3er Simposio Internacional de Cimentaciones Profundas 05-11-2015)

Volume Prepared by ISSMGE Technical Committee - 214 Volumen preparado por el Comité Técnico TC-214 de la ISSMGE

For / para

3rd International Conference on Deep Foundations Deep Foundations and Soil Improvement in Soft Soils

3er Simposio Internacional de Cimentaciones Profundas Cimentaciones Profundas y Mejoramiento Masivo en Suelos Blandos

November 11-12th, 2015, Mexico City

Edited by / Editado por Norma Patricia López Acosta

Technical Committee

TC-214

Copyright, México, 2015 Sociedad Mexicana de Ingeniería Geotécnica, A.C. Valle de Bravo No. 19 Col. Vergel de Coyoacan, 14340 México, D.F., MÉXICO Tel. +(52)(55)5677-37-30, Fax+(52)(55)5679-36-76 Página web: www.smig.org.mx Correo electrónico: [email protected]

Prohibida la reproducción parcial o total de esta publicación, por cualquier medio, sin la previa Autorización escrita de la Sociedad Mexicana de Ingeniería Geotécnica, A.C. Total or partial reproduction of this book by any medium requires prior written consent of the Sociedad Mexicana de Ingeniería Geotécnica, A.C. Las opiniones expresadas en este volumen son responsabilidad exclusiva de los autores. Opinions expressed in this volume are the sole responsibility of their authors.

Collaborators (Editing and Formatting)/ Colaboradores (Edición y Formato): A.R. Pineda Contreras, E. Martínez Hernández y A.L. Espinosa Santiago. ii

SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C. COUNCIL OF HONOR / CONSEJO DE HONOR Leonardo Zeevaert Wiechers† Raúl J. Marsal Córdoba† Alfonso Rico Rodríguez† Enrique Tamez González Guillermo Springall Caram Edmundo Moreno Gómez Carlos Jesús Orozco y Orozco† Luis Vieitez Utesa Gabriel Moreno Pecero Raúl López Roldán Raúl Flores Berrones Luis Miguel Aguirre Menchaca† Gabriel Auvinet Guichard Luis Bernardo Rodríguez González Raúl Vicente Orozco Santoyo Alberto Jaime Paredes Mario Jorge Orozco Cruz Juan Jacobo Schmitter Martín del Campo Héctor M. Valverde Landeros

CONSULTIVE COUNCIL / CONSEJO CONSULTIVO José Francisco Fernández Romero Rigoberto Rivera Constantino Walter Iván Paniagua Zavala Juan de Dios Alemán Velásquez David Yáñez Santillán

BOARD / MESA DIRECTIVA 2015-2016 Raúl Aguilar Becerril President Norma Patricia López Acosta Vice-President Carlos Roberto Torres Álvarez Secretary Celestino Valle Molina Treasurer María del Carmen Suarez Galán Nilson Contreras Pallares Miguel Figueras Corte Aristóteles Jaramillo Rivera Technical Assistants iii

SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C. ORGANIZING COMMITTEE / COMITÉ ORGANIZADOR 3rd International Conference on Deep Foundations / 3er Simposio Internacional de Cimentaciones Profundas

iv

Walter I. Paniagua Zavala

ISSMGE

Juan Paulín Aguirre

ISSMGE

Norma Patricia López Acosta

SMIG

Mary Ellen Large

DFI

Theresa Engler

DFI

Vernon Schaefer

G-I

2015

November 11-12th, Mexico City


vi

Foreword On behalf of the Technical Committee TC-214 of the International Society for Soil Mechanics and Geotechnical Engineering (ISSMGE), it is a privilege to present this volume for the 3rd International Conference on Deep Foundations (Deep Foundations and Soil Improvement in Soft Soils), held in Mexico City, November 11-12th, 2015. This time, four organizations have joined efforts to produce it: the above mentioned TC-214 (Foundations Engineering for Difficult Soft Soil Conditions), the Mexican Society for Geotechnical Engineering (SMIG, which hosts the TC-214), the Deep Foundations Institute (DFI), and the GeoInstitute of ASCE. In two previous events, SMIG and DFI had collaborated in 2011 and 2013 to organize the First and Second International Conference on Deep Foundations, with very good acceptance in the geotechnical community. The purpose of merging different entities is multiple: to foster collaboration between countries, to continue the technological and scientific knowledge transference, and to promote different points of view from geotechnical professionals, including academicians, consultants, contractors and equipment manufacturers. Therefore, the material presented hereby, includes a wide spectre of the deep foundations and soil improvement current knowledge, with special emphasis in soft soils. Three main topics are recognized: Deep foundations, Excavations, and Soil Improvement. From state of the art of geotechnical research to case histories, the papers presented herein give a general -and presentperspective on this matter. My gratitude to all attendees, speakers, exhibitors and members of the Organizing Committee, for their interest, collaboration and hard work in this event.

Walter I. Paniagua TC-214, Chair Pilotec, SA de CV

vii

viii

Introduction The interest for constructing high-rise buildings in urban zones and the necessity to build structures in difficult subsoil conditions requires engineers to look for efficient solutions. Scenarios are challenging the projected structures that are affected by significant natural forces, such as those imposed by wind, earthquakes or sea waves. The construction of deep foundations and the soil improvement works have proved to be efficient alternatives to handle these situations. In many cases, especially when loads and mechanical elements are of important magnitude, the design of deep foundations is mandatory. For example deep foundations are employed when weak or otherwise unsuitable soil exists near the subsurface and the vertical loads must be carried to depth deposits. Deep foundations have other uses, for example they are used to resist scour, sustain axial loading by side resistance in strata of granular soils or competent clay, allow above-water construction, support lateral forces, improve the stability of slopes, reduce settlements and other special purposes. The most used deep foundations are driven piles and drilled shafts. In other cases, when the resistance or deformability conditions of soils are not allowable for the project, the use of techniques for soil improvement are required. They help to reduce total or differential settlements, increase axial and lateral bearing capacity and, in some cases, help to avoid an undesirable soil behavior, such as liquefaction, swelling, among others. The 3rd International Conference on Deep Foundations (3rd ICDF), held for the third time in Mexico City, is a space to present recent experiences related to deep foundations and in this occasion it includes the topic of massive improvement in soft soils. The aim of this conference is to promote the most recent technical and scientific developments and to share experiences in the design and construction of deep foundations and improvement techniques of soils. The papers received for the 3rd ICDF include subjects such as cases history, foundations for high-rise towers, geo-construction techniques, special deep foundations, cases of improvements on different subsoils and deep excavations, among others. There is no doubt that the lectures on these topics will also increase our knowledge of soil behavior. My sincere acknowledgment to all authors for their invaluable contributions as well as to the Organizing Committee for their efforts to achieve a successful Conference. The Mexican Society for Geotechnical Engineering is proud to hold the 3rd ICDF on Mexico City and will collaborate continuously with the Deep Foundations Institute, the Geo-Institute and the International Society for Soil Mechanics and Geotechnical Engineering through its ISSMGE Technical Committee TC-214, in order to promote the dissemination of the geotechnical knowledge.

Raúl Aguilar Becerril Presidente SMIG – Mesa Directiva 2015-2016

ix

3rd

International Conference on Deep Foundations Deep Foundations and Soil Improvement in Soft Soils

Technical Committee

TC-214

Contents Page Foreword............................................................................................................................

vii

Introduction........................................................................................................................

ix

SESSION 1. DEEP FOUNDATIONS

1

Effects of varved deposit on driven piles at a LNG terminal site Efectos de depósitos estratificados en pilotes hincados en el sitio de la terminal LNG LIN Guoming, HUANG Yanbo & LIN Cheng.............................................................. Low noise and low vibration press-in piling method in soft soil in congested urban areas Método de piloteo de baja presión de vibración y bajo ruido en suelos blandos en áreas urbanas congestionadas TAKUMA Takefumi..................................................................................................... The use of displacement piling technology in soft soil conditions El uso de tecnología de pilotes de desplazamiento en condiciones de suelo blando MARINUCCI Antonio & CHIARABELLI Marco......................................................... Rescate de una cimentación de pilas con inclusiones rígidas Pile foundation retrofit with rigid inclusions SEGOVIA José, PANIAGUA Walter y LÓPEZ Germán............................................... Foundation design and construction for high-rise Towers in Mexico City Diseño de la cimentación y construcción de Torres de gran altura en la Ciudad de México DEMING Peter W., NIKOLAOU Sissy, POLETTO Raymond J. & TAMARO George J..................................................................................................... Deep foundations in Mexico City soft soils Cimentaciones profundas en suelos blandos de la Ciudad de México AUVINET-GUICHARD Gabriel & RODRÍGUEZ-REBOLLEDO Juan-Félix.............. The use of micropiles technology in soft soil conditions El uso de tecnología de micropilotes en condiciones de suelo blando PAGLIACCI Federico.................................................................................................

SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

3

11

19

31

39

51

67

xi

3rd International Conference on Deep Foundations

Contents

Deep Foundations and Soil Improvement in Soft Soils

Page Pore pressure build-up due to pile driving in clayey deposits Desarrollo de presión de poro debido al hincado de pilotes en depósitos arcillosos MENDOZA Manuel J., RUFIAR Miguel, IBARRA Enrique & OROZCO Marcos........................................................................................................... Geotechnical design of the foundation for an office building located at the transition zone Diseño geotécnico de la cimentación para un edificio de oficinas localizado en la zona de transición ARENAS Fernando & CUEVAS Alberto.........................................................................

SESSION 2. EXCAVATIONS

77

85

91

A historic capitol and a deep excavation Un capitolio histórico y una excavación profunda MASSOUDI Nasser & SLIWOSKI Richard.................................................................... The support of a 25 m deep excavation in difficult ground conditions using Single Bore Multiple Anchor technology Soporte de una excavación de 25 m de profundidad en condiciones de terreno difícil usando tecnología de anclaje múltiple con barreno único MOTHERSILLE Devon & OKUMUSOGLU Bora......................................................... The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas El uso de muros MSE rellenados con concreto celular ligero en áreas sísmicas de terrenos blandos PRADEL Daniel & TIWARI Binod.................................................................................

107

SESSION 3. SOIL IMPROVEMENT

115

Soil improvement around the world – Applications and solution examples Mejoramiento de suelos alrededor del mundo – Aplicaciones y ejemplos de solución GERRESSEN F............................................................................................................... Principles and application of soil mixing for ground improvement Principios y aplicación de la técnica soil mixing para mejoramiento de suelo WILK Charles M............................................................................................................. Sustitución dinámica aplicada en turbas de la península de Yucatán Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan CIRION ARANA Alfredo, CHATTE Rémi & PAULÍN AGUIRRE Juan......................... Transforming marginal land to support a world class development in Panama Modificación de suelos marginales para apoyar proyectos de clase mundial en Panamá LANGONI Gustavo & ARCHABAL Roger.....................................................................

xii


93

97

117

123

127

131

3rd.


Session 1: Deep foundations

Technical Committee

TC-214

Technical Committee

TC-214 3ER SIMPOSIO INTERNACIONAL DE CIMENTACIONES PROFUNDAS

Sociedad Mexicana de Ingeniería Geotécnica

Noviembre 11-12, 2015 – México, D. F.

Effects of varved deposit on driven piles at a LNG terminal site Efectos de depósitos estratificados en pilotes hincados en el sitio de la terminal LNG Guoming LIN1, Yanbo HUANG2, and Cheng LIN 3 1

Senior Consultant, Terracon, 2201 Rowland Avenue, Savannah, GA 31404; [email protected] Geotechnical Engineer, Terracon, 2201 Rowland Avenue, Savannah, GA 31404; [email protected] 3Project Geotechnical Engineer, Terracon, 2201 Rowland Avenue, Savannah, GA 31404; [email protected]

2Staff

ABSTRACT: More than 6000 steel pipe piles are required for a proposed large LNG (Liquefied Natural Gas) Terminal project. The regasification facility includes three 3.5 billion cubic feet (100 million cubic meters) double-wall storage tanks for a daily send-out capacity of 1.2 billion cubic feet (33 million cubic meters) and a pier designed to berth ships with a capacity of 200,000 cubic meters. The subsurface conditions at the tank locations were explored with a combination of 12 soil test borings (STB), 34 cone penetration test (CPT) soundings and eight dilatometer test (DMT) soundings. The geotechnical study also included field vane shear testing, pore pressure dissipation testing and laboratory testing. The site subsurface conditions feature a layer of 90-foot thick very soft to stiff clayey silts with interbedded thin sand seams (varved deposit). Characterization of this varved deposit layer, especially its shear strength and preconsolidation history, is critical to the foundation design and construction for this project. However, the unique structure of the clayey silts presents difficulties in defining some of its properties such as time rate of consolidation and undrained shear strength. This paper presents the subsurface exploration program and the methods used to characterize the clayey silts from the field and laboratory testing results. The preconsolidation history of this layer was evaluated using several different approaches. This paper discusses the potential downdrag force and its implication in the pile design and construction. A statistical procedure is developed to analyze axial pile capacities using SPT and CPT based methods and pile capacities obtained from different methods are compared and discussed.

