Scholarly article on topic 'Liquefaction analysis of alluvial soil deposits in Bedsa south west of Cairo'

Liquefaction analysis of alluvial soil deposits in Bedsa south west of Cairo Academic research paper on "Earth and related environmental sciences"

Share paper
Academic journal
Ain Shams Engineering Journal
{"Soil liquefaction" / Settlement / "Lateral spreading" / Earthquake / "Alluvial deposits"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — Kamal Mohamed Hafez Ismail Ibrahim

Abstract Bedsa is one of the districts in Dahshour that lays south west of Cairo and suffered from liquefaction during October 1992 earthquake, Egypt. The soil profile consists of alluvial river Nile deposits mainly sandy mud with low plasticity; the ground water is shallow. The earthquake hypocenter was 18km far away with local magnitude 5.8; the fault length was 13.8km, as recorded by the Egyptian national seismological network (ENSN) at Helwan. The analysis used the empirical method introduced by the national center for earthquake engineering research (NCEER) based on field standard penetration of soil. It is found that the studied area can liquefy since there are saturated loose sandy silt layers at depth ranges from 7 to 14m. The settlement is about 26cm. The probability of liquefaction ranges between 40% and 100%. The presence of impermeable surface from medium cohesive silty clay acts as a plug resisting and trapping the upward flow of water during liquefaction, so fountain and spouts at weak points occurs. It is wise to use point bearing piles with foundation level deeper than 14m beyond the liquefiable depth away from ground slopes, otherwise liquefaction improving techniques have to be applied in the area.

Academic research paper on topic "Liquefaction analysis of alluvial soil deposits in Bedsa south west of Cairo"

Ain Shams Engineering Journal (2014) xxx, xxx-xxx

Ain Shams University Ain Shams Engineering Journal


Liquefaction analysis of alluvial south west of Cairo

Kamal Mohamed Hafez Ismail Ibrahim *

Civil Engineering Dep., Suez Canal University, Egypt Received 20 August 2013; revised 30 January 2014; accepted 1 February 2014

soil deposits in Bedsa


Soil liquefaction; Settlement; Lateral spreading; Earthquake; Alluvial deposits

Abstract Bedsa is one of the districts in Dahshour that lays south west of Cairo and suffered from liquefaction during October 1992 earthquake, Egypt. The soil profile consists of alluvial river Nile deposits mainly sandy mud with low plasticity; the ground water is shallow. The earthquake hypocenter was 18 km far away with local magnitude 5.8; the fault length was 13.8 km, as recorded by the Egyptian national seismological network (ENSN) at Helwan. The analysis used the empirical method introduced by the national center for earthquake engineering research (NCEER) based on field standard penetration of soil. It is found that the studied area can liquefy since there are saturated loose sandy silt layers at depth ranges from 7 to 14 m. The settlement is about 26 cm. The probability of liquefaction ranges between 40% and 100%. The presence of impermeable surface from medium cohesive silty clay acts as a plug resisting and trapping the upward flow of water during liquefaction, so fountain and spouts at weak points occurs. It is wise to use point bearing piles with foundation level deeper than 14 m beyond the liquefiable depth away from ground slopes, otherwise liquefaction improving techniques have to be applied in the area.

© 2014 Production and hosting by Elsevier B.V. on behalf of Ain Shams University.

1. Introduction

Liquefaction was first initiated by Seed [1] through experimental studies on sand samples. It is the process of reduction of shear strength for low plastic loose cohesionless soil. Pore pressure buildup due to static or cyclic stress applications. The soil looses contact between its grains and upward flow of water takes place. If the magnitude of pore-water pressure generated

equals the total vertical stress, the effective stress becomes zero and the soil is said to have liquefied. The possibility of its occurrence depends on the initial void ratio or relative density of sand and the confining pressure (Seed [1]). Formation of sand boils and mud-spouts at the ground surface by seepage of water through ground cracks or in some cases by the development of quick sand conditions over substantial areas (Seed and Idriss [2]). Housner and Jennings [3] discussed the formation of sand boils in terms of soil porosity, permeability, elasticity, and degree of consolidation. Sand boils were attributed to non-homogeneity in permeability near the ground surface. Scott and Zuckerman [4] presented both experimental and analytical studies on the mechanics of liquefaction and sand boil formation in sandy soil deposits. They found that the presence of silt or a similar fine grained layer at the surface (above

* Tel.: +20 1001525472.

