earthquake master plan has been prepared based on assessment from several aspects, i.e. basic sciences, engineering and risk analysis, and social and legal aspects. The guidelines for tsunami evacuation had been finished and submitted to BNPB in 2013. Lastly, the Indonesia National Research Center for Earthquake has been initiated in June 2016.

In this chapter, the authors describe the historical evidence of how the ground conditions in Tokyo have changed over the past 150 years and those changes would influence the area if a mega-earthquake hits Tokyo in the near future. In the recent 150 years, Tokyo was developed as a highly industrialized city protected from flooding by dykes, flood-gates and reclaimed land which have been built in progressive manners. Water needed in industrialized low-land area of Tokyo was partly obtained by deep-well pumping of the groundwater during a period of 100 years from 1873 to 1973. The excessive groundwater withdrawal resulted in lowered groundwater level down to a maximum depth of about 60 m below sea water level, which led to serious land subsidence in the low-land area particularly during a period of 1960–1973 when the withdrawal was extremely heavy. The latest big earthquake struck Tokyo area in 1923 when the altitude of Tokyo low-land area was still higher than the sea water level. Since then Tokyo low-land area has experienced the following three major changes: (i) land subsidence (max: about 4.5 m) due to the excessive groundwater pumping, (ii) construction of extremely extensive underground networks of lifelines, railways, roads and shopping areas, and (iii) increasing danger of failure of the riverine levees due to the recent increase in the torrential rainfalls typically accompanied by typhoons that strike Japan several times a year because of increasing typhoon-activation power supplied from the sea water 1 or 2° (Celsius) warmer than before. These changes made the low-land area of Tokyo much more susceptible to flooding than the time of 1923. In case that the sea water of Tokyo Bay flows onto the low-land area of Tokyo through possible breakage of some part of the existing seawalls and/or the flood gates, the depth of the water in major part of the area is expected to reach about 4 m at the deepest area and about 1.5 m on average. This may cause unacceptably serious situation in which huge number of people will suffer from the submersion of subways when the sea water flows into the underground facilities. About two million local residents in the low-land area submerged by flood water need to move to higher areas because almost all lifeline systems become out of service in the low-land area. Many of submerged underground facilities will not work anymore even after drying off and therefore will have to be replaced by new ones. About ten million people currently using subways in their daily commuting, business trips etc. every day will have practically no alternative means of transport. In the worst scenario, this will give extremely serious effects on enormous numbers of economic activities not only in Tokyo but also in the entire Japan.

Recently a number of macro-element models have been formulated for assessing the performance of shallow foundations during earthquake loading. These provide a computational tool that represents the nonlinear dynamic behavior of the foundation in a manner much simpler than finite element modelling; consequently, they are useful for preliminary design. The basis of this chapter is the shallow foundation moment-rotation pushover curve, which is bracketed by the rotational stiffness at small deformations, determined by the small strain stiffness of the soil, and the moment capacity, which is a function of the soil shear strength and the vertical load carried by the foundation. Between these two limits there is a curved transition. The paper argues that when the vertical load carried by an embedded foundation is a small fraction of the vertical bearing strength, the moment-rotation behavior dominates the response. This means that the structure-foundation system can be reduced to a single degree of freedom (SDOF) model.

The form of the shallow foundation moment-rotation curve obtained from experimental and computational modelling is approximately hyperbolic; the nonlinear shape is due in part to the nonlinear deformation of the soil beneath the foundation but also to gradual loss of contact between the underside of the foundation and the soil below. The paper proposes a generalization of the pushover curve to give a shallow foundation cyclic moment-rotation relationship. The hysteretic damping properties of the model, as a function of the foundation rotation amplitude, are demonstrated as is the relation between secant stiffness and foundation rotation.

This chapter shows how the model can be applied in numerical simulation of structure-foundation systems subject to earthquake time histories. The significance of the maximum displacement (foundation rotation) in relation to the damping and residual rotation at the end of the earthquake record are discussed.