1 INTRODUCTION 1.1 Project information The proposed LNG Terminal includes three 150,000 cubic meter (944,000 barrel) double wall insulated LNG storage tanks, process equipment consisting of compressors and vaporizer, buildings, pipelines, impoundment dikes, roads, and a parking lot. A jetty will be built for unloading LNG tankers and a breakwater may be built to provide a sheltered area for the tanker. The tanks are designed to store liquefied natural gas (LNG) at a pressure of 2.0 psig and a temperature of -270°F. The tanks will have an outer concrete wall (122 feet in inner radius) and an inner steel tank. The project will require a permit from the Federal Energy Regulatory Commission (FERC). 1.2 Site description The site is located on the southeast bank of the Delaware River in Logan Township, Gloucester County, New Jersey. The property, approximately 175 acres, is primarily an agricultural soybean field

with several gas and liquid petroleum pipe lines that traverse the Delaware River and make landfall on the northwestern end of the property. The site is generally flat and had been used for disposing of dredge spoil from the Delaware River before the 1960s. 1.3 Geotechnical testing The subsurface conditions of the site were explored with a combination of 12 soil test borings (STB), 34 cone penetration test (CPT) soundings, and eight dilatometer test (DMT) soundings. Field vane shear tests were performed at three STB locations. In conjunction with the CPT soundings, pore pressure dissipation tests were performed at various depths within four CPT locations were measured at the center of the three tanks. The laboratory testing program consisted of soil index testing, consolidation and triaxial shear strength testing, and chemical analyses.


4

Effects of varved deposit on driven piles at an LNG terminal site

1.4 Subsurface Stratigraphy The subsurface stratigraphy is generalized in the following table.

sizes and thickness. The CPT soundings take readings at 2-centimeter intervals, which is relatively accurate in determining the interfaces of the soil strata. Figure 2 is a contour map showing the bottom elevation of the clayey silt layer varying approximately from -86 to -93 feet (NAVD) under Tank 1.

Table 1. Generalized Subsurface Stratigraphy. Layer No.

Soil Type

1

Dredged Fill

2 3

Sand/Gravelly Sand Clayey Silt with Interbedded Sand Seams Sand and Gravel Residual Clayey sand

4 5

Average Elevation (ft, NAVD) 7 to 3

H* (ft)

Geologic Period

4

Recent Dredge Spoil Quaternary Quaternary

3 to -3 -3 to -93

6 90

-93 to -113 -113 to -148

20 35

6 Metamorphic Rock Below -148 *H is the average thickness of soil layer

Quaternary Tertiary and Cretaceous

--

Layer 3, termed varved deposit, has a thickness of 85 to 95 feet and contains clayey silts with numerous interbedded thin fine sand seams as shown in Figure 1. Geologically, the soils were deposits of recent age as a result of warmer temperatures and rise of ocean levels after the Glacial Period. The fine grained soils became fertile ground for vegetation which resulted in variable amounts of organics within this layer. Due to its great thickness, the shear strength of this layer can greatly affect the pile capacities. Furthermore, potential downdrag force is a concern if the layer is underconsolidated or will undergo additional settlements from the surface loads.

Figure 2. Bottom of the Varved Clay Layer.

2 CHARACTERIZATION OF THE VARVED DEPOSIT (LAYER 3) 2.1 Soil index properties and classification The Layer 3 soils are mostly classified as low plasticity silts (ML) and high plasticity silts (MH) with occasional classification of high plasticity clays (CH) or clayey sands (SC). Table 2 summarizes the soil index and classification properties. Table 2. Summary of Soil Index and Classification Properties.

Range Average

# 200 Passing (%)

Natural Moisture Content (%)

Liquid Limit

Plastic Limit

Organic Content (%)

50~98 83

20~90 55

27~94 54

12~50 28

1.7~7.1 3.7

2.2 Consistency

Figure 1. Photo of Clayey Silt with Interbedded Sand Seams.

SPT blow count, CPT tip resistance, and DMT modulus were used to characterize the consistency of the clayey silt. In general, the clayey silt exhibited slightly increasing consistency with depth from soft at the top to firm near the bottom.

Samples taken from the SPT samplers and Shelby tubes allow visual observations of characteristics of the sand seams, such as the depth intervals, particle SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

LIN G. et al.

5

can be considered to increase from 400 psf at the top of clayey silt to 1100 psf at the bottom of the clayey silt with an average value of 750 psf.

Figure 3. Unit Weight with Depth.

2.3 Unit weight

Figure 4. Undrained Shear Strength versus Depth (Exponential Regression).

As shown in Figure 3, the unit weight derived from DMT soundings agrees reasonably well with the results of laboratory tests performed on Shelby tubes. The total unit weights were found to increase slightly with depth in the clayey silt layer from 100 pcf at the top to 110 pcf at the bottom.

2.5 Compressibility

2.4 Undrained shear strength The undrained shear strengths of the clayey silts were obtained using three different methods: laboratory triaxial tests, field vane shear tests, and correlations from CPT data. Due to the effect of thin sand seams within the test specimens, almost all unconsolidated undrained (UU) triaxial tests resulted in a sloped failure envelope rather than a horizontal failure envelope typical for normally consolidated clays. The undrained shear strength was then interpreted as the shear stress corresponding to the effective in-situ overburden stress on the failure envelope. The uncorrected undrained shear strength obtained from field vane shear tests was approximately two times as large as the undrained shear strength from the UU triaxial tests. Using the Correction procedures of Bjerrum (1972) as revised by Aas et al. (1986), the corrected vane shear strength values agree reasonably well with the laboratory test results, as shown in Figure 4. The trend of undrained shear strength increasing with depth can be approximated exponentially or linearly. In CPT soundings, the shear strength is related to the cone tip resistance by a cone factor Nkt. An Nkt value of 15 to 18 was obtained by matching the undrained shear strength derived from the CPT data to the best fitted curve from Figure 4. Previous studies by others indicated the cone factor Nkt generally ranges between 15 and 20 (ESOPT 1974 and 1982, ISOPT 1988). Undrained shear strength

Laboratory consolidation tests were performed using both conventional incremental loading procedures (ASTM D-2435) and constant strain rate (CSR) method (ASTM D-4186). The conventional consolidation tests yielded an average compression index (Cc) of 0.566 with an average initial void ratio (eo) of 1.535, which corresponded to an average compression ratio [Cc/(1+eo)]of 0.223. The CSR consolidation tests measured an averaged compression ratio of 0.218. Constrained modulus of compression (M), derived based on the empirical correlations with DMT data (Schmertmann, 1988) and CPT data (Senneset et al., 1989 and Kulhawy and Mayne, 1990), varied approximately between 25 and 60 tsf for the clayey silts. The constrained modulus derived from the three CSR consolidation tests averaged about 38 tsf. 2.6 Time rate of consolidation The time rate of consolidation was characterized using laboratory consolidation tests and in-situ CPT pore pressure dissipation tests. Theoretically, the drainage path for the laboratory consolidation tests is in the vertical direction while the pore pressure dissipation tests measure pore pressure dissipation in the horizontal direction. The vertical coefficient of consolidation (CV) measured in the consolidation test, varied between 0.1 and 0.5 ft2/day around the in-situ overburden stress. The coefficients of consolidation in the horizontal direction, Ch, were calculated based on a method proposed by Mayne (2002). The Ch, values varied from 0.41 ft2/day to 245 ft2/day. The large variation of the Ch values


6


suggests that the rate of pore pressure dissipation (i.e. consolidation) is greatly affected by the interbedded sand seams within the clayey silts 3 PRECONSOLIDATION HISTORY OF THE VARVED DEPOSIT (LAYER 3) The consolidation history of the varved deposit, i.e., whether the clayey silts are underconsolidated, normally consolidated or overconsolidated, is essential for the foundation design of this project. This condition will affect the magnitude of settlement of fills and shallow foundations and whether to consider downdrag forces on deep foundations. Due to the great thickness of the clayey silt layer and large number of piles to be used on this project, this consideration has substantial impact on the design of this project. Geologically, the clayey silts (varved deposit) were deposited along the shores of the Delaware River during the late Pleistocene Period. The deposition of the layer was a long slow process taking place more than ten thousand years ago. It is logical to assume that the silts and clays had consolidated under the self-weight of the material. The dredged fills were last deposited at the site in the 1960s. From the geotechnical standpoint, the clayey silts are more than 80 feet in thickness, which would require a long time to consolidate. However, the interbedded sand seams would function as horizontal drainage paths to facilitate consolidation. As such, the key question is if the consolidation of the clayey silts has completed under the weight of the dredged fill placed more than 30 years ago. Several different approaches were taken in evaluating the preconsolidation history of the clayey silt layer at the site.

curve using linear regression intercepts the strength axis at 75 psf (very small). The average strength gain is 15.9 psf per foot. Dividing the average strength by the average effective vertical stress gives a c/p ratio of around 0.37, a reasonable value for a normally consolidated or slightly overconsolidated silt. 3.3 Overconsolidation ratio In laboratory consolidation tests, smaller load increments were added in the vicinity of the existing overburden pressure to fine-tune the compression curves for the determination of preconsolidation pressure. The preconsolidation stresses, determined using the Casagrande procedures, indicated OCR values ranging between 0.9 and 2.1 with an average of 1.26. The OCR values were also derived from both CPT and DMT using empirical correlations (Powell and Cuarterman, 1988; Kamei & Iwasaki, 1995). The OCR of the clayey silts derived from DMT and obtained through consolidation tests fall mostly between 1 and 3, as shown in Figure 5. Therefore, the clayey silts are considered normally to slightly over-consolidated based on the overconsolidation ratio

3.1 Atterberg limits The relationship between the natural moisture contents and Atterberg Limits can be used as an approximate indication of soil’s preconsolidation history. Moisture contents that are well above the liquid limit at depths of tens of feet usually indicate underconsolidation. Moisture contents near the plastic limit at shallow depths usually indicate overconsolidation. A statistical analysis of the laboratory test results performed on 30 Shelby tube samples indicated an average natural moisture content of 55 percent and an average liquid limit of 54 percent. The natural moisture contents are very close to the liquid limits. These properties lead to the conclusion that the soil is not significantly overconsolidated or underconsolidated. 3.2 Undrained shear strength The undrained shear strengths generally increase with depth (effective vertical stress). The best fit

Figure 5. OCR from DMT and Consolidation Test.

3.4 Consolidation theory The Cv and Ch values obtained from the laboratory consolidation tests and derived from the field pore pressure dissipation tests vary greatly with depths and locations. The thickness of clay layer between two drainage paths (sand seams) also vary greatly. Therefore, a conservative model was used to predict the time required for the consolidation of the clayey silts under the weight of the dredged fill. Using the Terzaghi’s one dimensional consolidation theory and a Cv of 0.01 ft2/day and a layer thickness of 10 feet between top and bottom drainage paths (both values are considered conservative based on the field and


7

LIN G. et al.

laboratory test results), a time of 5.8 years is required to reach 90 percent consolidation. Considering the dredged fill was last placed on site more than 30 years ago, it is reasonable to believe that the clayey silts had consolidated under the weight of the dredged fill. 4 PILE DESIGN CONSIDERATIONS Steel pipe piles without concrete fill are considered the most suitable foundation system for this project (Yang et al., 2003). All piles are required to be embedded into the sand and gravel (Layer 4). Preliminary design performed by the design engineer (CB&l) requires an ultimate axial compression capacity of 200 tons and 280 tons for 18 and 22 inch diameter piles, respectively.

based on the factor recommended by American Petroleum Institute (API, 1993). For the clayey silt layer defined in the Driven program, undrained shear strength of 500 and 1000 psf, corresponding to adhesion values of 400 and 800 psf based on Tomlinson’s (1980) method, were used as the lower and upper bound values, respectively. Figure 8 presents comparison of the ultimate axial capacities using the two CPT based methods versus the Driven program. For the CPT based methods, the average pile capacities obtained through the statistical procedure are presented. It appeared the ultimate axial pile capacities calculated from the three different methods agree relatively well with each other. Back-calculation based on the pile capacity vs. depth curves indicated the average adhesion values of the clayey silt were 500 and 750 psf for the French Method and the Eslami & Fellenius Method, respectively.

