Earthquakes and soil liquefaction in flood stories of the ancient Near East

Flood stories in the Hebrew Bible and the Koran appear to be derived from earlier flood stories like those in the Gilgamesh Epic and still earlier in the Atrahasis. All would have their source from floods of the Tigris and Euphrates rivers.

The Gilgamesh Epic magnifies the catastrophe by having the flood begin with winds, lightning, and a shattering of the earth, or earthquake. Elsewhere in Gilgamesh, an earthquake can be shown to have produced pits and chasms along with gushing of water. It is commonly observed that earthquake shaking causes water to gush from the ground and leaves pits and open fissures. The process is known as soil liquefaction. Earthquake is also a possible explanation for the verse “all the fountains of the great deep (were) broken up” that began the Flood in Genesis. Traditionally, the “great deep” was the ocean bottom. A more recent translation substitutes “burst” for “broken up” in describing the fountains, suggesting that they erupted at the ground surface and were caused by an earthquake with soil liquefaction. Another relation between soil liquefaction and the Flood is found in the Koran where the Flood starts when “water gushed forth from the oven”. Soil liquefaction observed erupting preferentially into houses during an earthquake provides a logical interpretation if the oven is seen as a tiny house. A case can be made that earthquakes with soil liquefaction are embedded in all of these flood stories.     

Lithomorphic stresses and cleavage of loess

Lateral stresses existing in soil deposits are important clues to recent stress history because they tend to remain “locked in” following adjustments in the overburden load: Thus a high lateral stress is partly retained following a reduction in vertical stress, as by glacial melting or erosion. Conversely, during loess deposition a low lateral stress should be retained if the soil structure is preserved under increasing vertical stress. This has not previously been verified in the field because of the difficulty of measuring lateral stresses in situ in an hydrocollapsible material.

Tests were conducted at two sites in Missouri river valley loess using theK_o Stepped Blade, which uses data extrapolation to give stress on a zero thickness blade. In basal loess where the moisture content was at or above the liquid limit,K_o = 1.0 ± 0.02. In the upper to middle loess zoneK_o is as low as 0.2 to 0.4, lower than the values of 0.4 to 0.5 that would be expected if the soil were normally consolidated. Near the ground surface, lateral stresses are higher soK_o > 1, attributed to post-depositional weathering and the presence of smectite clay minerals. The vertical sequence of lateral stress from high to low to high again should contribute a tension-induced cleavage if lateral confinement is removed by excavation or erosion. This stress condition is most pronounced in loess close to its source, where the potential tension zone was found to extend to a depth of about 7 m compared to 4 m at the other site. Loess cleavage appears to result from stress relief and is not a unique directional property of this material.     

Crack damage stress as a composite function of porosity and elastic matrix stiffness in dolomites and limestones

Twenty-five uniaxial compression tests were performed to determine stress at onset of dilation, referred to herein as “the crack damage stress,” in heterogeneous dolomites and limestones. A simplified model for crack damage stress (σ_cd) is developed here using porosity, elastic modulus, Poisson's ratio and three empirical coefficients. The model shows that when porosity decreases and elastic modulus increases, σ_cd rapidly increases and approaches its maximum value. On the other hand, when porosity increases and elastic modulus decreases, σ_cd rapidly decreases and approaches its minimum value. The proposed model is validated for six heterogeneous limestone and dolomite formations which are widely distributed in Israel.     

Liquefaction-induced structures in Quaternary alluvial gravels and gravelly sediments, NE Brazil

We have identified numerous well-preserved elutriation and fluidization structures probably induced by liquefaction in Quaternary gravels and gravelly sediments of braided fluvial channel deposits in the Rio Grande do Norte and Ceará states, northeastern Brazil. They show evidence of upward-directed water escape after sediment deposition and before sediment compaction. Among the several types of structures observed, the most frequent are pillars, pockets and dikes. These structures range in width from a few centimeters to as much as 4 m, and in height from 60 cm to 4 m. Dikes, pillars and pockets are systematically associated. Clastic dikes vented large quantities of sand to the upper layers or the surface; pebbles and cobbles from the host rock sank into the dikes and formed pillars and pockets. Pockets form the root part; pillars form the intermediate part and dike, the upper part of the composite structure. The morphology of the structures in sectional and plan views indicates a 3D geometry composed of a tabular dike and pillar that present a downward continuous transition to a bowl-shaped pocket. This “stratigraphy” of liquefaction features is different from that usually presented in the current literature.