E-mail address:

Peer review under responsibility of Ain Shams University.

2090-4479 © 2014 Production and hosting by Elsevier B.V. on behalf of Ain Shams University.

the liquefied layer) is conducive to the generation of sand boils. In contrast to "piping", sand boils were observed to propagate from the source of pressure to the outlet by a mechanism of cavity formation. Adalier [5] also demonstrated that stratified soil profiles are conducive to sand boil formation. It was shown that low permeability and cohesion of an overlying upper layer may lead to the formation of large sand boils, as the extruded water mainly travels through cracks and weak zones within this upper layer.

Cyclic testing of a wide range of soils found to liquefy in Adapazari during the Kocaeli earthquake confirmed that these fine-grained soils were susceptible to liquefaction. It is not the amount of "clay-size" particles in the soil; rather, it is the amount and type of clay minerals in the soil that best indicate liquefaction susceptibility (Bray et al. [6]).

Fig. 1 shows recommendation given by Bray et al. [6] and Seed et al. [7] according to atterberg limits for soil to liquefy and that plasticity index appears to be a better indicator of liquefaction susceptibility.

— Loose soils with PI < 12 and wc/LL > 0.85 were susceptible to liquefaction.

— Loose soil with 12 < PI < 18 and wc/LL > 0.8 were systematically more resistant to liquefaction.

— Soils with PI > 18 tested at low effective confining stresses were not susceptible to liquefaction.

— The location of a soil on the Casagrande plasticity chart and, or in combination with, the use of the "C" descriptor, (USCS) (e.g. CH, CL, SC, and GC) are considered as non-liquefiable.

— Liquefiable fine-grained soils should have LL <35 and plot below the A-line or have PI < 7.

— Seed et al. [7] stated that soils with LL < 37 and PI < 12 are potentially liquefiable, and those with 37 < LL < 47 and 12 < PI < 20 require laboratory testing.

Ishihara [8] studied other factors which control liquefaction and/or cyclic mobility such as:

Confining pressure, initial static shear stress and stress-path.

Plito [9] found that soils with LL < 25 and PI < 7 are liquefiable, and soils with 25 < LL < 35 and 7 < PI < 10

Figure 1 Atterberg limits of fine-grained soil reported by Bray et al. [6] to have ''liquefied'' at 12 building sites during the 1999 Kocaeli earthquake and recommendations by Seed et al. [7].

are potentially liquefiable, and soils with 35 < LL < 50 and 10 < PI < 15 are susceptible to cyclic mobility. Cyclic mobility of clay may depend upon plasticity index, wc/LL ratio, Initial static shear stress, confining pressure, and stress path.

Several recent earthquakes indicate that many cohesive soils had liquefied. These cohesive soils had clay fraction less than 20%, liquid limit between 21% and 35%, plasticity index between 4% and 14% and water content more than 90% of their liquid limit. Kishida [10] reported liquefaction of soils with up to 70% fines and 10% clay fraction during Mino-Owar earthquake. Andrews and Martin [11] evaluate liquefactions of fine-grained soils as given in Table 1

Ishihara et al. [12] had set up a criterion to stipulate a threshold value for the thickness of a non-liquefiable surface layer to avoid ground damage due to liquefaction, as shown in Fig. 2. Although this figure is believed to be speculative and should not be used for design purposes, it provides initial guidance in this matter for sites having a buried liquefiable sand layer with a standard penetration resistance of less than 10 blows per foot (0.3 m). It should also be noted that even though the thickness of a non-liquefiable surface layer exceeds the threshold thickness shown in Fig. 2, the ground surface may still experience some settlement which may be undesirable for certain settlement-sensitive structures. Like all of the empirical curves, this figure is based on just three case histories, may need to be modified as more data become available.

In order to induce extensive damage at level ground surface from liquefaction, the liquefied soil layer must be thick enough so that the resulting uplift pressure and amount of water expelled from the liquefied layer can result in ground rupture such as sand boiling and fissuring (Ishihara et al. [12]; Dobry [13]). If the liquefied sand layer is thin and buried within a soil profile, the presence of a non-liquefiable surface layer may prevent the effects of the at-depth liquefaction from reaching the surface.