4.1 Axial Pile capacities 0

Eslamic & Fellenius (1997) Direct CPT Method, 10 to 90% of Normal Distribution

20

40

Depth (ft)

The axial capacities will be largely dependent on the adhesion between the pile and the clayey silt layer (Layer 3). Several methods were used to estimate the axial capacities of driven piles under static loads. These methods included calculating adhesion and friction values between pile and soil based on the soil strength from laboratory and field testing, directly from the CPT results, and based on SPT blow counts using a computer program FHWA Driven 1.2 (2001). Cone penetration tests with pore pressure measurements (CPTu) are considered probably the best in-situ test method for the design of axially loaded piles (Hannigan, et al. 1997). Various calculation methods based on CPT data were reviewed and compared, and two methods were selected to estimate axial pile capacities for this project: the method developed by Bustamante and Giasenelli (1982), also called the LCPC method or the French method and the method proposed by Eslami and Fellenius (1997). Before the calculation, the depths associated with CPT data were adjusted to a ground surface at 6 feet NAVD. To account for the variations of the soil conditions at this site, a statistical analysis was performed by assuming pile capacities based on the CPT at different soundings would have a normal distribution at the same depths. The procedure generated an upper bound, lower bound and an average pile capacity at a given depth by eliminating the samples which significantly deviate from the main group (more specifically, values of probability density function less than 10 percent). Figures 6 and 7 present the ultimate pile axial capacities for 18-inch steel pipe piles calculated using the two CPT methods and the statistical procedure. An average undrained shear strength of 750 psf was derived from the laboratory and field tests, which corresponds to an adhesion of approximately 656 psf

60

80

100

120 0

50

100

150

200

250

300

350

400

450

500

Ultimate Compression Capacity (tons)

Figure 6. Ultimate Compression Capacities for 18-in Steel Pipe Pile Using CPT methods by Eslamic & Fellenius method.

An ultimate compression capacity of 200 to 280 tons was recommended for 18-inch and 22-inch OD pipe piles, respectively, after the piles are driven into the sand and gravel layer. The pile length will vary with elevations of the top of the sand and gravel layer as well as the depth of penetration into this layer. A pile tip elevation between -95 and -108 feet NAVD can be expected for the purpose of preliminary design and estimates.


8


the piles to be embedded into the sand and gravel layer (Layer 4) even though the piles may achieve the required pile capacities with the pile tips above this layer. 5 SUMMARY AND CONCLUSIONS

Figure 7. Ultimate Compression Capacities for 18-inch Steel Pipe Pile by French Method.

1. The subsurface conditions for the proposed BP Crown Landing LNG terminal site were explored by using a combination of 12 soil test borings, 34 cone penetration test soundings and eight dilatometer soundings. The site features a thick clayey silt layer with interbedded fine sand seams. The strength and compressibility characteristics of this clayey silt layer have a major impact on the costs and safety of the pile foundation. The soil properties were tested using a series of field vane shear tests, pore pressure dissipation tests and laboratory tests. 2. The combined use of SPT, CPT and DMT soundings was a well thought-out choice for subsurface exploration. The samples from the SPT samplers and Shelby tubes allowed the engineers to closely examine the characteristics of the interbedded layers, such as the depth intervals, thickness and particle sizes of the sand seams. CPT soundings provided more accurate determination of the depth of soil interface and continuous data for deriving other engineering properties and subsequent pile capacity calculations.

Figure 8. Comparison of Compression Capacities by Different Methods.

4.2 Downdrag Force The preliminary grading plan shows the final grade in the tank areas will be at or near the existing site grade. There will be minimal increase of stresses in the soils from the fill and grading in the tank area. Based on analyses presented in the previous section, it was concluded that the clayey silts are normally consolidated. No downdrag force was considered for the piles underneath the tanks. However, there are uncertainties associated with potential secondary consolidation and long-term decomposition of organic materials in the soils. The risk associated with the above uncertainties is considered relatively small. For the piles embedded in the sand and gravel layer (Layer 4), there is an extra 100 to 150 ton compression capacity available to offset this highly unlikely but potential downdrag force. The Construction quality control program will require all

3. The undrained shear strength of the clayey silts was determined using three different methods: field vane shear tests, laboratory triaxial tests and empirical correlations from the CPT soundings. The results from different methods agree reasonably well after proper corrections. 4. The consolidation history of the clayey silts was analyzed qualitatively based on the geological and geotechnical considerations and quantitatively using laboratory and field test results. The analyses consistently indicate that the clayey silts are normally consolidated or slightly overconsolidated. No downdrag force was considered necessary for piles supporting the tanks, where no stress increase from fill or grading is anticipated. 5. Axial pile capacities were calculated using three different methods: based on adhesion and friction values of the soils, directly calculated from the CPT data and SPT blow count values. Results from these methods were evaluated and compared. Cone penetration test data was considered the best method in calculating pile capacities. A statistical analysis procedure was developed to account for the variations of the soil conditions to present the pile capacities in upper and lower bounds.


LIN G. et al.

6 ACKNOWLEDGEMENTS The authors would like to acknowledge the following organizations and individuals for their help and support for this project: BP and Mr. Junius Allen, CB&I and Messrs Greg Bertha and Donald Barrs, Dr. Felix Yokel, Site Blauvelt Engineers, Geotesting Express, and many current and former Terracon/WPC colleagues Messrs. William Wright, Wu Yang, Edward Hajduk, Jian Fang, Thomas Casey, William Anderson and Donovan Ledford.

9

Powell, J. J. M., & Quarterman, R. S. T. (1988). “The Interpretation of Cone Penetration Tests in Clays, with Particular Reference to Rate Effects”, Penetration Testing, Balkema, Rotterdam, The Netherlands, Vol. 2: 903-909. Tomlinson, M. J. (1980). Foundation Design and Construction, Pitman, London, UK Yang, W., Fang, J., and Lin, G.M. (2003). WPC’s Geotechnical Report Liberty LNG Project. Savannah, Georgia:

REFERENCES Aas, G., Lacasse, S., Lunne, T., & Hoeg, K. (1986). “Use of In-situ Tests for Foundation Design on Clay”, Publikasjon-Norges Geotekniske Institutt, (166): 1-15. American Society for Testing and Materials (2004). Standard Test Methods for One-Dimensional Consolidation Properties of Soils Using Incremental Loading, ASTM D-2435. American Society for Testing and Materials (2006). Standard Test Method for One-Dimensional Consolidation Properties of Saturated Cohesive Soils Using Controlled-Strain Loading, ASTM D4186. API (1993). “Recommended Practice for Planning, Designing and Constructing Fixed Offshore Platforms – Working Stress Design”, 20th Edition, American Petroleum Institute, Washington, DC. Bjerrum, L. (1972). “Embankments on Soft Ground”, Proceedings of Performance of earth and earthsupported structures, ASCE. 1-54. Bustamante, M., & Gianeselli, L. (1982). “Pile Bearing Capacity Prediction by Means of Static Penetrometer CPT”, Proceedings of the Second European Symposium on Penetration Testing. 493-500. Eslami, A., & Fellenius, B. H. (1997). “Pile Capacity by Direct CPT and CPTu Methods Applied to 102 Case Histories”, Canadian Geotechnical Journal, Vol.34(6): 886-904. Federal Highway Administration’s Driven 1.2 computer program (2001). Blue-Six Software, Inc. Hannigan, P. J., Goble, G. G., Thendean, G., Likins, G. E., & Rausche, F. (1997). “Design and Construction of Driven Pile Foundations-Volume I & II”, FHWA-HI-97-014, Washington, D.C. Kamey, T. & Iwasaki, K. (1995). “Evaluation of Undrained Shear Strength of Cohesive Soils Using a Flat Dilatometer”, Soils and Foundations, Vol 35(2): 111-116. Mayne, P. W. (2002). “Equivalent CPT Method for Calculating Shallow Foundation Settlements in the Piedmont Residual Soils Based on the DMT Constrained Modulus Approach” from http://www. geosystems.ce.gatech.edu/~geosys/.


Technical Committee




Low noise and low vibration press-in piling method in soft soil in congested urban areas Método de piloteo de baja presión de vibración y bajo ruido en suelos blandos en áreas urbanas congestionadas

Takefumi TAKUMA1 1Giken

America Corp.

ABSTRACT: Pile driving causes ground vibrations and noise by pile driving machines used during construction. Conventional pile driving has all but disappeared from urban construction because of the emissions of deafening noise and earth shattering vibrations. The “Press-in Piling Method” was developed to solve most of the problems associated with pile driving in urban construction. This innovative piling technique allows sheet piles and other types of prefabricated piles or panels to be hydraulically pressed into the ground using the reaction force generated by the previously installed piles’ surface friction with soil and the system’s own weight. The Press-in pile driving machine ‘walks’ on top of the pile wall, gripping on previously installed piles while installing the next pile immediately adjacent to the one just installed. While the method is highly suited for soft ground, the system can also efficiently deal with hard soil, such as dense sand, stiff clay, gravels, cobbles, boulders and soft rock with attachments without another set of large equipment for predrilling. Some of the urban projects require pile driving in low head room or with very small clearance from existing structures. In other cases, the pile driving may have to be conducted without an access road to the piling location. This paper presents as to how the Press-in Piling Technology can effectively mitigate the negative environmental impacts associated with pile driving on urban infrastructure projects along with case studies in Japan, U.S.A. and Mexico.

1 INTRODUCTION The Press-in Piling Method was invented out of pure necessity back in 1975 in Japan, where very many infrastructure projects were simultaneously built nationwide due to the country’s rapid economic expansion as well as the government’s policy at that time. A sheet pile driving project in a regional city of Kochi, some 500 kilometers west of Tokyo, was forced to shut down due to a noise and vibration complaint filed by a local resident who lived right next to the project. He had to rest during the daytime due to his night time work. This incident prompted Akio Kitamura, who was the president of the local foundation contractor involved, to start thinking about an alternative pile driving method which would not generate noise or vibration. By collaboration with a local inventor who was dubbed “Edison of Kochi”, he built the first Press-in piling machine. Although the original purpose for creating the pile driver was to utilize it for his company’s sheet pile driving projects, foundation contractors in other regions of the country, who were also looking for a low noise and low vibration pile driving method, started to ask

about the equipment. That was the beginning of the success story of the Press-in Piling Method. By now, it has been widely used not only in Japan but also in many parts of Asia, Europe and North America, providing environmentally-friendly solutions to numerous foundation projects. 2 HOW DOES THE PRESS-IN PILING WORK? The Press-in Piling Method typically utilizes reaction force derived from a few previously installed piles to hydraulically push the next pile into ground (see Figure 1). The Press-in piling equipment of this type grips the top of already-driven piles to drive the new one and moves forward or backward on its own (See Figure 2). Due to the fact that this method is not using vibrating or purcussive force to drive the piles, it is regarded as a very environmentally-friedly piling method.


12

Low noise and low vibration press-in piling method in soft soil in congested urban areas

Figure 1. Principle of the Press-in Piling Method.

 Low noise and practically vibration free.  The equipment size is relatively small and its clamping points on the pile are much lower than those of other piling methods. This enables the equipment to work in physically tight working conditions, horizontally and vertically.  With attachments it can effectively drive piles into hard soil.  It can achieve much more accurate pile installation thanks to a combination of the better control of the pile and the lower clamping points compared to other piling methods.

Its advantages are;

Figure 2. Sequence of the Press-in Pile Driving.

3 PRESS-IN PILE DRIVING IN HARD SOIL

3.1 Water Jetting Dense sand, stiff clay, gravel, cobbles and boulders are difficult to drive piles into. In some cases, limestone, mudstone and weathered rock layers may exist in pile lines. The high pressure water jet and crush auger attachments as part of the Press-in piling technology are very useful tools to drive piles

into some of these hard soil. The high pressure water jetting will be quite effective in dense sand and silt layers. A small nozzle attached to the toe of each pile blasts out a small diameter of high pressure water to create a pilot hole in a hard soil, loosen it and lubricate the pile surface, reducing pile’s skin friction. See Figure 3.


