liquefaction mitigation in silty soils using stone columns supplemented with wick drains

List of Tables Table 2.2 Selected Vibro Compaction Projects as a Liquefaction Countermeasure 13 Table 2.3 Selected Vibro Replacement Stone Columns Table 2.5 Selected Deep Dynamic Compact

LIQUEFACTION MITIGATION IN SILTY SOILS

USING STONE COLUMNS SUPPLEMENTED WITH WICK DRAINS

BY THEVACHANDRAN SHENTHAN

December 2005

Dissertation submitted to the Faculty of the Graduate School of State University of New York at Buffalo

In partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Department of Civil, Structural, and Environmental Engineering

UMI Number: 3203927

3203927 2006

UMI Microform Copyright

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company

300 North Zeeb Road P.O Box 1346 Ann Arbor, MI 48106-1346

by ProQuest Information and Learning Company

To My Wife Suganya…

ACKNOWLEDGEMENTS

I wish to express my deepest gratitude and sincere thanks to my academic advisor – Dr

S Thevanayagam, for his invaluable guidance, inspiration, as well as support and continuous encouragement, without which this work would not have been possible

I would also like to thank my dissertation committee members Dr P.K Banerjee and Dr

S Ahmad, professors at University at Buffalo, and my external review committee member Dr G.R Martin, professor at University of Southern California, for their continuous guidance and invaluable feedback on my research study

I would also like to thank graduate student J Liang, who helped a lot in carrying out cyclic triaxial tests, and graduate student W Jia, who helped me in numerical modeling Graduate students, T Kanagalingam and R Nashed are also thanked for helping me in generating the field and laboratory tests database, and supporting me through out my research

I also need to thank my parents, brother, sisters, my parents in law, sister in law, and brothers in law When I began to pursue a higher degree here, SUNY at Buffalo, it was they who kept on giving me courage and support

Funding for this research work was provided by FHWA / MCEER Highway Project-094, and the financial support is greatly appreciated

Last, but not least, I would like to thank my wife Suganya and my sons Vasheeigaran and Vagish for their everlasting love, support and encouragement

1.0 Introduction 01

2.1.3 Vibro Concrete Columns 19

2.5.4 Limitations 50

3.3 Post-Liquefaction Pore Pressure Dissipation and Densification 55 3.4 Recent Developments in Understanding Liquefaction

6.2.1 Pore Pressure Generation 103

7.4 Equipment and Basic Construction Procedure 153

E- SPT – Relative Density Relationships 316

List of Tables

Table 2.2 Selected Vibro Compaction Projects as a Liquefaction Countermeasure 13 Table 2.3 Selected Vibro Replacement Stone Columns

Table 2.5 Selected Deep Dynamic Compaction

Table 2.6 Selected Deep Blasting Projects as a Liquefaction Countermeasure 22 Table 2.7 Selected Compaction Grouting

Table 2.8 Selected Gravel Drains Projects as a Liquefaction Countermeasure 26

Table 2.10 Selected Jet Grouting Projects as a Liquefaction Countermeasure 32 Table 2.11 Selected Permeation Grouting

Table 2.12 Selected Deep Soil Mixing Projects as a Liquefaction Countermeasure 35

Table 5.1 Anticipated Post-Densification Cyclic Strength 95

Table 6.4 Comparison of Densification Processes using

List of Figures

Fig 2.1 Applicable Grain Size Ranges for Different Stabilization Methods 12

Fig 2.8 Schematic Diagrams of (a) Jet, (b) Permeation, and

Fig 2.11 Relationship between Greatest Pore Pressure Ratio (rg)

Fig 2.12 A Part of the Stone Column Design Chart Developed by Onoue (1998) 46

Fig 2.13 Required Area Replacement Ratio for Liquefaction Reduction by

Stress Redistribution to achieve a Post-Treatment FS=1 48

Fig 4.3 GEOCOMP Triaxial Apparatus 65

Fig 5.1 Pore Pressure Generation Data for Sands and Clayey Silts

Fig 5.2 Pore Water Pressure Generation Data (This Study) Along with

Fig 5.10 Volumetric strain Due to Remolding of Soil Following Liquefaction 92

Fig 5.11 Volumetric Strain of Sand and Sandy silts vs (a) Relative Density

(b) Eq Rela Density (FC<FCth) and (c) Eq Rela Density (FC>FCth) 93

Fig 6.3 Radiation of Vibratory Energy - Schematic 107 Fig 6.4 Defin of Radii Used in the Analysis, and Stress States around the Probe 109