Fig. 3 shows the fault mechanism of 12 October 1992 earthquake southwest of Cairo, Egypt. It occurred on Monday at 15:09 local tim. It was a damaging earthquake of magnitude Mw = 5.8 and took place in Dahshour region, about 18 km SW of downtown Cairo at coordinates 29.77 N, 31.07 E and was followed by a sequence of aftershocks (Kamal et al. [14]).

The earthquake area lies in the northern part of the western desert tectonic zone, which forms part of the African plate. The focal depth was 23 km. The only earthquake that had occurred in this region is the 4.9 Ms event October 1920 at 29.50N, 31.30E, and the only known earthquake during this historical period had occurred in August 1847 A.D. and destroyed 3000 houses and 42 mosques in Cairo and Northern Egypt (Kebeasy et al. [15]).

Fig. 4 shows large sand-boil craters that had occurred in agricultural field at locations 2.5 km away from the Nile, and 1.0 km west of El-Beleda village (Elgamal et al. [16]).

Throughout centuries, the Nile River flooded the plains along its path every summer until the construction of the Aswan High Dam in 1971, so the age of sediments can occur in late Holocene. Table 2 shows evaluation of liquefaction according to age of deposits (Yould and Perkins [17]).

Natural deposits of alluvial and fluvial origins generally have soil grains in the state of loose packing which are young, weak and free from added strength due to cementation aging.

Youd and Hoose [18] stated that, as a rule of thumb, alluvial deposits older than late Pleistocene (10,000-130,000 years)

Table 1 Criteria recommended by Andrews and Martin [11], for evaluating the liquefactions of fine-grained soils.

LL < 32 LL > 32

Minus 2 im fraction < 10% Minus 2 im fraction P 10% Susceptible to liquefaction Further studies are required (consider non-plastic clay sized grains) Further studies required (consider plastic non-clay sized grain) Not susceptible to liquefaction

Figure 2 Proposed boundary curves for site identification of liquefaction-induced damage (Ishirhara et al. [12]).

Figure 3 Fault mechanism of 12 October 1992 earthquake southwest of Cairo (Kamal et al. [14]).

are unlikely to liquefy except under severe earthquake loading conditions, while late Holocene deposits (1000 years or less) are most likely to liquefy, and earlier Holocene (100010,000 years) deposits are moderately liquefiable.

Figure 4 Largest sand valve crater due to soil liquefaction in Bedsa (Elgamal et al. [16]).

A list of selected methods for ground improvement and structural solutions to reduce hazards from liquefaction example, excavation and/or compaction, in situ ground densifica-tion, ground treatment such as jet grouting, gravel drains, berm, dikes, deep foundations and reinforced shallow foundations were studied by Ledbetter [19].

2. Methodology

Based on empirical method introduced by NCEER [20] for assessment of liquefaction potential of soil, this procedure essentially compares the cyclic resistance ratio (CRR) at a given depth with the earthquake-induced cyclic stress ratio (CSR) at that depth from specified design earthquake. Reasonable estimates of liquefaction potential can be made based on simple in situ test data such as standard penetration values (N0 or (N^^.

2.1. Evaluation of CRR for liquefiable soil

Values of cyclic resistance ratio (CRR) were originally established from empirical correlations using extensive databases for sites that did or did not liquefy during past earthquakes where values of (N1)60 could be correlated with liquefied strata.

Base line chart defining values of CRR as a function of (N1)60 for earthquake of magnitude 7.5 were given by Seed et al. [21].

2.1.1. Corrected S.P.T. (standard penetration test) measurements

(Ni)6o = Nspt x CN x Ce x Cb x Cs x Cr

Table 2 Estimated susceptibility of sedimentary deposits due to liquefaction, during strong seismic shaking (Yould and Perkins [17]).

Type of deposit Distribution of cohesionless deposits By age deposits

<500 year Holocene Plaistocene Pre-plaistocent

River channel Flood plain Alluvial fan Delta and fan-delta Coastal delta Locally variable Locally variable Widespread Widespread Widespread Very high High Moderate High Very high High Moderate Low Low High Low Low Low Low Low Very low Very low Very low Very low Very low

CN= J— 6 2.0

actual energy at top of drill rod 0.6 * Theoretical maximum SPT hammerenergy ER "60"

CN is correction factor; Pa is the atmospheric air pressure. ER is the energy ratio; CE was given by Seed et al. [21]. CB is correction for boring diameter given by Robertson and Wride [22]. CS is sampler correction = 1.0 for standard sampler and 1.2 otherwise. CR is correction due to loss of energy through reflection in short lengths of drill rod. CR = 0.75, (15 + z)/24 and 1.0 for Z 6 3 m, 3 < Z < 9 m and Z > 9 m respectively, where Z is the length of drill rod in meters (Robertson and Wride [22]).