Field data suggest that both the location and the geometry of the features were controlled by sedimentary properties rather than joints and small faults. The size and abundance of these features suggest that they were formed by great events rather than localized mechanisms. Field evidence also indicates that these features are the product of fluidization and elutriation and may have been induced by liquefaction processes associated with seismic shaking. A nonseismic origin related to elutriation processes, however, cannot be ruled out for some of the features.     

基于耗散能量的饱和黄土动孔压模型

通过一系列不排水动三轴试验探究了饱和黄土振动液化过程中孔隙水压力和累积耗散能量的演化模式,并讨论了围压、动应力幅值和固结应力比对其演化过程的影响。结果表明：饱和黄土的孔隙水压力和耗散能量随着循环荷载作用逐渐累积。固结围压抑制孔隙水压力增长而消耗更多能量;更大的动应力幅值使得孔隙水压力增长更快而消耗能量更少;等压固结下,孔隙水压力增长至围压从而触发初始液化,而偏压固结下,通常先达到振动液化应变标准而孔隙水压力并没有增长至围压水平,并且固结应力比越大,液化时孔隙水压力越小,消耗能量也更少。归一化孔隙水压力u/σ₀’与累积耗散能量W/W_f之间关系受围压、循环应力比和固结应力比影响较小,可统一用双曲线模型表示。     

Simulation analysis of ground liquefaction induced by earthquake

A dynamic effective stress analysis with the finite element method has long been recommended to predict the liquefaction phenomena of sandy soil by authors and Zienkiewicz et al. as well as the similar approaches by the others. Our approach of the analysis is summarized in the first.

Until recently, however, these approaches has not commonly been used as the means of design, although its capability of prediction is appreciated by geotechnicians. This method has been neglected because of the lack of verification studies of soil models of sand and mathematical formulation for boundary problems of liquefaction phenomena. Therefore the verification of the numerical method to evaluate liquefaction potential are urgent requirement of the recent engineering practice. To respond this requirement, extensive numerical studies on the liquefation simulations are performed by DIANA program for shaking table tests which have been conducted by the authors. The test models are soil-structure type models with combination of homogenous ground and partially improved ground by compaction.

The good performance of our approach is proved by the results of numerical simulation showing good agreement with experimental data in terms of response acceleration, excess pore pressure, and deformation profile. It is also demonstrated that the numerical results can provide substantial information to understand the mechanisms of soil ground behavior which is not easily obtained by experiments.

The procedure to identify soil constants for the reflecting surface model is also reported in details.     

饱和砂土爆炸液化特性研究

基于有效应力动力分析法,在Byrne有效应力弹塑性模型的基础上,提出了一个能够考虑主应力轴旋转、饱和砂土含有少量气体、饱和砂土液化后的应变软化和应力重分布特性的弹塑性模型。将该模型编制成分析模块,并与通用岩土分析软件FLAC接口,进而对饱和砂土分别在单点、两点（微差）和多点（微差）爆炸地震波荷载作用下进行数值模拟分析,分别考虑了水平、微倾以及斜坡场地等3种工况,并且对爆炸地震波荷载与天然地震波荷载作用下饱和砂土的动力特性进行了对比研究。数值模拟结果表明,该模型能够很好地表现饱和砂土的爆炸液化气特性;不同动载和不同场地条件下,饱和砂土表现的动力特性以及液化行为也不尽相同。     

Liquefaction potential of level ground: Deterministic and probabilistic assessment

Two important aspects of the liquefaction of level ground are discussed; the liquefaction potential of a site due to specified earthquake motions and the total probability of liquefaction at a site due to the entire seismic environment surrounding the site. Dynamic effective stress analysis is used to explore two important case histories of liquefaction due to specified ground motions. The probability of liquefaction is assessed by a method which estimates the joint probability of occurence of earthquake magnitude and acceleration amplitude sufficient to cause liquefaction.     