TAKUMA T.

13

Figure 3. Press-in Piling with Integrated High Pressure Water Jetting for Dense Sand and Silt.

3.2 Crush Auger Attachment The integrated auger, which simultaneously drills into hard soil as the pile is pressed in, allows the pile to be advanced by loosening, crushing and partially

removing the hard soil to accomplish the smooth pile driving. Figure 4 shows how the auger system works.

Figure 4. Press-in Piling with Integrated Crush Auger Attachment for Hard Soil.


14


4 NON-STAGING PRESS-IN PILING FOR PROJECTS WITH LIMITED ACCESS The Press-in pile driving can be done by having all the necessary equipment moving on top of alreadydriven piles. In other words, a line of driven piles can be used as construction road by the Non-staging Press-in Piling Method. Pile Runner transports the piles from a faraway access point to the pile driver.

Clamp Crane picks up the pile from Pile Runner and pitches it to the pile driver. See Figure 5. The method is highly useful for projects that have limited access, such as those above shallow or deep water, ones on a sloped embankment without the need for staging or dense jungle without any construction road.

Figure 5. Non-staging Press-in Piling Method with Clamp Crane and Pile Runner.

5 CASE STUDIES 5.1 Myoshoji River Restoration Project (Gekitoku-1 Section), Tokyo, Japan Myoshoji River is one of Kanda River’s tributaries located approximtely 15 km northwest of downtown Tokyo. Although only 9.7 km long and its watershed being relatively small, the river runs through densely populated residential and commercial areas of the city. Very heavy rainfall (263 mm) on one September evening in 2005 flooded more than 3,300 units of buildings in the area. To reduce such flooding in the future, this project was to widen the river to increase the drainage capacity and to reduce the flood risk by installing 634 of 1,000 mm diameter tubular piles into the existing concrete retainig walls. The pile depth varied from 11 to 22 m. Rotary Press-in Piling equipment was utilized to effectively drive the piles into the concrete retaining walls without removing them. Each pile had cutter bits welded at the toe of the pile to facilitate the cutting operations. Due to the roads on both sides of the river being quite narrow

for site access and also to the fact that they had to be kept open for the local traffic almost all the time, the Non-staging Press-in Piling Method was adopted. The tube piles were delivered to the project’s material handling point by a flat bed truck and tranferred to Pile Runner that subsequently traveled on the rail placed on the pile top to Clamp Crane’s pick-up point without blocking the road traffic. Figure 6 shows the project’s sectional view and Figure 7 shows site’s soil conditions containing dense sand, sandy gravel and consolidated silt layers with SPT values at or higher than 50 at 8 m and continuously beyond 12 m below ground. The Rotary Press-in Piling equipment used a small quantity of water as “lubricant” for efficient rotary cutting operations. Figure 8 shows the Rotary Pile Driver at work just in front of a local clinic.


TAKUMA T.

15

Figure 6. Sectional View of the Project. Figure 8. Pile Driving in Densely-populated Area.

5.2 Sandalwood Canal Improvement Project (Hodges Bl. from Beach Bl. to Atlantic Bl., Project No. P-80-01) in Jacksonville, Florida, U.S.A.

Figure 7. Soil Conditions.

This project was to repair the damaged earthen levees by an earlier flooding as well as to increase the drainage capacity of an existing canal by widening/deepening with sheet piles driven into the levees running through a densely populated area of the City of Jacksonville, Florida. In order to minimize noise/vibration and also to reduce in-stream exposure of the equipment during construction, two units of the Press-in pile drivers were used and the work was done during the dry season of winter. The widths of levee shoulders were relatively narrow (approximately 3 m, see Figure 9) for a truck crane to maneuver through, so a 10 ton capacity Clamp Crane was used for hoisting sheet piles to the pile drivers. The soil conditions were primarily sandy with the SPT values of between 10 and 45 as shown in Figure 10. The noise and vibration during the sheet pile driving were limited by the specifications in the following manner. “The hydraulic press-in equipment shall not produce more than 70dB of noise, at a distance of 25 feet from the equipment, while in operation. It shall not produce any measurable vibration at the ground surface, at a distance of 25 feet from the equipment, while in operation.” Approximately 950 pairs of 7.0 m to 9.0 m long Zshaped PZC18 type sheet piles were driven without causing damages to the nearby homes. Figures 11 and 12 show the jobsite before and during the sheet pile driving. The channel was sandwiched by rows of houses. SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

16


Figure 9. Typical Cross Section.

Figure 10. Soil Conditions.

Figure 11. Sandalwood Canal After Vegetation Removal.


TAKUMA T.

17

Figure 12. Two Units of Press-in Pile Drivers With Clamp Crane.

5.3 Foundation Reinforcement Work of San Juan de Ulua Fortress, Veracruz, Mexico

deal with the harder layer in order to drive 18 meter long U-shaped LX32 sheet piles.

The 16th century fortress’ foundation needed to be rehabilitated and also protected to allow nearby canal’s widening construction. To protect and reinforce the foundation, sheet piles were driven to form permanent retaining walls outside the perimeter of the fortress standing in sea water. Vibration from pile driving had to be minimized by any means not to damage this invaluable historical landmark. The agency in charge decided to adopt the Press-in Piling Method to achieve this goal. Sheet pile alignments were on the west, south and southeast sides of the fortress as shown on Figure 13.

San Juan de Ulua Fortress

Figure 14. Boring Data. Sheet Pile Line

Figure 13. Plan View of the Sheet Pile Wall.

Although the water depth was relatively shallow (3 to 4 meters) and the soil is generally soft, there were some harder layers due to dense sand mixed with shells as seen on Figure 14. The high pressure water jet attachment was employed to effectively

Sheet piles were driven right against the fortress’ foundation on east and west sides. There was one Y-shaped connecting point with four sharp angle corners of sheet pile walls. A barge-mounted crane was used to pitch the sheet piles to the Press-in pile driver working on top of the sheets as shown on Figures 15 and 16.


18


6 CONCLUSION Conventional pile driving methods not only cause hated noise and vibration but also are not useable on many projects in congested and densely populated cities due to various local conditions. On the other hand, the Press-in Piling Method can achieve the project’s goal in harmony with such urban environment. With attachments and auxiliary systems, it has wide range of applications in highly congested conditions. With ever growing population, the author believes that major cities in the world including Mexico City and other populous cities in the country will be greatly benefitted by adoption of this method. It has shown the ability to preserve the nation’s historical landmark in Veracruz. Figure 15. Section View of Sheet Pile Driving Work.

REFERECES White, D., Finlay, T., Bolton, M., and Bearss, G. (2002), “Press-in Piling: Ground Vibration and Noise During Piling Installation”, Proceedings of the International Deep Foundation Congress, ASCE Special Publication 116 Motoyama M., Goh, T. and Yamamoto, M. (2005) “Silent Piling Technology and Its Application in Hong Kong”, Proceedings of the 2005 conference of the Hong Kong Branch of the Chartered Institution of Highways and Transportation Figure 16. Profile of Sheet Pile Driving Work.

The sheet piles were driven accurately without causing damage to the foundation or the fortress’ structure. The operator was operating the pile driver on its own staging affixed to the machine. See Figure 17.

Figure 17. Sheet Pile Driving at Southwest Corner of the Fortress.


Technical Committee




The use of displacement piling technology in soft soil conditions El uso de tecnología de pilotes de desplazamiento en condiciones de suelo blando Antonio MARINUCCI, PhD MBA PE1, & Marco CHIARABELLI1 1Soilmec

North America, Houston, Texas, USA

ABSTRACT: Displacement piles are cast-in-pace, reinforced concrete piles that are formed with little or no soil removal, where the soil is displaced radially into the soil by the drilling tool. This piling technique is applicable in soft-to-firm ground conditions – in loose to medium dense sands and in cohesive soils where the undrained shear strength is less than about 100 kPa (2,000 psf). There are many benefits to the use of displacement piles, including low vibrations during pile construction, minimal amount of soil removal, no need for stabilizing fluids (slurry), and improvement of the load resistance especially in side friction. The benefits of displacement piling make this technology ideal in contaminated and/or urban environments. This paper provides an overview of the various types of displacement piles that have been used, applicability of the technology, and general requirements for types of equipment and tooling needed. In addition, practical examples of the technology and recent advancements to displacement piling tooling will be presented as mini case histories. .

1 OVERVIEW Drilled displacement piles (DDP) refers to a specialized technology in which a bored pile is constructed using a process in which (1) a specially designed tool is advanced into the ground using both rotation and downward thrust (“crowd force”) to displace the in situ soil radially outward into the surrounding formation, and (2) concrete is injected and steel reinforcement (if required) is inserted to fill the created hole and provide structural stiffness. A key benefit of DDP is the minimal amount of drill spoils generated, which provides a cost effective and practical solution for sites with contaminated soils (e.g., typically found at landfills, brownfield sites, and industrial facilities). In addition to the reduced environmental impact, other advantages of DDP such as proven reliability, relatively rapid construction, high daily production, minimal noise associated with DDP, and minimal ground vibrations have contributed to the increased use of the technique especially for construction in urban areas, in congested spaces, and in close proximity to existing structures. DDP has been used as structural foundation elements (e.g., support column loading) and for ground improvement (e.g., column-supported embankments) on both commercial and public work type projects. The maximum diameter and depth that can be achieved are directly related to the capability of the drill rig used to construct the DDP. As reported in the literature, displacement piles with diameters ranging from about 300 to 800 mm (12- to 32-inches) and to a maximum depth of approximately 35 m (115 ft).

1.1 Description and Classification The myriad types of bored piles are typically classified according to the qualitative amount of disturbance resulting from the piles’ construction, which can range from non-displacement type to a complete or full displacement type of pile. Drilled shafts fall under the non-displacement type of piling, and continuous flight auger (CFA) piles can be categorized under either non-displacement or partial displacement type depending on whether (a) the concrete/grout is injected under pressure, and (b) the ratio of the outer diameter of the hollow drill stem to the diameter of the borehole is greater than about 50-60%, in which case a greater amount of soil will be displaced radially and compacted into the borehole wall. That is, a narrower hollow stem will result in minimal-to-partial displacement of the soil, while a wider hollow stem will result in a greater amount of soil being displaced into the surrounding soil during drilling. DDP can be listed under either partial displacement or full displacement type according to the installation method and/or the type/shape of tooling used to create the pile, which can be grouped as essentially cylindrical shaped (Figure 1a) or screw shaped (Figure 1b).


20

The use of displacement piling technology in soft soil conditions

1.2 Benefits provided by Displacement Piles Various practitioners (Basu et al, 2010; Paniagua, 2006; Baxter et al, 2006; Bottiau, 2006; Brown, 2012; NeSmith, 2004; Pagliacci and Chiarabelli, 2015; etc.) have extolled the benefits realized through the use of displacement piles, which include the following:

(a)

(b)

Figure 1. Schematic of (a) cylindrical shaped and (b) screw shaped displacement piles.

A schematic of a representative displacement tool is shown in Figure 2, which highlights many of the common components found on modern DDP tools. In general, modern displacement tools will contain the following common elements (Figure 2): (1) displacement body, which is an enlarged section near the bottom of the drill string that facilitates soil movement radially outward, thereby displacing it into the surrounding soil, (2) a drilling tip attached to the bottom of the drill string that is used to loosen the soil during the advancement of the tool (if a re-usable drilling tip is used, a pivoting gate located near the bottom of the tool or drill string is utilized for the injection of the concrete or grout; otherwise, with a sacrificial drilling tip, the concrete or grout is injected through the bottom of the drill string), (3) a hollow stem drill string with a diameter smaller than or equal to the diameter of the displacement body, (4) a lower auger segment with partial flights that moves the soil upward toward the displacement body, and (5) an upper auger segment with partial flights that moves the soil downward toward the displacement body.

Figure 2. Schematic of a representative DDP tool delineating many of the common components found on most modern DDP tools (DeWaal displacement pile tool shown; modified from Basu et al, 2010).