This method is used basically used for clean sand soil, but if fines exist a correction D(N1)60 can be used to find a clean sand equivalent (N1)60

(N1)60 = (N1)60 + A(M)60 (4)

A(N060 = 0 for FC < 5%,

D(N1)60 = 7 x(FC - 5)/30 for 5 < FC < 35%

D(N1 )60 = 7 for FC > FC > 35%, where FC is the percentage of fines

Relative densities may be estimated from correlations with either S.P.T or C.P.T results. A modified version of Meyerhofs [23] correlation was used to estimate relative densities of clean sand from SPT blow counts

16 x ^(N)^ = 14 x ^(NU [(N1)«, 6 42]

where Dr is the relative density of a clean sand as a percentage; (Ni)60 normalized S.P.T., N value corrected for the rod energy ratio ~60% reference energy, overburden effective stress ~100 kPa reference effective stress, rod length, borehole diameter, and sampling method, as discussed by Youd and Idriss [24]; and (N1)78 equals to (N1)50/1.3.

2.1.2. Cyclic resistance ratio

Blake [25] proposed CRR for clean sand based on equivalent (N1)60 for earthquake of magnitude 7.5, as shown in Fig. 5. A value of (N1)60 > 30 indicates non-liquefiable soil.

CRRr = 1 = expl

CRR for any earthquake of other magnitude Mw other than

CRR = CRRm=7 5 x MSF (7)

ForMw < 7.0, MSF = 103 x M;3:46

ForMw > 7.0, MSF = 10224 x M-2:56

For liquefiable soil deposits subjected to significant overburden or static shear stresses, the calculated CRR should be multiplied by two additional factors, Krdepends on confining pressure and Kadepends on static horizontal shear stresses, vertical effective stress, and relative density of soil), but they are not included in the analysis.

2.2. Cyclic stress ratio (CSR) induced by earthquakes

CSR = 0.65 amax x x rd

g r eff

where "g" is the gravity acceleration, ot is the total stress, of is the effective stress and rd is a reduction factor given by Blake [25].

(1.0 - 0.411z05 + 0.04z + 0.00175z15) (1.0 + 0.417z0 5 + 0.0573z - 0.0062z15 + 0.00121z2)

2.3. Prediction of liquefied thickness

The factor of safety against liquefaction FSiiq is defined by Ishihara [26].

FSliq —


The soil predicted to liquefy is FSliq 6 1.0 (prone to liquefaction), otherwise no liquefaction.

2.4. Liquefaction Potential Index (LPI)

The Liquefaction Potential Index (LPI) was developed and presented by Iwasaki et al. [27]. The LPI is defined as:

LPI — Fi W(z)dz

where W(z) = 10-0.5z, F1 = 1 - FS for FS <1.0, F1 = 0 for FS >1.0 and z is the depth below the ground surface in meters.

N1(60) c

Figure 5 Proposed CRR7.5 curve for clean sand (Blake [25]).

2.5. Lateral displacement index (LDI)

Cmax dz

where LDI is the lateral displacement index, Zmax = maximum depth below all the potential liquefiable layers with a calculated FS < 2.0, and ymax = maximum shear strain.

Although LDI has the units of displacement, it provides an index to quantify potential lateral displacements for a given soil profile, soil properties, and earthquake characteristics.

2.6. Lateral displacement (LD)

The actual magnitude of lateral displacement depends on both LDI and geometric parameters characterizing ground geometry. The ground inclination and the presence of near-by free face height affect lateral spreading of the ground surface during liquefaction as in Fig. 6 (Lelio et al. [28]). Below 2H or if FS > 2, the lateral spreading is ignored. Table 3 shows some case histories for LDI and LD (Zhang et al. [29]).

Figure 6 Ratio of measured LD to LDI versus L¡H for case histories with level ground and a free face: standard penetration test-based data, (Lelio et al. [28]).

2.7. Post-liquefaction settlement

For level ground conditions the settlement can be computed from the volumetric reconsolidation strains induced as the excess pore water pressures dissipate. The curve proposed by Ishihara and Yoshimine [30] in Fig. 7 indicates that volumetric reconsolidation strains can range between about 4.5% for very loose sand to 1.2% for very dense sands. These curves are recommended for estimating post-liquefaction settlements.