Bioerosion and carbonate mud production on high-latitude shelves

Low-latitude carbonate muds often are composed either of entire units of skeletons (e.g., algal muds) or of precipitates, whereas high-latitude carbonate muds are bioerosional or result from maceration. Bioerosion at high latitudes is most intense in the photic zone, particularly down to 25 m depth. Shelly substrata may be crushed, bitten, drilled, bored or scraped. Clionid sponges, endolithic algae, acmaeid gastropods and regular echinoids are the most significant agents. Clionids produce distinctive facetted carbonate silt chips when boring, which have been described from both high- and low-latitudes. Faecal pellets break down to yield mud-sized carbonate particles that are more irregular than those produced by maceration. Exhumed infaunal bivalves are often preferred to epifaunal organisms as substrata. Bioerosion occurs very rapidly; shells may be totally infested with boring algae in three months. A “moth-eaten” appearance therefore does not denote a relict grain. Reliable rates of fine sediment production are not yet available.

The mud fraction of northwest European shelf sediment generally contains 10–20% CaCO₃, though an inshore and offshore belt with higher values may be identified. Some Holocene supratidal mud-flats exceed 50% CaCO₃. Much of the shelf represents a modern-day equivalent of the “calcareous shale” facies common in the geological record. Instances of synsedimentary cementation are not uncommon, particularly in association with heavily burrowed muds.     

Experimental study of the grading characteristic effect on the liquefaction resistance of various graded sands and gravelly sands

The liquefaction vulnerability of fine- to coarse-graded saturated sand and gravelly sand is evaluated by stress control cyclic triaxial laboratory tests. Cyclic triaxial test on reconstituted specimens is carried out with relative density of 45 %. Cyclic triaxial tests performed by sinusoidal wave form with 1-Hz frequency and with the effective consolidation pressure equal to 100 kPa. Cyclic stress ratio which caused initial liquefaction in 15 cycles recognized as the liquefaction resistance in cyclic triaxial test of present study. The result had been used to find a relationship between liquefaction resistance and the grading characteristics (e.g., coefficient of curvature and coefficient of uniformity) of various graded sands and gravelly sands. Then, it is realized that a relationship between liquefaction resistance and any of the sizes (i.e., D ₁₀, D ₃₀, D ₅₀, D ₆₀) appears to be more practical.     

Engineering geologic and geotechnical analysis of paleoseismic shaking using liquefaction effects: field examples

The greatest impediments to the widespread acceptance of back-calculated ground motion characteristics from paleoliquefaction studies typically stem from three uncertainties: (1) the significance of changes in the geotechnical properties of post-liquefied sediments (e.g., “aging” and density changes), (2) the selection of appropriate geotechnical soil indices from individual paleoliquefaction sites, and (3) the methodology for integration of back-calculated results of strength of shaking from individual paleoliquefaction sites into a regional assessment of paleoseismic strength of shaking. Presented herein are two case studies that illustrate the methods outlined by Olson et al. [Engineering Geology, this issue] for addressing these uncertainties.