Seismic ground motions and behavior of excess pore water pressures on a reclaimed land during the 2011 off Pacific coast of Tohoku earthquake are studied numerically by using a computer program, “FLIP ROSE” (Iai et al, Soils Found 32(2):1–15, 1992; Int J Numer Anal Meth Geomech 35:360–392, 2011). Long duration time of 30 min which includes both the mainshock and the aftershock is taken into account in the dynamic analysis considering build-up and dissipation of excess pore pressures as well. The site studied is located in Tokyo port, Japan, where seismic ground motions are obtained with vertically arrayed seismographs by Tokyo metropolitan government. It is shown that observed peak ground accelerations at each depth of ground are reasonably simulated. The ratio of excess pore water pressure is calculated as maximum of about 0.4 during aftershock, agreeing with the fact of no significant liquefaction at the site. It is noted that significant rise of excess pore water pressure is calculated in the aftershock, suggesting significant effect on liquefaction.

A series of effective stress dynamic analyses considering permeability was performed in areas with and without liquefaction damage at reclaimed land in Urayasu city in the Tokyo Bay area due to the 2011 Tohoku Earthquake off the Pacific coast. Liquefaction damage was significant because the ground shook with earthquake motions for an extremely long duration (more than 2 min for the main shock) followed by an aftershock about 30 min later. In particular, the delayed sand boil, which occurred much later than the main shock, was monitored by a security camera at the site. In typical earthquake motions with a shorter duration, the permeability of the ground is ignored in the seismic response analyses. However, whether the effect of permeability can be ignored becomes questionable when the duration of earthquake motion is extremely long and additional effects occur due to an aftershock such as the case with the 2011 earthquake. The results of the effective stress analyses incorporating the effect of permeability are consistent with the behavior of the ground with and without liquefaction damage. In particular, the delayed sand boiling at the liquefied site is well simulated by the effective stress analysis for a specific combination of permeability assigned above and below the groundwater table, indicating that it is important to carefully evaluate the permeability in the analysis of liquefaction.

Effective stress analysis of the damage to river embankments during the 2011 Great East Japan earthquake (M = 9.0) is performed to study the applicability of the analysis method. The multi-spring model and the cocktail glass model are used for undrained and partially drained analysis, respectively. Both are defined in the framework of the strain space multiple mechanism model. The earthquake motion in this particular earthquake lasted more than 2 min, which is much longer than typical strong earthquake motions. Three damaged river embankments, 4–8 m high, constructed on liquefiable foundation soil and subjected to bedrock peak ground motions ranging from 0.15 to 0.6 g with and without the effect of an aftershock, are studied. The settlements of the embankment range from 0.8 to 4.3 m with and without differential settlements. The analysis results are mostly consistent with the damage to the river embankments both in terms of settlements and deformation modes. In particular, the partial drainage analysis using the cocktail glass model is applicable to the behavior of a river embankment subject to a seismic motion with a long duration, including the additional effect of the aftershock.

Numerous quay walls were damaged due to ground motions and subsequent tsunami during the 2011 Great East Japan Earthquake off the Pacific Coast of Tohoku, especially in the area south of the epicenter. Although it is assumed that liquefaction due to the ground motions occurring before the tsunami magnifies the tsunami damage, the severity of liquefaction and related damage to the quay walls before the tsunami is unknown. Various numerical simulations are performed for understanding the damage mechanisms caused by the ground motions using a strain space multiple mechanism model, called the cocktail glass model, as a constitutive model for soils. If a sheet pile type quay wall can retain structural stability and the excess pore water pressure dissipates before the arrival of a tsunami, the quay wall will not collapse due to a tsunami attack. However, a caisson-type quay wall, even if the quay wall retains its structural stability, may deform due to rubble foundation scouring caused by a tsunami attack. In addition, the cocktail glass model, which can consider the effect of pore water migration and dissipation, is applicable to evaluate the seismic performance such as deformation of quay walls against strong, long-duration ground motions.

The end bearing capacity of a pile is modeled for two-dimensional effective stress analysis by initially conducting a vertical loading test in a centrifugal field and then simulating the vertical loading test by three-dimensional analysis to validate the three-dimensional analysis. Finally, a model, which uses nonlinear spring elements for two-dimensional analysis, is proposed and validated via case studies. In the proposed model, the nonlinear spring elements are characterized by the hyperbolic relationship determined in the three-dimensional analysis.