3. Soil bore log and liquefaction analysis

Village of Bedsa consists of 1-2 stories and was severely damaged during the seismic excitation. No indications of lateral spreading were observed at this site of fairly flat terrain. The soil ejected to ground surface during liquefaction was thought to be fine sandy silt, (Elgamal et al. [16]).

The bore log in Fig. 8 shows the soil profile in Bedsa site and its corresponding S.P.T values. The soil at the surface starts with medium brown silty clay up to depth 5.5 m with average S.P.T in the order of 9 and has unconfined strength

about 1.0 kg/cm2, then it changes to gray silty clay with traces of sand up to 7.5 m depth, follows that loose gray sandy silt up to depth 15 m, almost this layer is prone to liquefaction, follows that yellow coarse to medium dense sand up to depth 30 m. The ground water table is 0.8 m below the ground surface (Elgamal et al. [16]).

The software NOVOLIQ [31] is designed for soil liquefaction analysis during earthquake. It supports multilayer as well as single layer stratigraphy. It is used to analyze Bedsa site and to check its potential to liquefaction. The input data are the following: soil bore log profile, percentage of fines for every layer, unit weight, S.P.T. values as shown in Fig. 8, layers that are prone to liquefaction have to be defined (the upper 7.5 m silty clay layers are not prone to liquefaction), magnitude of 12th October earthquake equals to 5.8, fault distance 18 km, post-liquefaction topography L/H = 4.

4. Results and discussion

Fig. 9a shows the row and corrected N(60) S.P.T values based on clean sand equivalent for different soil layers with depth.

Table 3 Case histories with nearly level ground and a free face (Zhang et al. [29]).

Case history Dagupan 1990 Moss landing 1989 Jensen plant 1971 Niigata 1964 Alaska 1964

No. of sites 3 3 3 2 1

No of LD data 7 6 13 66 1

LD (cm) 50-600 30-125 2-100 41-1015 157

Accuracy of LD (± cm) >50 >10 47 72 10-50

LDI (cm) 79-220 28-114 5-20 246-637 173

H (cm) 4-115 1.9-2.4 10.4-17.2 4.9-5.2 4.9

L/H 3.8-27.3 6.3-23.5 8.7-30.5 5.1-36.2 6.2

Mw 7.6 7.0 6.4 7.5 9.2

Amax (g) 0.2 0.25 0.55 0.19 0.33

No. of S.P.T. and C.P.T. 3 (S.P.T.) 7 (C.P.T.) 20 (S.P.T.) 47 (S.P.T.) 3 (S.P.T.)

Note: C.P.T. = Cone penetration test.

Figure 7 Recommended relationships for volumetric reconsolidation strains as a function of maximum shear strain and relative density (Ishihara and Yoshimine [30]).

The reduction factor rd distribution with depth is shown in Fig. 9b according to Eq. (9). The effective and the total overburden pressure distribution are shown in Fig. 9c.

The cyclic resistance ratio CRR, the cyclic stress ration CSR and the factor of safety of soil layers are shown in Fig. 10a-c. It is noticed that the layers that are prone to liquefaction have a safety factor less than one and this is achieved for soil layers lying between 7 and 14 m. Also the probability of liquefaction PL% shown in Fig. 10d, ranges between 40% and 100% for that range.

Post-liquefaction maximum soil shear strain and volumetric strain are shown in Fig. 11a and b. It is noticed that these values occurs in the layer of loose sandy silt that are prone to liquefaction that range from 7 to 14 m depth. Fig. 11a and b shows that the maximum soil shear strain reaches about 55%, while the maximum soil volumetric strain reaches about 4.2%. Fig. 10c shows the soil settlement due to liquefaction process. The soil settlement for depths greater than 14 m is zero, while it increases gradually at shallow depths till reaching a maximum value 26 cm at depth about 7 m then it remains constant toward ground surface since the surface silty clay layers are defined as non-liquefying layers. Fig. 11d shows that

(■ dtnolM Iwl wakM mi Mvrad from tu oofTMpondkiQ quO

Figure 8 Soil profile near sand boils at Bedsa (Elgamal et al. [16]).

Figure 9 Variation of (a) SPT, (b) reduction factor rd, and (c) overburden and effective pressure with depth.