Fig 6.11 Pore Pressure Changes within the Soil during Stone Column Installation 127 Fig 6.12 Effect of Hydraulic Conductivity of Soil on the

Fig 6.13 Effect of Wick Drains in Densification during Stone Column Installation 130

Fig 7.4a Site Layout and Instrumentation locations (A, B and C) – Schematic 155

Fig 7.6 Seismic Detector Response Curve for the Geophone Used in the Project 158

Fig 7.12 Absolute Radial Acceleration at A Corresponding to Probe Distance 164

Fig 7.14 Absolute Radial Acceleration at C Corresponding to Probe Distance 165

Fig 7.19 Comparison of the Simulated and Measured Post-Improvement Densities169

Fig 8.1 Composite Vibro-Stone Columns – Design Charts based on Soil Densities173

Notations

b = Factor representing influence of fine grains; One half of stone column spacing

Cv = Coefficient of consolidation; Coefficient of consolidation in vertical direction

(Dr,c)eq = Equivalent relative density (FC<FCth)

(Dr,f)eq = Equivalent relative density (FC<FCth)

Ec = Cumulative energy loss per unit volume of soil

EL = energy per unit volume required to cause liquefaction

e = Global void ratio

ec = Intergranular void ratio

(ec)eq = Equivalent intergranular void ratio

ef = Interfine void ratio

(ef)eq = Equivalent interfine void ratio

emax,hf = Maximum void ratio of host fines

emax,hs = Maximum void ratio of host sand

emin = Minimum void ratio

FCth = Threshold fines content

Hdr = Length of longest drainage path

Ir = Rigidity Index (G/Su)

kh = Hydraulic conductivity of soil in horizontal direction

ks = Hydraulic conductivity of soil

kw = Hydraulic conductivity of gravel wells

Lw = Coefficient of well resistance

m = A coefficient (0<m<1)

(N1)60 = Corrected SPT blow counts

(N1)60cs = Normalized clean sand equivalent SPT

NL = Number of cycles required for initial liquefaction (5% double ampl Ax Strain)

Nq = Normalized effective cone tip resistance (qc’/σv0’)

P0 = power rating of the vibratory probe

qc’ = Effective cone tip resistance

Rd = Size disparity ratio (D/d)

Re = Cavity Radius during expansion

r = Radial distance from the center of stone columns

(ru)avg = Average excess pore pressure ratio

(ru)max = The greatest excess pore pressure ratio

S = Normalized unit membrane compliance

Tbd = Time factor used in the numerical analysis

Tv = Time factor

ug = Excess pore pressure generated by ground motion or cavity expansion

ush = Shear induced excess pore pressure

W = Energy per unit time passing through a unit area of spherical surface at radius r

Wo = η0P0

w = Energy loss per unit time per unit volume of soil at distance r

Z = Vertical distance from the surface

Zdr = Vertical nearest distance to the drainage path

ψe = Pressure at the cavity wall

∆σz = Change in vertical total stress due to applied load

α = Constant, Coefficient of attenuation

γw = Unit weight of water

η0 = probe efficiency

σ’vo = Initial effective confining stress

σh0 = Insitu horizontal stress

σr = Stress in radial direction

σθ = Stress in angular direction

τs = Shear stress experienced by soil

ABSTRACT

Vibro replacement stone columns are in use to mitigate liquefaction hazards in sandy soils for almost three decades There are three mechanisms that help reduce liquefaction potential of a sandy soil improved using stone columns During stone column installation sandy soils densify due to installation vibration Further, the stiffness of the composite improved soil increases leading to a reduction in cyclic shear stress induced on the soil surrounding the stone columns during earthquakes In addition, pore pressures generated

in the soil during earthquakes are quickly dissipated through the highly permeable stone columns These combined mechanisms reduce the liquefaction potential of the improved soil Sandy soil sites improved using stone columns have performed well during earthquakes However, its effectiveness in silty soils is limited Recent case histories show stone columns supplemented with wick drains work well in such soils This study focuses on three aspects: (i) examining the reasons for the sub-performance of stone columns in silty soils, identifying key soil parameters that hinder the effectiveness of stone columns, and developing means to improve the effectiveness of this method in silty soils including provision of supplementary wick drains, (ii) developing a numerical model to simulate stone column installation with and without wick drains, and qualitatively evaluate the degree of ground improvement, and (iii) verifying the numerical simulation results using case histories and field experimental studies, and developing modified design charts and guidelines for designing stone columns with and without wick drains to improve sands and silty soils