• Environmentally friendly because minimal amount of drill spoils produced return to ground surface, thereby lowering both the risks associated with transport of spoils (especially contaminated material) and the cost of disposal; • Minimal vibration induced during the construction of the displacement pile because the rotary drilling technique does not induce large vibrations into the soil; • Even in loose soils, the borehole can be formed without need of steel casing and/or slurry; • Cleanness of the working platform, lowering the risk to injury of onsite personnel; • Compared with non-displacement bored pile techniques, the concrete overbreak is significantly lower; and • Compared to non-displacement bored pile techniques, higher unit side friction and end bearing resistance can be achieved through the compaction of the surrounding soil, which results in a lower cost (per ton of load). 2 INSTALLATION-INDUCED CHANGES During the construction, the soil surrounding the DDP will undergo changes to its stress state (e.g., change in void ratio) as a function of the soil type, original stress state and consistency, shape of the tool, and installation method. The changes are directly caused by the loading imposed on the soil by the tool from both radial compactive/torsional stresses and vertical shearing stresses during the advancement and extraction of the tool. Ultimately, the ground improvement induced by the installation process results in larger unit values of side shear. Consequently, the load-displacement response of a displacement pile is comparatively stiffer than that of a similarly sized non-displacement pile; therefore, compared with similarly sized non-displacement piles, DDPs will be able to achieve a given load resistance at a shorter length. In loose to medium-dense cohesionless soils, the compactive effort of the tool produces greater radial displacement and results in a decrease of the void ratio (higher relative density than initial) relatively soon after construction is completed. Brown (2005) reported that DDPs “increase the horizontal stress in the ground and densify sandy soils around the pile during installation…[achieving] a measure of ground improvement around the pile.” For soft to stiff (displaceable) cohesive soils, the soil will be deformed


MARINUCCI A. et al.

plastically, may require some time to realize the increase in shear strength as the soil undergoes consolidation within the affected zone. In sensitive soils, the disturbance caused by the displacement tool during advancement and extraction could result in remolding and formation of residual shear planes. In partially saturated and fully saturated soils, the advancement and extraction of the displacement tool may generate excess pore water pressures in the soil surrounding the pile. For cohesionless soils with minor fines content (<15%), the dissipation of these induced excess pore water pressures will be relatively rapid. For cohesionless soils with appreciable fines (>15%) and for cohesive soils, however, the dissipation of the induced excess pore water pressures will require time to dissipate, which will depend on the length of the drainage path. Brown (2012) warned that the construction of DDPs “may induce excess pore water pressures in the surrounding soil, which could result in water intrusion into a newly constructed pile as the excess pore pressure dissipates.” 3 APPLICABILITY The technique is ideally suited for a wide spectrum of soil conditions ranging from sandy gravel to clay, with the caveat that the soil is able to be both displaced and compacted. In cohesionless soils, displacement piles are appropriate in loose to medium-dense soil conditions, where the relative density (Dr) is less than about 65%, the cone tip resistance (qc) is less than about 14 MPa (2,000 psi), and/or SPT N-values are less than 30 blows/0.3 m (or 30 blows/ft). Due to the compactive nature of the displacement tool, the void space is decreased, the structure is reorganized, and the relative density increased, which has a positive effect on the behavior of the DDP. NeSmith (2002) reported that the installation of displacement piles in dense cohesionless soils (qc greater than 14 MPa or 2,000 psi) can be difficult and uneconomical. In cohesive soils, displacement piles are appropriate in soft to stiff soil consistency, where the undrained shear strength does not exceed about 100 kPa (2,000 psf) or where SPT N-values are less than about 10 blows/0.3 m (or 10 blows/ft). During installation, the cohesive soils undergo plastic deformation and are compacted. Stiff cohesive soils, however, are difficult to compact. Brown (2012) indicated that “residual soil profiles, weak limerock formations, cemented sands, and stiff clays are soil types that favor easy construction” of displacement piles. Reporting on conditions in the United Kingdom, Baxter et al (2006) described that applicable conditions for the use of DDP include sites with “alluvium, soft clays, loose sand, or chalk.” DDPs should be considered as an alternative to conventional CFA piles in instances where a weak

21

layer is underlain by a stronger, more competent layer(s) at moderate depths, and where potential soil mining and effects from ground vibrations would be a concern (Brown, 2005). According to Bustamante and Gianeselli (1998), the performance of DDP may be compromised due to possible difficulties encountered during installation in very loose sandy soils or very soft clayey soils (characterized by SPT N-values<5 blows/0.3 m (5 blows/ft) or qc<1 MPa (145 psi)). According to Brown (2005), the use of DDP may impractical or problematic in predominantly saturated, fine-grained and plastic soils where the advancement and extraction process will cause extensive remolding of the soils, which could be deleterious to the soil structure, shear strength, and performance of the pile. In some instances, displacement piles may be used even in the presence of incompressible and/or nondisplaceable layers (e.g., dense sands and gravel, overconsolidated cohesive layers, weathered and soft rock) provided that (1) the thickness of the incompressible layer is less than about 1 to 1.5 m (3 to 5 ft), and (2) the incompressible or nondisplaceable or layer is located at the lowermost strata that is to be treated (Soilmec, 2012). 4 GENERALIZED CONSTRUCTION PROCEDURE 4.1 Advancement Phase – Drilling During the drilling phase, the tool and drill string are rotated clockwise and penetrate the ground using the single rotary drive and crowd force provided by the drill rig, causing the material to move upward as the tool moves downward. A drilling tip (appropriate for the anticipated ground conditions) is attached to the bottom of the drill string is used to loosen the soil during the advancement of the tool and to prevent soil from entering and plugging up hollow stem. As the drill string advances deeper into the soil, the lower auger flights move the soil upward toward the displacement body, which then displaces and compacts the soil radially outward into the borehole wall and surrounding formation. The drilling phase continues until the desired depth is achieved. The maximum achievable depth is limited by the capabilities of the drill rig: (a) the pull up/extraction force, (b) the maximum available rotary torque, and (c) the height of the drill mast, which can be extended using a Kelly extension and/or by the addition of jointed sections to the drill string. 4.2 Extraction Phase – Concreting Once the desired depth has been achieved, the displacement tool and drill string are extracted while concrete or grout is simultaneously pumped through the hollow stem and injected into the void created by the tool. The concrete or grout is injected either


22


through a grout port (with disposable plug) or a pivoting gate located near the bottom of the tool or drill string (used with a re-usable drilling tip) or through the bottom of the drill string (used with a sacrificial drilling tip). The pressure used to pump the concrete or grout should be correlated to the ground conditions at the pile location, and should be established using a preproduction test program. NeSmith (2002) indicated that for loose to medium-dense cohesionless soils, the initial “target lift-off” grout pressure should be within the range of about 140 to 210 kPa (20 to 30 psi) and between 70 to 105 kPa (10 and 15 psi) for the remainder of the grouting. As the bottom of the drill string approaches the ground surface, the pressure should be gradually decreased according to the in situ stress state. When the tool is within about 1.5 to 3 m (5 to 10 ft) of the surface, the pressure should be decreased to zero and the pumping stopped. As described in Section 5, rotation of the displacement tool may or may not occur, depending on the tool and technique utilized. The concrete mix or the grout mix contains many similar components: Portland cement, aggregate (fine aggregate for grout), water, fly ash, and admixtures. The admixtures affect and control the rate of hydration (for workability and set time), and the water reducers (e.g., plasticizers) affect the amount of water needed for fluidity and flowability to ensure the fresh concrete can get to its intended location without clogging the lines. Brown (2005) reported that the concrete or grout mix should provide “that all solids remain in suspension without excessive bleed-water, must be capable of being pumped without difficulty, penetrate and fill open voids in the adjacent soil, and allow for insertion of the steel reinforcement.” Therefore, a high slump, fluid concrete or grout mix (with aggregates of fine gravel with a máximum particle size of 18 mm (¾-inches)) should be used. Through the action of the displacement tool, the borehole wall is compacted and is relatively smooth, which reduces the concrete overbreak and eliminates risk of over-augering. 4.3 Insertion of Steel Reinforcement Depending upon the technique used to construct the DDP, the steel reinforcement (cages, bars, beams, etc.), when required, can be placed either prior to or after the extraction of the tool and concreting. For displacement tools that have a large internal passage (Figure 3), it is possible to insert the steel reinforcement inside the hollow stem prior to the placement and injection of concrete. In this instance, the tip at the end of the tool is sacrificial, and serves as the injection location for the concrete. The tool is then extracted during the concreting process. For most techniques, however, the steel reinforcement is inserted after the tool has been extracted, the concreting completed, and while the concrete is still fresh. Depending upon depth, reinforcement configuration, and fluidity of the

concrete, the reinforcement may need to be pushed or vibrated into place.

Figure 3. Photograph of sacrificial tip, lower section of a cylindrical displacement tool, and the internal passage within the hollow stem of the tool.

5 TYPES OF DISPLACEMENT PILES & TOOLS As Brown (2012) explained, “the torque and crowd required to construct a drilled displacement pile is substantial…the energy required to install the pile is related to the resistance of the soil to the displacement, and so the piles are often installed to a depth that is controlled by the capabilities of the drilling rig.” For the installation of conventional DDPs, modern hydraulic drilling/piling rigs are capable of producing high torque (≥500 kN-m or ≥370,000 ft-lb) and large crowd forces (450 kN or 100,000 lb), which are needed for the desired pile diameters and depths. Paniagua (2006) provided a detailed history of the evolution of DDPs and the principal advancements realized during each of these three generations. Basu et al (2010) present a comprehensive narrative and thorough review for many of the different tools and techniques used to construct displacement piles in Europe and North America. It is important to note that the advancements made in DDP technology are the direct result of contractors and equipment/tooling manufacturers developing practical solutions to actual problems and issues. The early methods (prior to the 1970s) used to construct displacement piling (“first generation” piles) focused on either soil removal during the advancement of the tool or on the insertion of large casing into the ground during advancement. Moreover, relatively low torque (50-100 kN-m or 37,000 to 74,000 ft-lb) was required by the drilling equipment to perform these piles, but the production was slow. The methods comprising these first generation piles include: Atlas piles, DeWaal piles, Franki VB piles, Fundex piles, and Olivier piles. The next version of displacement piles emerged during the 1970s and improved upon the production rate achievable during advancement by adding partial auger flighting near the bottom of the tooling. Methods comprising this second generation of


MARINUCCI A. et al.

displacement piling include: Pressodrill pile, Tubex pile, SVB pile, and SVV pile. Since the 1980s and even moreso during the last two decades, there has been advancements in the tooling (e.g., increased diameters, and design of the flights and body to increase production), techniques (e.g., reduced vibrations, spoils, and noise), and drilling equipment (e.g., greater torque and pulldown crowd force, which permit larger diameters and greater depths). Methods comprising this third generation of DDP include: Omega pile, Berkel Auger Pressure Grouted Displacement (APGD) pile, Menard controlled modulus columns, Trevi Discrepiles, and displacement piles constructed using the Soilmec Traction Compaction Tool (TCT). To highlight the main differences among select displacement piling methods, the following sections provide a succinct overview of select techniques. A schematic of four displacement piling methods (DeWaal pile Fundex pile, Omega pile, and Berkel APGD pile), photographs of different Soilmec Discrepile tools, and a schematic of the Soilmec Traction Compaction Tool are provided in Figures 4, 5, and 8, whereby both the displacement mechanism can be discerned and the differences in the shape of the completed DDP can imagined. 5.1 DeWaal Pile During the drilling phase, the De Waal displacement tool (Figure 4b) is advanced downward into the formation using clockwise rotation and crowd force. Once the desired depth is reached, concrete is placed into the hollow stem of the drill string to a prescribed distance above the ground surface (i.e., head), and then the sacrificial tip at the bottom of the

23

typically inserted into the borehole after the concrete has been placed but while it is still fresh. In some instances, the steel reinforcement can be inserted into the hollow stem prior to the placement of concrete. Typical diameters achievable with this technique range from about 300 to 600 mm (12- to 24-inches), and to a maximum depth of about 25 m (82 ft). 5.2 Fundex Pile For Fundex displacement piles, the tool with a conical auger tip (Figure 4c) is advanced downward into the formation with clockwise rotation and crowd force, thereby displacing the soil radially outward. Once the desired depth is reached, steel reinforcement is inserted into the hollow stem, the sacrificial drilling tip, which forms the enlarged pile bottom, is released, and concrete is placed into the hollow stem. The tool/drill string are then extracted using an oscillating up-anddown motion along with a 180° clockwisecounterclockwise rotation, while ensuring the concrete is maintained at desired level within casing. The withdrawal process produces a nearly smooth shaft. The possible diameters and lengths for Fundex piles range from about 450 to 675 mm (17.5- to 26.5inches) and to a maximum depth of about 35 m (115 ft), respectively (Basu et al, 2010). 5.3 Omega Pile As shown in Figure 4d, the diameter of the flange along the length of the Omega tool and partial auger flights increases gradually and similarly from both ends of the tool toward the displacement body (with maximum diameter). For displacement piles constructed using this method, the tool (with a varying diameter) and drill string are advanced downward into the formation using clockwise rotation and crowd force. Once the desired depth has been reached, concrete is injected under pressure and the sacrificial tip is released. During the extraction of the tool while maintaining a clockwise rotation, concrete is injected under pressure until the tool and drill string are fully extracted from the borehole. The reinforcement cage is inserted (assisted by vibratory means and/or downward force) down into the fresh concrete. For some Omega piles, it is possible to place the steel reinforcement (e.g., cage or bar) into the drilling stem prior to the placement of concrete (Bottiau, 2006). 5.4 Berkel (APGD) Pile