Figure 10 Variation of (a) CRR7 5, (b) CSR, (c) Fs, and (d) probability of liquefaction PL with soil depth.

Figure 11


Variation of (a) maximum shear strain (%), (b) volumetric strain (%), (c) settlement, and (d) liquefaction index LDI with soil

Excessive settlement and rotation of isolated footing if it is near a spout

Shallow Ground water

Figure 12 Pore water upward flow extruding ejected water soil mixture though spouts.

the maximum lateral displacement index reaches about 250 cm (refer to Table 3) and does not reach the ground surface. It is ignored for surface cohesive soil (not prone to liquefaction) where safety factor is relatively high p 2.

Fig. 12 shows that in Bedsa the presence of impermeable surface cohesive silty clay layer resists and trap the upward flow of water to the surface causing accumulated rise of pore water pressure under it, as a result and due to non-homogeneity at weak zones, shear failure and spouts occurs and water fountain accompanied with underground liquefied sandy silt soil are extruded.

In Bedsa it is wise to consider differential settlement as a guide in foundation design. Isolated footings may undergo excessive settlement and rotation if it is near a spout. Pile foundation with foundation level deeper than 15 m (liquefied soil

depth equals 14 m from study) is preferable to resist the expected 26 cm total settlement and any expected differential settlement but it has to be away enough (L/H > 40, Fig. 6, as given by Lelio et al. [28]) from canals or drains and not to be constructed in steep ground slopes to minimize LD on pile stability, otherwise special liquefaction improving techniques can be performed.

5. Conclusions

The applied analysis based on S.P.T. shows that the soil profile of Bedsa as one of Dahshour districts is prone to liquefaction due to the presence of nearby fault 18 km away and due to the presence of prone to liquefaction loose sandy silt layers that lie between 7 and 14 m. Silty clay layers which extend from ground surface to 7 m depth are not liquefiable which agree with boundaries given by Seed et al. [7] for liquefiable soils. Presence of surface silty clay soil with low coefficient of permeability behaves as an impermeable plug trapping upward flow of water during liquefaction process as a result craters at weak zones occurs and ejected water mixed with loose underground sandy silt is spread at the ground surface.

The expected total settlement is 26 cm, so it is wise to choose pile foundation with foundation level deeper than 14 m the liquefiable soil depth and to be constructed in sites that are away (L/H > 40) from any canals or drains and

not in steep ground slopes to minimize lateral displacement which affects pile stability, otherwise other special improving soil techniques can be suggested.


[1] Seed HB. Soil liquefaction and cyclic mobility evaluation for level ground during earthquakes. J Geotechn Eng Div ASCE 1979; 105(2):201-55.

[2] Seed HB, Idriss IM. Ground motions and soil liquefaction during earthquakes. Oakland (CA): Earthquake Engineering Research Institute Monograph; 1982.

[3] Housner GW, Jennings PC. Generation of artificial earthquakes. ASCE J Eng Mech Div 1964;90:113-50.

[4] Scott RF, Zuckerman KA. Sand blows and liquefaction, The Great Alaska Earthquake of 1964-engineeringpublication 1606. Washington DC: National Academy of Sciences; 1972, p. 179-89.

[5] Adalier K. Post-liquefaction behavior of soil systems. I.S. Thesis, Dept of Civil Engineering, Rensselaer Polytechnic Institute, Troy, NY; 1992.

[6] Bray Jonathan D, Sancio RB, Reimer MF, Durgunoglu T. Liquefaction susceptibility of fine-grained soils. In: Proc 11th int conf on soil dynamics and earthquake engineering and 3rd inter conf on earthquake geotech engrg, vol. 1. Berkely, CA; 2004. p. 655-62.

[7] Seed RB, Cetin KO, Moss RES, Kammerer AM, Wu J, Pestana JM, et al. Recent advances in soil liquefaction engineering. A unified and consistent framework. In: 26th Annual ASCE Los Angelos geotechnical spring seminar, Long Beach, California. Keynote Presentation; 2003.

[8] Ishihara K. Liquefaction of subsurface soils during earthquakes, department of civil engineering. Tokyo; May 1974.

[9] Plito C. Plasticity based liqurefaction criteria. In: Proc 4th int conf on recent adv in geotech earth engrg and soil dynamics. San Diego; 2001.