The first case study is for a site near Memphis, Tennessee, wherein cone penetration test data from side-by-side locations, one of liquefaction and the other of no liquefaction, are used to readily discern that the influence of post-liquefaction “aging” and density changes on the measured in situ soil indices is minimal. In the second case study, 12 sites that are at scattered locations in the Wabash Valley and that exhibit paleoliquefaction features are analyzed. The features are first provisionally attributed to the Vincennes Earthquake, which occurred around 6100 years BP, and are used to illustrate our proposed approach for selecting representative soil indices of the liquefied sediments. These indices are used in back-calculating the strength of shaking at the individual sites, the results from which are then incorporated into a regional assessment of the moment magnitude, M, of the Vincennes Earthquake. The regional assessment validated the provisional assumption that the paleoliquefaction features at the scattered sites were induced by the Vincennes Earthquake, in the main, which was determined to have M7.5. The uncertainties and assumptions used in the assessment are discussed in detail.     

饱和兰州黄土液化过程中孔压和应变发展的试验研究

采用WF12440型空心圆柱扭剪仪,用反压饱和法对初始饱和度较低的原状黄土进行饱和,进行室内原状黄土饱和液化试验研究,探讨了饱和兰州黄土液化过程中孔隙水压力、轴向应变、应力-应变滞回圈的发展规律。结果表明,对初始饱和度较低的原状黄土,反压饱和法使孔压系数B值达到0.95以上,即土样完全饱和;兰州黄土在均压固结条件下液化的孔压发展,开始时上升速率较缓慢,循环数一定后会出现孔压迅速增高的现象直至达到有效围压;应力-应变滞回特性随着振动次数的增加发生变化,塑性逐步增大;当轴向应变小于2%时,孔压增长缓慢;此后,孔压上升速率加大,3%应变可以出现在初始液化前;接近液化时偏应力为负值时的有效应力大于正值时的有效应力。     

Source regions of granites and their links to tectonic environment: examples from the western United States

This review, in honor of Ilmari Haapala's retirement, reflects on lessons learned from studies of three granitic systems in western North America: (1) Mesoproterozoic samples from west Texas and east New Mexico; (2) Laramide granitic systems associated with porphyry-copper deposits in Arizona; and (3) granites of the Colorado Mineral Belt. The studies elucidate relationships amongst tectonic setting, source material, and magma chemistry.

Mesoproterozoic basement samples are from two different felsic suites with distinct elemental and isotopic compositions. The first suite, the “plutonic province”, is dominantly magnesian, calc-alkalic to alkali-calcic, and metaluminous. It has low K₂O/Na₂O and Rb/Sr, and Nd model ages of 1.56 to 1.40 Ga. The second suite, the “Panhandle igneous complex”, is magnesian, metaluminous, alkalic, and is part of the Mesoproterozoic belt of magmatism that extends from Finland to southwestern United States. Samples from the Panhandle igneous complex demonstrate three episodes of magmatism: the first pulse was intrusion of quartz monzonite at 1380 to 1370 Ma; the second was comagmatic epizonal granite and rhyolite at 1360 to 1350 Ma. Both of these rock types are high-K to slightly ultra-high-K. The third pulse at 1338 to 1330 Ma was intrusion of ultra-high-K quartz syenite. Nd model ages (1.94 to 1.52 Ga) are distinct from those of the “plutonic province” and systematically older than crystallization ages, implying a substantial crustal input to the magmas.

At the Sierrita porphyry-copper deposit in the Mazatzal Province of southeastern Arizona, trace element, Sr, and Nd isotopic compositions were determined for a suite of andesitic and rhyolitic rocks (67 Ma) intruded by granodiorite and granite. Isotopic composition and chemical evolution are well correlated throughout the suite. Andesite has the least negative initial ε_Nd (−4.3) and lowest ⁸⁷Sr/⁸⁶Sr_i (0.7069). It is also the oldest and chemically most primitive, having low concentrations of Rb, SiO₂, and high concentrations of transition elements. These parameters change through the system to the youngest unit (granite), which has the most negative ε_Nd (−8.5), the highest ⁸⁷Sr/⁸⁶Sr_i (0.7092), and is chemically most evolved. Correlation between chemical and Nd isotopic evolution probably resulted from a continuous process of progressive assimilation, in which mafic magmas invade and incorporate continental crust. Deposits in Arizona with ε_Nd values more negative than the −8.5 of Sierrita lie in the older Yavapai province in the northwestern part of the state. The difference in the most negative epsilon Nd implies that Nd isotopic signature is sensitive to the age of the Precambrian domain.