(a)

(b)

(c)

(d)

Figure 4. Schematic of select Displacement Pile methods: (a) DeWaal Pile, (b) Fundex Pile, (c) Omega Pile, and (d) Berkel APGD Pile (modified after Basu and Prezzi, 2005).

tool is released. The tool and drill string are extracted using clockwise rotation while maintaining a constant head of concrete within the hollow stem, resulting in a relatively smooth surface. The steel reinforcement is

For the Berkel Auger Pressure Grouted Displacement (APGD) pile, the tool (Figure 4e) is advanced downward into the formation with clockwise rotation and crowd force. Once the desired depth is reached, high-strength grout is injected under pressure through the hollow stem of the drill string. Once the initial target pressure is achieved, the extraction of the tool maintaining a clockwise rotation and grouting of the


24


borehole is initiated. Pressurized grouting is continued along clockwise rotation of the tool until the tool and drill string are fully extracted from the borehole. After the tool is removed from the borehole and while the grout is still fresh, the steel reinforcement is inserted in the grouted hole. Essentially, two types of APGD piles can be constructed: (1) partial displacement piles in loose to dense sands (with N<50) where the diameter ranges from 300 to 500 mm (12- to 20-inches) and to about 17 m (80 ft) in length, and (2) full displacement piles in loose-to-medium dense sands (corresponding to SPT blow count N<25) where the diameter can range from 300 to 450 mm (12- to 18-inches) and to about 24 m (79 ft) in length (NeSmith, 2002).

the downward spiraling auger flights. Eccentric-type roller tools are able to form boreholes with diameters ranging from about 450 to 600 mm (17.5- to 26.5inches).

5.5 Trevi Discrepile The cylindrical (“Cilindrico”) displacement tool shown in Figure 5a is well suited for soft ground conditions: loose to medium-dense sands, soft clays, and organic soils. The conical (“Conico”) displacement tool is shown in Figure 5b, and is well suited for stronger ground conditions: medium-dense to dense sands and stiff clays. The conical tool is modular, and is composed of four primary sections: (1) a drilling tip fitted with teeth appropriate for the soil conditions being drilled, (2) a lower section with right-handed partial auger flights that move the soil upward toward the displacement body, (3) a central cylindrical body that stabilizes and displaces the soil radially outward thereby producing the actual pile diameter, and (4) an upper section with left-hand partial auger flights that move the soil above the tool downward toward the displacement body. Different pitch lengths can be used on the auger flighting depending upon the anticipated soil conditions: (1) a short pitch is preferable for very loose and fine sands, soft clays, and organic soils, (2) a medium pitch is suitable for medium-dense sands and medium clays, and (3) a long pitch is preferable for medium-dense to dense cohesionless soils, medium to stiff clay, and sandy gravel. Cylindrical and conical displacement tools are able to form boreholes with diameters ranging from about 350 to 600 mm (13.5- to 26.5-inches). The eccentric roller (“Pirucca”) displacement tool shown in Figure 5c is ideally suited for soft to medium ground conditions: loose to medium-dense sands, soft clays, and organic soils. Given the eccentric nature of the tool, only a portion of the roller is in contact with the soil at a time, while the remaining main portion of the tool/drill string is away from the wall, thereby decreasing the frictional resistance acting on the tool/drill string. As a result, less torque is needed to rotate and advance the tooling/drill string, which allows the use of smaller/lighter drill rigs (lower operational and transport costs). Soil that is not compacted into the borehole by the eccentric roller is moved downward toward the displacement body by

(a)

(b)

(c) Figure 5. Different types of Soilmec Discrepile tooling: (a) cylindrical tool, (b) conical tool, and (c) eccentric roller tool. The requirements for proper selection of a drill rig capable of constructing displacement piles using the Discrepile tools include (a) a rotary head capable of delivering rotation at about 20 to 25 rpm, (b) a rotary head capable of delivering at least 200 to 250 kN-m (147,000 to 185,000 ft-lbs) of torque, (c) a pull down system with a crowd force of at least 200 kN (45,000 lb); and (d) a pull-up system capable of providing an extraction force of at least 200 kN (45,000 lbs). The concrete pump should be sized according to the expected extraction rate of the drill string, the volume of the void created by the displacement tool as it is extracted, and the required pressure range that will be used during the injection of the concrete.


MARINUCCI A. et al.

5.5.1 Case History – La Prua Business Center; Rimini, Italy To support the proposed mixed-use residential and business structures for the new La Prua Business Center located along the waterfront in Rimini (Figures 6 and 7), the design engineer estimated that more than 20,000 linear meters (65,000 linear feet) of bored and/or driven piling would be required to support the building structures. At this site, the general subsurface profile consisted of a layer of silt with sand and clay to a depth of about 3.1 m (10 ft) underlain by a 7.8 m (25.5 ft) thick layer of medium-coarse sand with silt and rounded pebbles, which was underlain by a 14.05 m (46 ft) thick layer of clayey silt with traces of organics. Beneath the clayey silt layer, there is a 2.1 m (6.9 ft) thick layer of medium sand with silty gravel, underlain by a 1.1 m (3.6 ft) layer of gravel with sand and silt, which is then underlain by a 6.8 m (22.3 ft) thick layer of clayey silt and clay with interbedded lenses of sand.

25

The original foundation options (Figure 8) included bored piles stabilized with bentonite slurry during drilling and driven piling. However, it was deemed that the driven piling operations would have caused excessive environmental concerns (e.g., noise and vibrations) to the nearby residents and surrounding buildings, respectively. In addition, there was concern that the bored pile operations, especially the use of bentonite slurry, would have issues related to the cleanliness of the jobsite and effect on the surrounding roads resulting from the trucks transporting the excavated and removed the spoils.

Figure 8. A portion of the plan view of the foundation structure layout beneath the curved structure and the middle structure.

Figure 6. Aerial photograph of the location (outlined in red) for the proposed mixed use, residential-commercial structure.

As an alternative to conventional piling, the geotechnical specialty contractor, Trevi S.p.A., proposed using unreinforced displacement piles installed in a diamond shaped pattern for ground improvement beneath the structures (Figure 9). There was concern expressed by the owner / engineer that the unreinforced displacement pile elements would not provide structural support should it be required or needed. To ensure adequate support for the structures and allay any concern by the owner, the contractor performed a compression test on a sacrificial, non-production displacement pile (Figure 10). As shown in Figure 11, at a load of about 176 tons, the top and creep displacements were approximately 8.5mm (5/16 inch) and 3mm (1/8 inch), respectively.

Figure 7. Axonometric illustration of the proposed mixed use, residential-commercial structure.


26


(a) (a)

(b) Figure 9. (a) Trucated plan view showing the pattern of the installed displacement piles, and (b) photograph of two exposed tops of completed displacement piles.

(b) Figure 11. Graphical depiction of the load-displacement behavior from (a) a compression load test on an unreinforced displacement pile, and (b) a creep load test performed with a constant load maintained for 60 minutes.

In total, about 1,600 piles with a diameter of about 600 mm (24-inches) and an approximate length of 25 m (82 ft) were installed using a conical displacement tool and a Soilmec SR-65 drill rig. Due to the compactive nature of the tool and the resulting face of the borehole wall, the concrete overbreak was kept to a minimum, as anticipated, and averaged between 5% and 10% above theoretical. In addition, the drill spoils that needed to be disposed were also kept to a minimum, where the disposal volume averaged between 5% and 10% of the total volume of installed piling. 5.5.2 Case History – Monselice Hospital; Monselice, Italy Figure 10. A photograph of the compression test load frame. Note: a steel sleeve was used to laterally restrain/support the exposed portion of the displacement pile.

Located southeast of the Euganean Hills and southwest of Padua in Italy, the Monselice Hospital is located in a town with a population of 18,000 inhabitants and in an area with substantial geothermic activity, leading to the popularity of local hot springs


MARINUCCI A. et al.

and mud spas. By working with nature, the building’s designers incorporated the use of the underground geothermal resource to heat and cool the hospital. The foundation for the hospital was constructed using displacement piling technology to maximize the benefit of minimal removal of the drilling spoils, noise, and vibration. The geology below the hospital was very amenable to displacement piling construction because it was characterized by fine to mediumdensity sand (SPT N-values≤45) and silty clay with the groundwater table located about 0.9 m (3 feet) below the ground surface. The displacement piles were installed by Trevi S.p.A. using a Soilmec SR-80 drill rig fitted with the displacement pile kit and the conical displacement tool (Figure 12). The in situ soil was displaced and compacted radially during downward penetration during the drilling phase.

Figure 12. Installation of displacement piles using a SR-80 drill rig and a conical displacement tool.

During the extraction phase, concrete was injected at the tip of the tool as the drill string was extracted. When the tool was out of the hole, the reinforcement cages equipped with geothermal loops were then installed in the fresh concrete. The geothermal loops were incorporated into the displacement piles to take advantage of the stable subterranean temperatures where to provide the hospital with heat in the winter and air conditioning in the summer. The piles were about 600 mm (24-inches) in diameter and ranged in length from about 17 to 24 m (55 to 80 ft), resulting in a total of about 44,000 linear meters (144,300 linear feet) of displacement piling. The average daily production of displacement piling was about 16 piles per day or about 350-380 m per day (1,150 to 1,250 ft/day). To provide control and monitoring during the drilling and concreting phases, the contractor utilized the monitoring system (Soilmec Drilling Mate System, “DMS”) that was integrated into the SR-80 drill rig.

27

The monitoring system allowed the drill rig operators and jobsite personnel to monitor and accurately control the machine (e.g., drilling parameters and rig performance) in real time. Data from the DMS was also streamed to a remote computer where jobsite managers were able to monitor, process, and document the project information. 6 SOILMEC TCT – A NEW DDP TOOL For conventional displacement piling, as described above, where the soil is compacted during the drilling or penetration phase, a large, heavy, powerful drill rig is required to provide the crowd force and torque needed to achieve the desired displacement, diameter and depth. Therefore, an effective manner to reduce the need for very large drill rigs for this work is to reduce the required crowd force and torque needed to push the drill string and tooling into the ground. Soilmec S.p.A. patented a displacement piling technique using the Traction Compaction Tool (TCT) where the soil is compacted during the extraction phase of the pile construction instead of during the drilling phase, thereby reducing the amount of torque and crowd force required to turn the tooling and penetrate the ground. As shown in Figure 13, the TCT is composed of three main parts: (1) the lower section contains a drilling tip fitted with appropriate teeth (based on soil type) to facilitate penetration into the ground, a flight auger, a short drill string, and a concrete pivot gate that is connected to the hollow stem of the drill string; (2) the middle section contains flights and borehole stabilizer that is allowed to partially freely rotate around the hollow stem of the drill string, and are used to displace and compact the soil during extraction and concreting; and (3) the upper section is fitted with flights (only a few) rigidly connected to the drill string, which facilitate movement of the soil toward the displacement body during extraction. The upper and lower sections of the TCT are fixed and turn simultaneously. The special shape of the lower tip creates a separation with respect to the flight of the central tool’s portion, and the mechanical gate separates the tip section (where the concrete flows out from the hollow stem) from the main section of the tool. By maximizing the drill equipment operability with the TCT, comparatively smaller rigs can be used to achieve similar sized (diameter and depth) displacement piles constructed using conventional displacement piling tools/drill rigs, resulting in reduced operating and transport costs without sacrificing quality and productivity. In addition to the benefits achievable with conventional DP methods, the advantages of constructing DPs using the TCT include the construction of larger diameter elements (up to 800 mm (32-inches)), the use of smaller/lighter


28

(a)


(b)

Figure 13. Execution phases for Soilmec Traction Compaction Tool (TCT) – (a) drilling phase – minor and (b) extraction phase - soil is displaced and compacted radially into the surrounding soil.

drilling rigs, and higher quality of finished element because the mechanical gate separates the soil being displaced from the concrete being injected. During the development and trial phases of development, it was determined that displacement piles with diameters ranging from about 400 to 800 mm (16- to 32-inches) could be constructed using the TCT. Moreover, similar to conventional displacement piling and depending on the capability of the drill rig, the maximum achievable depth using the TCT is approximately 32.5 m (106 ft). (Note: at this time, the TCT has not been utilized in large-scale, production work; consequently, comparative case histories, load tests, and other performance data are not yet publicly available.) Similar to conventional displacement pile construction, there are two distinct phases utilized during the construction of TCT displacement piling. First, during the drilling phase, the TCT (in open position, Figure 14a) is advanced into the ground and the disturbed in situ soil fills the flight of the tool but there is little-to-no compaction occurring during this phase. Second, during the concreting phase, the TCT (closed position, Figure 14b) is extracted and the soil along the length of the pile is displaced radially into the borehole wall. During the drilling phase, the tool and drill string are rotated in a clockwise rotation and penetrate the ground using the single rotary drive and crowd force provided by the drill rig, causing the material to move upward as the tool moves downward. Compared to the length of a typical continuous flight auger, the length of the TCT is short and the flights are few, which reduces the total friction that develops between the soil and the tool (on the flights and hollow stem). As such, the torque and crowd force needed by the drilling rig to turn and advance the tool are also reduced.