[10] Kishida H. Characteristics of liquefied sands during mino-owari, tohnankai, and fukui earthquakes. Soils Found. 1969;9(1):75-92.

[11] Andrews DCA, Martin GR. Criteria for liquefaction of silty soils. In: Proc 12th WCEE. Auckland, New Zealand; 2000.

[12] Ishihara K, Shimuzu K, Yamada Y. Pore water pressures measured in sand deposits during an earthquake. Soils Found 1981;21:85-100;

Jaeger JC. Elasticity, fracture and flow. London: Methuen and Co. Ltd.; 1969.

[13] Dobry R. Liquefaction of soils during earthquakes. National Research Council (NRC). Committee on earthquake engineering. Report no CETS-EE-001, Washington DC; 1985.

[14] Kamal MA, Hesham MH, Ahmad SA, Ezzeldien MI. Seismo-logical aspects of the Cairo earthquake, 12th October 1992. Ann. Geofis. 2000;43(3).

[15] Kebeasy RM, Maamoun M, Albert RNH, Megahed M. Earthquake activity and earthquake risk around Alexandria, Egypt. Bull IISEE Jpn 1981;19:93-113.

[16] Elgamal A-W, Amer MI, Adalier K, Abulfadl A. Liquefaction during the October 12, 1992 Egyptian Dahshour earthquake. In: Proceedings: 3rd international conference on case histories in geotechnical engineering. St. Louis, Mo; 1993. p. 14-8.

[17] Youd TL, Perkins M. Mapping liquefaction-induced ground failure potential. J Geotech Eng Div, ASCE 1978;104(GT4).

[18] Youd TL, Hoose SN. Historic ground failures in Northern California triggered by earthquakes. US geological survey professional paper 993; 1978. p. 175.

[19] Ledbetter RH. Improvements of liquefiable foundation conditions beneath existing structures. Technical report REMR-GT-2. US army corps of engineers, waterways experiment station, Vicks-burg, Mississippi; 1985.

[20] 20-.NCEER. Proceedings of the NCEER workshop on evaluation of liquefaction resistance of soils. Technical report NCEER-97-0022. National Center for Earthquake Engineering Research. Buffalo, New York; 1997.

[21] Seed HB, Tokimatsu K, Harder LF, Chung RM. Influence of SPT procedures in soil liquefaction resistance evaluations. J Geotech Eng Div ASCE 1985;3(GT12).

[22] Robertson PK, Wride CE. Evaluating cyclic liquefaction potential using the cone penetration test. Can Geotech J 1998;35:442-59.

[23] Meyerhof GG. Discussion on research on determining the density of sands. In: Proc 4th int conf of soil mechanics and foundation engineering, vol. 3. London; 1957. p. 110.

[24] Youd TL, Idriss IM. Liquefaction resistance of soils: summary report from the 1996 NCEER and 1998 NCEER/NSF workshops on evaluation of liquefaction resistance of soils. J Geotech Geo-environ Eng ASCE 2001;127(10):297-313.

[25] Thomas F. Blake. Liquefaction analysis. NCEER workshop. Fugro west inc, Ventura California; 1997.

[26] Ishihara K. Stability of natural deposits during earthquakes. In: Proceedings of the 11th international conference on soil mechanics and foundation engineering, vol. 1. San Francisco; 1985. p. 321-76.

[27] Iwasaki T, Arakawa T, Tokida K. Simplified procedures for assessing soil liquefaction during earthquakes. In: Proc conference on soil dynamics and earthquake engineering. Southampton; 1982. p. 925-39.

[28] Mejia Lelio H, Hughes David K, Sun Joseph I. Liquefaction at moss landing during the 1989 Loma Prieta earthquake, earthquake engineering. In: Tenth world conference. Balkema, Rotterdam; 1992. ISBN 9054100605.

[29] Zhang G, Robertson PK, Brachman RWI. Estimating liquefaction-induced lateral displacements using the standard penetration test or cone penetration test. J Geotech Geo-environ Eng © ASCE/AUGUST 2004/861.

[30] Ishihara K, Yoshimine M. Evaluation of settlements in sand deposits following liquefaction during earthquakes. Soils Found 1992;32:173-88.

[31] NovoLiq software, programming Alireza Afkham designed for soil liquefaction analysis during earthquake, first released on September 2009. Address: 4188 Hoskins Road, North Vancouver, BC, Canada. <>.