The granites from the Colorado Mineral Belt were emplaced during the transition from Laramide convergence to mid-Tertiary extension. Three different groups of granites are recognized. The first is Laramide and was formed during assimilation-fractional crystallization involving lower crustal mafic source materials; the second and third groups are mid-Tertiary and represent intracrustal melting of heterogeneous sources. This change in source regions and melt regimes in transition from convergence to extension is fundamental to the Mesozoic and Cenozoic evolution of western North America.     

南京砂强度特征与静态液化现象分析

在松散、中密和密实状态下,以南京粉细砂三轴固结不排水试验结果为基础,进行了强度、变形与静态液化特征的分析。松散南京砂强度特性表现出典型的应变软化,当轴向变形小于1 %时强度达到最大值,而后急剧降低;在50,100 kPa围压时发生了静态液化。但随着固结压力的增大,静态液化消失。与南京砂具有相同土骨架的松散纯净砂却在低围压下未出现静态液化,其形成机制是：粉粒的存在未使土体孔隙比发生较大变化,却引起更大的体缩性;中密和密实南京粉细砂表现出加工硬化的强度特征,临界应力状态线倾角高达55°,具有较高的抗静态液化能力。     

CONE模型与地基动力液化的非线性有效应力分析

在分析建筑物地基的地震（或振动）液化时,为了简化分析土壤-地基的动力相互作用,引入Cone模型,在总应力分析的方法基础上,结合了Martin等提出的孔隙水压力发展模型,并且考虑土壤发生液化的动力非线性过程中材料刚度退化的影响,提出了可以直接求解各区域的孔隙水压力的有效应力简化分析法,并进行液化可能性判别。     

Origin and tectonic interpretation of multiple fault patterns

It is shown that in two-dimensional and three-dimensional deformation accommodated by fracture, the symmetry of the fault patterns is an intrinsic attribute because it reflects the symmetry of either stress or strain tensors. The deformation accommodated by sliding along pre-existing planes, when there is kinematic interaction between that planes, forms multiple fault pattern and multiple slickenline sets during a single deformation event. These fault patterns have no restrictions with respect to symmetry, number of fault sets or fault orientation.

The kinematic analysis developed here shows that an interacting system is formed by two cross cutting faults and three slickenlines. One slickenline must be parallel to the intersection line between the planes. Also, it is demonstrated that the slickenlines generally do not correspond to the shear stress solution on the planes. Thus, the interaction between planes does not satisfy the assumption of parallelism between shear stress and slip vector. We conclude that the inversion methods to calculate paleostress tensors can lead to erroneous interpretations in structurally complex zones with many pre-existing planes of weakness.

We propose four possibilities to form multiple fault patterns: (1) two or more events of faulting obeying Coulomb's law with a change of orientation of the principal stresses in each event; (2) reactivation of non-interacting planes according to the Bott (1959) model; (3) one three-dimensional strain event that obeys the “Slip Model”; this mechanism will form an orthorhombic four-fault pattern and two slickenline sets in a single strain event; and (4) one or more events obeying the interacting block model proposed here, with or without rotation of the principal stresses. We propose the last origin as the most common in continental regions.     

栗西沟尾矿坝地震响应及液化分析

采用不排水有效应力法,利用Geo-slope软件,对金堆城栗西沟尾矿坝进行了地震响应的综合分析与液化计算。结果显示: a .栗西沟尾矿坝在静态条件下是稳定的; b .在7度地震条件下,尾矿坝的加速度反应较小,其放大倍数为2.204; c .栗西沟尾矿坝内动剪应力和动孔压绝大部分是随着地震历时的增加而逐渐增大; d .坝内孔压比都较小,抗液化安全系数较大,但局部液化区的存在,仍可能影响到整个坝体的安全性,应在液化区采取加固措施。      

Einübertragungsnetz zur kluftstatistik

This paper describes the application and characteristics of a new net for crack statistics. The net, here called transmission net or “u-net” (“Übertragungsnetz”, “Ü-Netz”) is used in combination with a transmission table or “u-table” (“Übertragungstabelle”, “Ü-Tabelle”). Its purpose is to obtain a sphere of crack locations from a series of cracks, having been measured according to strike and inclination.