The strain space multiple mechanism model, which was originally developed for the cyclic behavior of granular materials such as sand, is adapted to idealize the stress–strain behavior of clay under monotonic and cyclic loads. Compared to the conventional elasto-plastic models of the Cam-clay type, advantages of the proposed model include (1) the arbitrary initial K₀ state can be analyzed by static gravity analysis, (2) the stress-induced anisotropy (i.e., the effect of initial shear) in the steady (critical) state can be analyzed based on Shibata’s dilatancy model (Ann Disaster Prev Res Inst Kyoto Univ 6:128–134, 1963), (3) over-consolidated clay can be analyzed by defining the dilatancy at the steady state based on the over-consolidation ratio, and (4) the strain-rate effects for monotonic and cyclic shears can be analyzed based on the Isotach/TESRA model proposed by Tatsuoka et al. (Soils Found 42(2):103–129, 2002) in a strain rate ranging from zero to infinity as well as by the conventional strain-rate effects of the secondary consolidation (creep) type. Simulations of the drained/undrained behaviors of clay under monotonic and cyclic loadings are used to demonstrate the performance of the proposed model.

A new constitutive model (cookie model) for clay was proposed to establish a prediction method to evaluate damages from complex disasters. The model was developed as an extension of a model (cocktail glass model) for liquefiable sandy soil. The models were installed on a two-dimensional dynamic effective stress analysis program. The program was used to apply a constitutive model (cookie model) to a clay layer under a highway embankment as an example. We could perform the self-weight analysis, the long-term consolidation settlement analysis continuously, and seismic response analysis within process of the consolidation settlement analysis. The results analytically indicated that earthquakes caused additional instantaneous settlement, additional long-term settlement, and additional increase in pore water pressure of the clay layer. The settlement advanced rapidly during the consolidation process, and rising of pore water pressure was observed for the embankment. These observational results could be analytically reproduced by applying seismic waves during the consolidation settlement analysis.

This chapter presents the finite strain formulation of a strain space multiple mechanism model for granular materials. Since the strain space multiple mechanism model has an appropriate micromechanical background in which the branch and complementary vectors are defined in the material (or referential) coordinate, the finite strain formulation is carried out by following the change in these vectors, both in direction and magnitude, associated with deformation in the material. By applying the methodology for compressible materials established in the finite strain continuum mechanics, decoupled formulation that decomposes the kinematic mechanisms into volumetric and isochoric components is adopted. Material (Lagrangian) description of the integrated form is given by a relation between the second Piola-Kirchhoff effective stress and the Green-Lagrange strain tensors; spatial (Eulerian) description by a relation between the Cauchy effective stress and the Euler-Almansi strain tensors. Material description of the incremental form is derived through the material time derivative of the integrated form. The counterpart in the spatial description is derived through the Lie time derivative, given as a relation between the Oldroyd stress rate of Kirchhoff stress and the rate of deformation tensor.

To accurately estimate the damage of soil-structure systems during earthquakes, reliable analytical methods and appropriate modeling of soils and structures are necessary. Herein seismic response analyses are performed on an embankment, a caisson-type composite breakwater, and a caisson-type quay wall to verify the applicability of the strain space multiple mechanism model, where both the total and updated Lagrangian (TL/UL) formulations are introduced based on the large deformation (finite strain) theory. Both infinitesimal and large deformation analyses are performed to examine the effect of geometrical nonlinearity. All of the computed results (e.g., deformation and excess pore pressure ratio) indicate that the TL and UL formulations are theoretically and numerically equivalent, validating the computer program for large deformation analysis. In addition, the large deformation analyses decrease the amount of deformations compared to the infinitesimal deformation analysis in these three cases. This tendency becomes more significant as the amplitude of the input ground motions increases. We also perform a centrifuge model test to investigate the large deformation mechanism of liquefiable sloping ground and compare the experimental results with numerical simulations. The comparison indicates that the large deformation analysis by the UL formulation reasonably simulates the experiment, whereas the infinitesimal deformation analysis overestimates the deformation.