Pore pressure generation, post-liquefaction dissipation, and densification characteristics

of an artificial silty soil and three natural silty soils were experimentally studied and compared with sand A careful analysis of such data indicates that liquefaction characteristics of silty soils and sands are not very different when compared using grain contact density indices as the basis for comparison However, post-liquefaction dissipation characteristics are very much dependent on grain size characteristics Low coefficient of consolidation associated with silty soils precludes faster pore pressure dissipation during stone column installation and therefore hinders densification around the stone columns during installation It also hinders drainage during earthquakes This appears to be the primary reason for the lack of effectiveness of stone columns in silty soils Numerical studies of pore pressure behavior of silty and sandy soils support this view

Based on the experimental results, a numerical model was developed to simulate the stone column installation process During installation, pore pressure generated due to the vibratory energy imparted into the surrounding ground was estimated, and the ground densification associated with pore pressure dissipation was calculated Several simulations were done for sands and silty soils with varying initial conditions improved using stone columns with and without wick drains The model was fine tuned and tested using case studies and field measurements

Design charts and design guidelines that were developed based on the extensive experimental and numerical study are presented Recommendations for improving the stone column design methodology, and for further research in this subject are presented

as well

1964 great Alaskan earthquake many highway and railway bridges were collapsed beyond repair mainly due to liquefaction and related ground deformations (Bartlett and Youd, 1992) Buckle (1995) reported of the distress to the bridge superstructure and collapse of several spans caused by lateral movement of several bridge piers due to lateral spread of the ground along the manmade waterfront area in Kobe Similar waterfront and transportation facility damage has been observed during the most recent Taiwan Chi-Chi

and Turkey earthquakes as well Such damages are reminders of the need to develop and

implement prudent remedial measures for liquefaction mitigation to retrofit transportation

and lifeline infrastructure systems

Many ground improvement techniques have been developed in the recent past (JGS

1996, Cooke and Mitchell 1999) These techniques can be categorized into four groups:

(a) densification (b) drainage, (c) reinforcement, and (d) cementation/solidification by

grouting (Table 1.1) Depending on site conditions, accessibility, sensitivity of adjacent

facilities, etc., each method has some advantages over the other

Table 1.1: Ground Improvement Techniques for Liquefaction Mitigation

Deep soil mixing Lime/cement injection

Densification technique using dynamic compaction requires open ground conditions and it is suitable for new construction sites Densification/drainage techniques using vibratory probe and grouting techniques do not require large open space and are suitable for retrofit applications for highway sites They are primarily suitable for sandy sites containing little or no silt Deep mixing and jet grouting techniques are suitable for silty soils as well Grout based techniques are usually more costly Reinforcement techniques have so far found limited applications to mitigate liquefaction They are often used for foundation stabilization (micropiles) or slope stabilization

1.1 Scope of this study

This thesis focuses mainly on one type of densification/drainage technique, namely vibro stone columns Vibro stone columns or vibro replacement stone columns are in use

to mitigate liquefaction hazards in sandy soils for more than twenty-five years (Dobson, 1987) In this technique a vibratory probe is inserted into the ground and vibrated Large gravel is fed into the ground upon retrieval of the probe Soil around the probe liquefies and is densified during this process Furthermore the ‘stone columns’ created act as reinforcements increasing the stiffness of the improved ground and reducing the magnitude of shear stress caused in the improved soil by an earthquake During an earthquake the pore pressures developed also dissipate through the stone columns All of the above processes reduce the liquefaction potential of the site Past experiences indicate that sandy soils improved using vibratory stone columns have performed well during earthquakes Mitchell et al (1995)

Densification/drainage techniques have found limited applications in silty soils Most recent attempts show that such techniques may be extended to silty soils when combined with other supplementary techniques such as prefabricated drains (Dise et al., 1994; Dumas et al., 1994; Luehring et al., 2000) This is a promising modification

However there are no guidelines or design methods available for such combined techniques At present they require pilot testing to study feasibility before making implementation decisions Rational design guidelines and design methods are needed Development of such design methods first requires a good understanding of the behavior

of silty soils compared to sands Pore pressure generation and densification behavior of such soils during installation of stone columns and during earthquakes in contrast with sands should be understood Based on such understanding quantitative design analysis methods need to be developed to assess pore pressure development, densification, and dissipation during stone column installation and during earthquakes when combined with supplemental wick drains Furthermore composite effect of stone columns on earthquake-induced shear stresses as compared to unimproved ground also need to be incorporated in such analysis methods These design methods must be evaluated and verified using field data so that they can be applied in practice with a higher degree of confidence Fig.1.1 schematically represents the various elements of this research