Once the desired depth has been achieved, the rotary drive turns the drill string in a counterclockwise direction; however, since the middle portion of the tool (i.e., “middle flights and stabilizer” in Figure 13a) is in intimate contact with the soil, the friction developed between the stabilizer and the ground does not allow this portion of the tool to rotate. In addition, as the drill string is rotated counter-clockwise, the lowermost flight (located below the stabilizer during the drilling phase) rotates into a position that essentially forms a cover plate above the concrete gate, effectively separating the soil to be displaced with the area to be concreted during extraction. During extraction, the soil above the cover plate along the tool is rotated downward toward the stabilizer, which is then compacted radially outwards and into the borehole wall while concrete is pumped through the hollow drill string and out the pivot gate. Through the action of the displacement tool, the borehole wall is compacted and is relatively smooth, which reduces the concrete consumption (i.e., overbreak) and eliminates risk of over-augering. If required, steel reinforcement is placed after the tool is extracted from the hole and while the concrete is still fresh.

(a)

(b) Figure 14. Soilmec R-625/SR-65 rig fit with the TCT advanced DP tooling: (a) tool in open position for the drilling phase, and (b) in the closed position that is used for the extraction/concreting phase


MARINUCCI A. et al.

7 QUALITY CONTROL During construction, it is essential to control and monitor the various parameters that affect the integrity and performance of a displacement pile. During the drilling phase, the important parameters include the drilling depth, penetration speed, rotation speed of the tool, inclination of the tool guide mast, rotary torque, and crowd force. During the extraction and concreting phase, the important parameters include the depth, lifting speed, rotation speed of the tool, inclination of the tool guide mast, rotary torque, extraction force, concrete pressure and flow, total volume of pumped concrete, and concrete overbreak. During extraction, care must be exercised to coordinate the concrete pumping rate with the extraction rate of the tooling/drill string. Necking and other integrity issues may occur if the tooling/drill string is extracted too quickly comparatively, which would be realized in a pressure decrease on the gauge at the concrete pump. The different parameters can be continuously monitored and recorded using automated monitoring systems that are integrated directly into the drilling rig, which are common on modern displacement drill rigs. Controlling and monitoring the various parameters during drilling and concreting assists with ensuring the quality of the finished product consistently meets project specifications. 8 CONCLUSION The use of drilled displacement piles has increased significantly during the past two decades as a result of various factors including advancements in tooling (e.g., increased diameters, increase production rates) and equipment capabilities (e.g., greater torque and pulldown force permitting larger diameters and greater depths). The various benefits resulting from the use of environmentally friendly displacement piles were presented, and include minimal drill spoils, reduced ground vibrations, and an increase in unit side friction and end bearing resistance. As long as the soil can be displaced and compacted, the technique is ideally suited for a wide range of ground conditions ranging from soft-to-firm ground conditions and from sandy gravel to clay. As such, displacement piling has been used for a wide array of applications on both public- and commercial-type projects ranging from unreinforced columnar elements for ground improvement purposes (e.g., column-supported embankments) and as structural foundation elements (e.g., support for column loading). For conventional displacement piles, the maximum diameter and depth that can be achieved range from about 300 to 800 mm (12- to 32-inches) and from about 24 to 35 m (80 to 115 ft), respectively, which are directly correlated with the capabilities of the drilling rig. This paper provided an overview of the evolution of displacement piles, various types of techniques and tools, applicable

29

ground conditions where the technology is suitable, and general requirements for the construction of displacement piles. Recent advancements to displacement piling (e.g., Traction Compaction Tool) were presented, where piles can be constructed to a maximum diameter and depth of 800 mm (32-inch) and 32.5 m (106 ft), respectively, which also facilitates the use of comparatively smaller/lighter drill rigs, and lower operating and transport costs. REFERENCES Basu, P., Prezzi, M., and Basu, D. (2010). “Drilled Displacement Piles – Current Practice and Design.” DFI Journal, v.4 n. 1, August. P. 3-20. Baxter, D.J., Dixon, N., Fleming, P.R., and Hadley, S.P. (2006). “The Design and Formation of Bored Displacement Piles – A United Kingdom Perspective.” Proceedings of the DFI/EFFC 10th International Conference on Piling and Deep Foundations, Amsterdam, Netherlands. Bottiau, M. (2006). “Recent evolutions in deep foundation technologies.” Proceedings of the DFI/EFFC 10th International Conference on Piling and Deep Foundations, Amsterdam, Netherlands. Brown, D.A. (2005). “Practical considerations in the selection and use of continuous flight auger and drilled displacement piles.” Advances in Designing and Testing Deep Foundations. Geotechnical Special Publication No. 132, ASCE, pp. 1-11. Brown, D.A. (2012). “Recent Advances in the Selection and Use of Drilled Foundations.” Proceedings of the GeoCongress 2012: State of the Art and Practice in Geotechnical Engineering. Geotechnical Special Publication No. 225. Sponsored by Geo-Institute of the ASCE. Hryciw, R.D., Athanasopoulos-Zekkos, A., and Yesiller, N., editors. Brown, D.A. (2005). “Selection and Design of Continuous Flight Auger and Drilled Displacement Piles.” Proceedings of the Geo-Frontiers Congress 2005. Geotechnical Special Publication No. 132 of the ASCE. Rathje, E.M., editor. pp. 1-11. Bustmante, M. and Gianeselli, L. (1998). “Installation parameters and capacity of screwed piles.” Deep Foundations on Bored and Auger Piles, BAP III. Balkema, Rotterdam. pp. 95-108. Chiarabelli, M. (2015). “Soilmec Introduces New, Advanced Technologies for Displacement Piling.” ADSC-IAFD Foundation Drilling Magazine. February/March, p. 34-37. NeSmith, W.M. (2002). “Design and installation of pressure-grouted displacement piles.” Proceedings of the Ninth International Conference on Piling and Deep Foundations, Nice, France, pp 561-567. NeSmith, W.M. (2004). “Application of Augered, Cast-in-Place Displacement (ACIPD) Piles in New York City.” Proceedings of the Deep Foundation


30


Institute (DFI) Augered Cast-in-Place Piles Committee Specialty Seminar, McGraw-Hill Building, New York, NY, pp 77–83. NeSmith, W.M. and Fox, J. (2009). “Practical Considerations for Design and Installation of Drilled Displacement Piles.” Proceedings of the Contemporary Topics in Deep Foundations: Selected Papers from the 2009 International Foundation Congress and Equipment Exposition. Geotechnical Special Publication No. 185 of the ASCE. Iskander, M., Laefer, D.F., and Hussein, M.H., editors. pp. 438 – 446. Pagliacci, F. (ed). (2015). Ground Engineering Technologies. The Trevi Group. pp. 328. Pagliacci, F. and Chiarabelli, M. (2015). “Monselice Hospital, Displacement Piling Case History.” Pile Buck Magazine, v.31, n. 2 P. 14-17. Paniagua, W.I. (2006). “Construction of Drilled Displacement and Auger Cast in Place Piles.” Proceedings of the International Symposium: Rigid Inclusions in Difficult Soft Soil Conditions, TC36, México DF Prezzi, M. and Basu, P. (2005). Overview of construction and design of auger cast-in-place and drilled Displacement piles. Proceedings of the DFI 30th Annual Conference on Deep Foundations, Chicago, U.S.A., pp. 497-512. Siegel, T.C., NeSmith, W.M., NeSmith, W.M., and Cargill, P.E. (2007). Ground improvement resulting from installation of drilled displacement piles. Proceedings of the DFI 32nd Annual Conference on Deep Foundations, Colorado Springs, U.S.A., pp. 129-138. Soilmec S.p.A. (2013). DMS – Drilling Mate System. pp. 28. Soilmec S.p.A. (2012). Displacement Piles Technology. pp. 16.


Technical Committee




Rescate de una cimentación de pilas con inclusiones rígidas Pile foundation retrofit with rigid inclusions José SEGOVIA1, Walter PANIAGUA2 & Germán LÓPEZ3 1Pilotec

S.A. de C.V., México S.A. de C.V., México 3Facultad de Ingeniería, UNAM, México 2Pilotec

RESUMEN: La construcción de una torre de consultorios médicos en el sur de la zona de transición de la ciudad de México, entre el “Pedregal de San Ángel” y el “Cerro de la Estrella”, motivó al diseñador geotécnico a resolver la cimentación con pilas coladas “in situ”. Sin embargo, debido a que la construcción de pilas la realizó una empresa con escasa experiencia en este tipo de trabajo, la supervisión de la obra solicitó la evalución de las pilas mediante pruebas de integridad. El resultado de esta revisión llevó a la conclusión de rediseñar la cimentación mediante una losa, utilizando las pilas solamente como inclusiones rígidas para reducir el asentamiento del edificio.

1.2 Ubicación geotécnica y solución de cimentación

1 ANTECEDENTES 1.1 Características del Proyecto El proyecto consiste en la construcción de la cimentación de dos torres de 6 niveles y azotea, las cuales se construirán en etapas. En la Figura 1 se muestra una planta con la localización general.

Figura 1. Localización de Hospital

El sitio se ubica dentro de la zona de Transición que se caracteriza por un costra superficial, subyaciendo por arcillas limosas de alta plasticidad y depósitos fluvio – lacustres y/o tobas areno – limosas de alta resistencia al corte y baja deformabilidad, Figura 2.

Figura 2. Ubicación del proyecto en la zona de Transición


32

Pile foundation retrofit with rigid inclusions

En la Fig. 3 se muestra una planta general del proyecto, con la ubicación de los sondeos de exploración realizados: dos sondeos mixtos, dos sondeo de muestreo selectivo y un tubo de observación del NAF.

De acuerdo con proyecto, la solución de la cimentación para el edificio fue con base en pilas coladas en el lugar, con una profundidad de 19.0m respecto al nivel de banqueta, apoyadas en los estratos resistentes indicados por el estudio de mecánica de suelos como fluvio – lacustres con diámetros de 0.8, 1.0 y 1.20 m. Las pilas de cimentación se complementaron con una losa tipo “waffle” con trabes de 0.80 m de peralte y dados de 1.6 m de profundidad para unir las pilas, Figura 5. Debido a diferentes circunstancias, la supervisión de la obra solicitó la revisión de la integridad estructural de las pilas y de su longitud de desplante, para lo que se solicitó la realización de una serie de pruebas de integridad (PIT) para determinar la sanidad de la pilas.

Figura 3. Planta general del proyecto

Los suelos encontrados de 0 a 5.4m, comprenden limos arcillosos y arenosos, de consistencia media a alta. De 5.4m a 18m, son suelos arcillosos y limosos de alta plasticidad, baja resistencia y alta compresibilidad, de consistencia blanda a media. Finalmente, de 18m y hasta 30m (máxima profundidad explorada) hay depósitos fluviolacustres, formados por arenas finas y limos arenosos, muy compactos y de consistencia muy rígida, Figura 4. El NAF se detectó a 2.15m de profundidad.