The “u-net” is composed of a grid subdivided into degrees or degree-grid (“Gradnetz”) and an equal-area-grid (“flächengleiches Netz”). Having replaced the direction σ of strike by its normal δ, the cracks are registered into the meshes of the degree-grid and are there counted. Then they are transmitted into the equal-area-grid according to the percent values of the u-table. From the pattern of frequency numbers (“Häufigkeitszahlen”) in the equal-area-grid the sphere of crack locations is obtained.

Particular specifications regulate the procedure for special measure values, i.e., those of the integral multiples of 5°, especially the angles of inclination τ = 0° and τ = 90° (see 1.4). With greater inaccuracies in measurements, one changes, by means of a given table (see Table V), to a degree-grid of 10°. With very small inaccuracies, on the other hand, the procedure may be simplified, the degree-grid becoming unnecessary (see 1.5). The meshes near the centre, being too long, may be avoided by an additional circle (“Zusatzkreis”—see 1.3).

The “u-net” was constructed in such a way that the spheres of crack locations report the real frequency distribution at all times, free from systematic errors. This is achieved by the method that all calculations follow the principle of area equality or area proportionality on the hemisphere (see 2.1). The procedure using the “u-net” can be adapted to differential accuracies of measuring. It is especially suitable for large numbers of cracks; it is simple in calculation and may easily be programmed for digital computers. Thus the “u-net” is advantageous for all applications in which a large number of cracks has to be dealt with. Such applications are very frequent in rock mechanics, in engineering geology for the purpose of foundation of large hydraulic buildings (dams, caverns), in petrography, tectonics and in geophysical investigations such as the determination of crack structures with a view to explaining micromagnetic occurrences, for instance. Furthermore, the “u-net” is applicable not only to crack statistics but also to other similar statistical methods, e.g., to the statistics of cristal axes or to geographic frequency analyses.     

Assessment and Prediction of Liquefaction Potential Using Different Artificial Neural Network Models: A Case Study

Soil liquefaction as a transformation of granular material from solid to liquid state is a type of ground failure commonly associated with moderate to large earthquakes and refers to the loss of strength in saturated, cohesionless soils due to the build-up of pore water pressures and reduction of the effective stress during dynamic loading. In this paper, assessment and prediction of liquefaction potential of soils subjected to earthquake using two different artificial neural network models based on mechanical and geotechnical related parameters (model A) and earthquake related parameters (model B) have been proposed. In model A the depth, unit weight, SPT-N value, shear wave velocity, soil type and fine contents and in model B the depth, stress reduction factor, cyclic stress ratio, cyclic resistance ratio, pore pressure, total and effective vertical stress were considered as network inputs. Among the numerous tested models, the 6-4-4-2-1 structure correspond to model A and 7-5-4-6-1 for model B due to minimum network root mean square errors were selected as optimized network architecture models in this study. The performance of the network models were controlled approved and evaluated using several statistical criteria, regression analysis as well as detailed comparison with known accepted procedures. The results represented that the model A satisfied almost all the employed criteria and showed better performance than model B. The sensitivity analysis in this study showed that depth, shear wave velocity and SPT-N value for model A and cyclic resistance ratio, cyclic stress ratio and effective vertical stress for model B are the three most effective parameters on liquefaction potential analysis. Moreover, the calculated absolute error for model A represented better performance than model B. The reasonable agreement of network output in comparison with the results from previously accepted methods indicate satisfactory network performance for prediction of liquefaction potential analysis.