First a laboratory experimental study was conducted on four silty soils (a) Ottawa sand-silt mix (silt content varying from 0% to100% by dry weight), (b) Sandy silt from New Jersey area, (c) Sandy silt from San Fernando area, CA, and (d) Sandy silt from Los Angeles area, CA Undrained cyclic triaxial tests followed by pore pressure dissipation were carried out to determine the cyclic strength, pore pressure generation, pre- and post-

liquefaction compressibility, coefficient of consolidation, and densification characteristics of these soils The data was used to assess the influence of silt content on such characteristics compared to clean sand

Based on this study, it was found that it is possible to densify silty soils by expediting pore pressure dissipation using pre-installed wick drains during vibro stone column installation Since it is difficult to analytically determine the degree of improvement for such a composite system, a numerical model was developed to simulate vibro stone column (with or without pre-installed wick drains) installation process This model also capable of analyzing pore pressure changes within the improved soil in the event of an earthquake Several simulations were done to compare its results with those reported in the literature for clean sands with vibro stone columns alone, and found to be on good agreement There is only a limited amount of data available for composite stone columns (vibro stone columns supplemented with wick drains) installed in silty soil sites Therefore, a field study was conducted at a silty soil site, Marina del Rey, Los Angeles County, CA These field observations were used to refine the numerical model

Using the new numerical simulation program composite stone column design charts were developed Design guidelines and sample design problems are reported as well

Task I SOIL RESPONSE CHARACTERIZATION

Cyclic strength and liquefaction behavior of silty soils

Pore pressure Generation, Dissipation & Densification

Hydraulic Conductivity, Compressibility, Coefficient of

Consolidation

Task IIA THEORY & COMPUTATIONAL

DEVELOPMENT

Densification (During Installation)

Drainage

Reinforcement Effect

Field Tests

Task IIB ASSESSMENT

Task III DESIGN CHARTS AND DESIGN GUIDE LINES

Fig.1.1 Research Tasks

1.2 Organization of Thesis

The second chapter presents a review of ground modification systems for liquefaction mitigation and a detailed review of current design methods for liquefaction mitigation using stone columns It also presents a brief summary of case histories of liquefaction mitigation and identifies the applicability and limitations of the various methods for liquefaction mitigation in silty soils Chapter three presents a review of recent understanding of liquefaction behavior of silty soils in contrast with sands

Chapter four presents the details of the laboratory experimental program conducted

to study the pore pressure generation and dissipation characteristics of silty soils

The fifth chapter presents an analysis of the experimental data and identifies major similarities and differences in the behavior of silty soils compared to sands Attention is also drawn to possible differences in natural soils compared to Ottawa sand-silt mixes The sixth chapter presents details of the numerical model developed to analyze pore pressure generation, dissipation, and concurrent densification behavior of soils improved

by stone columns with and without supplementary wick drains It also presents a parametric study and comparison of pore pressure response of sandy and silty sites improved using stone columns with and without wick drains

Chapter seven presents the field studies conducted to verify the numerical model Chapter eight presents composite stone column design charts and design guidelines This chapter also includes sample design problems

Chapter nine includes the concluding remarks and recommendations for further research in this subject matter This chapter is followed by a list of references cited in this report

Appendices are also presented at the end of this thesis Appendix A presents sample interpretation of laboratory experimental data; Appendix B presents the summary of all the laboratory experimental data collected as apart of this study; Appendix C presents test summary and interpretation for each laboratory test; Appendix D presents the data collected from the field test carried out at Marina del Rey, CA; Appendix E presents the correlation between corrected SPT counts and relative density for sands reported by Tokimatsu & Seed (1984), which was used in the numerical studies reported in this thesis; and finally, the Appendix F presents a sample input files used in the numerical simulations

Chapter 2

REVIEW OF GROUND IMPROVEMENT

TECHNIQUES

2.0 Ground Improvement Techniques to Prevent Liquefaction

Liquefaction potential of a soil can be mitigated either by increasing the cyclic strength of the soil by means of densification and/or cementation and solidification, by shear stress redistribution, i.e reduction in cyclic shear stress caused by an earthquake by modifying the stiffness of the soil, or by controlling the excess pore pressures that develop during an earthquake by providing shorter drainage paths Accordingly, currently used ground improvement methods can be categorized into the following categories: (a) Densification, (b) Drainage, (c) Reinforcement, and (d) Stabilization by grouting or admixtures Table 2.1 provides a summary of techniques that are used for liquefaction mitigation, soil types that can be treated, their suitability for highway applications, and typical costs Figure 2.1 illustrates in general terms, the types of ground improvement that are useful relative to different soil particle ranges (Mitchell and Wentz, 1991) A brief overview of each technique is also presented following this figure

Andrus and Chung, 1995

Fig 2.1: Applicable Grain Size Ranges for Different Stabilization Methods

(Source: Mitchell and Wentz, 1991)

2.1 Densification Methods

Commonly used densification methods are: (i) vibro compaction, (ii) vibro replacement, (iii) vibro concrete columns, (iv) deep dynamic compaction, (v) deep blasting, (vi) compaction grouting, and (vii) displacement piles

2.1.1 Vibro Compaction

Vibro-compaction, also known as vibroflotation, is the rearrangement of soil particles into a denser configuration by the use of powerful depth vibrators (Fig 2.2) During vibration the soil surrounding the vibrator is liquefied and is rearranged into a

denser array with concurrent dissipation of pore pressures developed during vibration It

is effective in highly permeable sandy soils This method is not suitable when the fines content exceeds 15-25% (Adalier, 1996, Welsh et al.,1987; Drumheller et al., 1997) Table 2.2 provides a summary of two case histories, where vibro compaction was used to mitigate liquefaction induced hazards

Fig 2.2: Typical Procedure for Vibro Compaction (Source: Hausmann, 1990)

Table 2.2: Selected Vibro Compaction Projects as a Liquefaction Countermeasure

Very loose to med

dense hydr placed sand fill with some thin layers of silt and clay

GWT @ 5.5 ft

Treatment area: 10 ft beyond each bldg

Depth: 30 ft

6.5 ft on a grid

of equilateral triangles

Med dense hyd

placed sand fill with occasional thin layers of silts and clays

GWT @ 5 ft

Treatment area: 20 ft beyond the structure edge

Depth: 20 ft

8 ft on a grid of equilateral triangles

Medium

(Source: Adalier, 1996; Mitchell and Wentz, 1991)

2.1.2 Vibro Replacement Stone columns

Vibro replacement technique is an extension of the vibro compaction method In this method vibrators are used to displace the deep soil laterally First the vibratory probe is inserted to the bottom elevation of the soil to be improved by vibration and jetting The vertical hole created by the vibrators is backfilled by gravels in increments of 1.5 to 2.5

ft, and the backfill is compacted by the vibrating probe, which simultaneously displaces the material radially into the soil until desired densification is achieved (Dobson, 1987; Mitchell and Wentz, 1991) Figure 2.3 illustrates a stone column-installation procedure

In a bottom-feed stone column construction method, the gravel backfill is discharged at the bottom of the hole through a pipe attached to the side of the vibroflot

Backfill

Loose granular soil

Stone Column

Vibrator

Fig 2.3: Vibro Replacement Stone Column Technique (Source: Dobson, 1987)

In addition to densification, the stone columns also provide drainage pathways for faster dissipation of excess pore pressures developed during installation as well as during earthquakes Meantime, the stone columns also function as reinforcing elements Therefore, the range of soils that can be improved using this technique is larger compared

to vibro compaction Figure 2.4 shows such a range along with gradations of sandy soils liquefied during past earthquakes

Fig 2.4: Range of Soils Treated by Vibro Stone Columns

(Source: Dobson, 1987)

Stone Columns have been in use to reduce liquefaction-induced hazards since 1974, when this technique was first utilized for prevention of liquefaction at Santa Barbara, CA Dobson (1987) reports several case histories, which give evident for the usage of stone columns in sands as well as silty soils to mitigate liquefaction hazards Post-modification evaluations indicated satisfactory improvement, though real performance would vary

during earthquakes Mitchell et al (1995) report that the sites improved by stone columns performed well during earthquakes However, attention should be given to that most of the sites reported have experienced lower levels of ground shaking compared to design levels Table 2.3 provides general information of some of the sites that were improved by vibro stone column technique Table 2.4 gives the performance details of these sites during earthquakes

Luehring et al (2000, 2001) report a recent case history where prefabricated vertical drains were used to enhance densification during vibro replacement stone columns installation in a test section consisting of low permeable silty soils Field observations and post-improvement evaluations indicated that the effectiveness of stone columns can

be increased in silty soils when combined with supplementary wick drains

Further analysis and a review of design methods used for stone columns are provided

in the chapter 2.5