Figura 5. Planta de pilas de cimentación del Hospital

2 PRUEBAS DE INTEGRIDAD 2.1 Generalidades La prueba de integridad es un ensayo para de determinar la variación de las características del concreto de las pilas de cimentación en toda su longitud. La forma usual del ensayo consiste en la colocación de un acelerómetro en el cabezal de la pila bajo prueba, y en la aplicación de golpes con un martillo instrumentado Figura 6. 2.2 Descripción de la prueba

Figura 4. Caracterización geotécnica y diagrama de esfuerzos

El acelerómetro se fija a la cabeza de la pila por medio de cera de petróleo. Los golpes del martillo generan una onda de compresión, que recorre la pila y sufre reflexiones al encontrar cualquier variación en las características del material. Esas reflexiones causan variaciones en la aceleración medida por el sensor. El equipo hace un registro de la evolución de esa aceleración con el tiempo.


SEGOVIA J. et al.

Figura 6. Equipo de medición para pruebas de integridad

Es usual la aplicación de varios golpes secuenciales a fin de que el equipo de la prueba de integridad obtenga un promedio de las señales correspondientes. Ello permite eliminar interferencias aleatorias o efectos anómalos, sobresaliendo en la señal las variaciones causadas por las reflexiones de la onda. Como la onda hace su recorrido con una velocidad fija (velocidad del concreto), al conocerse esa velocidad de propagación y el tiempo transcurrido entre la aplicación del golpe y la llegada de la reflexión correspondiente a las anomalías o a la punta, es posible determinar la localización exacta de éstas. Las vibraciones superficiales son grabadas en la parte superior del cabezal de la pila así como todas las reflexiones primarias. Considerando la naturaleza y los tiempos de observación de las reflexiones, es posible valorar la integridad de la pila y detectar anomalías. El reflector más profundo es la punta de la pila (parte baja), por lo que su reflexión es la última que puede observarse, Figura 7.

33

y su longitud, L, la reflexión de la pila es esperada al tiempo 2L/C. Las reflexiones se observan mediante un acelerómetro móvil, temporalmente adosado en el cabezal de la pila. La aceleración de la señal es digitalizada y almacenada en el equipo colector. Asimismo, la aceleración se integra numéricamente para producir un velocigrama. Por otra parte, midiendo la aceleración del martillo y al multiplicarla por su masa, es posible conocer el valor de la fuerza aplicada, F. Durante el impacto, cuando el martillo y la pila están en contacto, la fuerza es proporcional a la velocidad esperada del cabezal de la pila. La constante de proporcionalidad es la impedancia acústica Z en el cabezal. El cociente F/Z es llamado velocidad y es presentado conjuntamente con la velocidad observada V. Ambas deben de tender a ser iguales durante el impacto. La fuerza grabada después del impacto no tiene aún un significado práctico desde el punto de vista de la integridad de la pila. Para facilitar la localización de los reflectores, la señal se presenta en función de la distancia mediada desde el punto de impacto. Las reflexiones son generadas por las variaciones de la impedancia de la pila, Z, que puede calcularse con, 𝑍= 𝜌𝑐𝐶𝐴

(1)

donde A = área transversal de la pila; densidad del material que conforma la pila; C = producto que indica cambio de impedancia y que genera reflexiones. Con la geometría de la pila (área transversal y longitud), es posible detectar irregularidades en la pila, tanto la variación de la sección transversal a lo largo del fuste como los cambios en el tipo de material, mediante la interpretación del comportamiento anómalo observado entre la señal incidente y el reflector de la punta, Figura 7. 2.4 Interpretación de resultados Se estudiaron el 100% de las pilas. En las Figuras 8 y 9, se muestran algunas de las pilas ensayadas. La interpretación de los registros de campo se realizó con la ayuda de los programas de cómputo PIT-W-2003 y Profile-2003, donde es posible realizar el post-procesado de los registros tanto en el dominio del tiempo como en el de la frecuencia.

Figura 7. Esquema de la reflexión de ondas en caso de disminución de sección de una pila

2.3 Interpretación de la prueba Dada una estimación de la velocidad de onda que viaja por el material del que está compuesta la pila, C,


34


Figura 8. Pilas ensayadas

Figura 9. Pilas ensayadas

2.5 Resultados La interpretación de los registros de campo se realizó con la ayuda de los programas de cómputo PIT-W2003 y Profile-2003, donde es posible realizar el postprocesado de los registros tanto en el dominio del tiempo como en el de la frecuencia. Es importante mencionar que la prueba de integridad da solamente una calificación cualitativa de la pila, por lo que no debe utilizarse como un mecanismo de clasificación de la pila y de su proceso constructivo. Es importante indicar que las ampliaciones y rugosidades en las pilas no deben considerarse como un defecto, sino en la mayoría de los casos como un incremento en su capacidad de

carga tanto por fricción como por punta, con la condición de que exista continuidad en el elemento. La calificación de la calidad de la pila se agrupa en familias de calidad: A. Pila Buena: No se aprecian defectos obvios y la respuesta de la punta de la pila es clara para longitudes de pila de hasta 30 diámetros. B. Pila Mala: Existe una identificación clara de defectos en la pila, no se aprecia claramente el reflector de la punta de la pila después de eliminar el ruido de la señal, aun cuando se cumple con el criterio de que la longitud de la pila es menor que 30 veces su diámetro. En tales casos, para poder descartar una pila es recomendable llevar a cabo pruebas de detalle (pruebas de carga y/o sondeos de inspección), analizar profundamente el historial de construcción de la pila, así como realizar correcciones en el caso de que los defectos se muestren superficialmente y volver a efectuar los ensayes de integridad. C. Pila con posibles defectos. Los defectos en la pila no son claros. Es necesario llevar a cabo pruebas de integridad adicionales después de aplicar medidas correctivas en caso de que los defectos se localicen en la parte superior (gran longitud del armado que sobresale en el cabezal, imperfecciones del cabezal, deficiencia en el pulido de la superficie del cabezal en la zona donde se coloca el acelerómetro, etc.); en caso que los posibles defectos persistan, será necesario llevar a cabo pruebas de carga o sondeos directos (extracción de núcleos) para poder descartar dicha pila. D. Datos no concluyentes. No se tienen registros de calidad debido a imperfecciones en el cabezal de la pila (armado que sobresale, superficies mal pulidas, contaminaciones en el cabezal), a la alta resistencia del terreno localizado a lo largo de la pila o debido a la longitud de la pila por lo que el pulso reflector de la punta no sea observable (un criterio empírico para definir la longitud máxima de la pila para que pueda observarse dicho pulso reflector es que la longitud de la pila sea menor que 30 veces su diámetro). Velocigramas e interpretación de resultados Los velocigramas capturados en campo, fueron interpretados en gabinete mostrando los siguientes resultados. Se probaron la totalidad de las pilas de la estructura, 49 pilas; 22 pilas califican sin defectos, de acuerdo a las pruebas; 8 pilas muestran reducción en su diámetro o menor longitud que la de proyecto y 19 pilas se manifiestan con posibles defectos estructurales, Figura 10.


SEGOVIA J. et al.

35

DN P31-6

6: # 7 -0.15

0.00

0.15

0.30

cm/s 0

0

Wavelet Impedance: Mass:

2.92 m

2

4

5

6

8

10

10

12

15

14 15.04 m (4200 m/s)

Relative Vol.: Construct. Vol.: Construct. Area: Max Prof ile: Min Profile:

0.74 1.00 1.00 1.39 at 2.04 m 0.40 at 13.98 m

x 39 16

20

18

20

25

22 Magn

m

V 0.210 cm/s (0.350) F/Z 0.013 cm/s (0.016)

diam

Figura 12. PILA MALA. Se observan decrementos del diámetro de la pila a partir de los 8m de profundidad, así como un decremento de la longitud de la pila de 15.15 a 14m DN P30-6

6: # 7 -0.25

0.00

0.25

0.50

cm/s 0

0

2

4

Figura 10. Resultado de la interpretación de las pilas

Low Pass: Wavelet Impedance: Mass:

0.26 m 1.78 m

5

6

8

10

10

12

15

14 15.04 m (4200 m/s)

En las Figuras 11, 12 13 y 14, se presentan ejemplos de los velocigramas de campo y su interpretación de gabinete.


0.91 1.00 1.00 1.37 at 1.91 m 0.74 at 13.45 m

x 20 16

20

18

20

25

22 Magn

0.00

diam

Figura 13. PILA CON POSIBLES DEFECTOS, Se observan decrementos del diámetro a partir de los 12m de profundidad hasta la punta de la pila, así mismo se observa un decremento de la longitud de la pila de 15.45 a 15m. Se aprecia claramente la señal de la pila

DN P28-5

5: # 4 -0.40

m

V 0.173 cm/s (0.210) F/Z 0.012 cm/s (0.013)

0.40

0.80

cm/s 0

0

Low Pass: Wavelet Impedance: Mass:

0.36 m 0.42 m

5833 Hz 5000 Hz 1.143E+004 kN/m/s 0.9 kg

4

5 DN P34-6

4: # 8-175% -2.75

8

0.00

2.75

5.50

cm/s

0

12

16.11 m (4200 m/s) x 20

10


1.09 1.00 1.00 1.66 at 1.73 m 1.00 at 0.00 m

0

2

Low Pass: Wavelet Pivot Impedance: Mass:

1.76 m 1.56 m

1.143E+

4

16 6

5

15 8

20 10

12

Magn

V 0.365 cm/s (0.388) F/Z 0.009 cm/s (0.009)

24 m

20 diam

10

14 15.15 m (4200 m/s) x 20

16

Figura 11. PILA BUENA. Se observa un incremento importante de la longitud de la pila de 15.15 a 16.1m

18

15

20

22 Magn

V 0.929 cm/s (1.486) F/Z 0.005 cm/s (0.006)

m

diam

Figura 14. DATOS NO CONCLUYENTES. No se observa una señal clara



0.70 1.00 1.00 1.38 at 1.79 m 0.20 at 11.21 m

36


3 ALTERNATIVA DE SOLUCIÓN

Carga de la estructura

En las condiciones actuales las pilas de cimentación no se consideran aptas para trasmitir las cargas de la estructura a los depósitos resistentes del subsuelo de la zona; sin embargo, esta situación puede ser solventada utilizando las pilas como inclusiones rígidas que restrinjan la deformabilidad de los depósitos lacustres que se localizan entre 5.4 y 18.0 m de profundidad.

Loas de cimentación y plataforma de transferencia

3.1 Antecedentes de las inclusiones rígidas Las inclusiones rígidas han sido empleadas desde hace muchos años como un sistema para reducir asentamientos diferenciales e incrementar la capacidad de carga de estratos blandos; se han utilizado principalmente como solución de grandes plataformas de almacenamiento en Europa y Oriente, pero grandes ejemplos utilizando estacones de madera como inclusiones rígidas lo tienen muchos edificios del centro histórico de la Cd. de México. En los años recientes grandes desarrollos habitacionales han sido construidos utilizando este principio, principalmente en la zona de Aragón y en ciudades como Morelia, Michoacán y como ejemplo notorio de su utilización se tiene el rescate de la Catedral Metropolitana de la Cd. de México.

Inclusiones rígidas

Figura 15. Características de una cimentación con inclusiones

Figura 16 Esquema de solución

3.2 Características de la cimentación Una de las claras ventajas que se tienen al utilizar las pilas como inclusiones rígidas, es que se aprovechan las pilas que se encuentran construidas, modificando su concepto de trasmisoras de carga a reductoras de deformación. Bajo el concepto anterior se requiere que la losa de cimentación tenga la suficiente rigidez para trasmitir de una manera uniforme las cargas al subsuelo, además de dotar a la cimentación de una plataforma de transferencia entre la losa y las pilas de cimentación, Figura 15. La plataforma de transferencia está compuesta por una capa de tepetate compactado, suelo – cemento o relleno fluido de baja resistencia. Otro punto importante de la solución es que se mantienen los niveles arquitectónicos sin modificar ninguna parte del proyecto. 3.3 Diseño conceptual En relación con la magnitud de cargas que se trasmitirán a la solución de inclusiones se tiene que: La estructura pesa alrededor de 5,400 t Área de cimentación 995 m2 Descarga neta 5.4 t/m2

4 PROCEDIMIENTO CONSTRUCTIVO La implementación de la solución requierió de las siguientes acciones: Demoler las cabezas de las pilas hasta el nivel 3.95 m respecto al 0.0 de proyecto. Colocar como capa de transferencia un limo arenoso (tepetate) compactado al 95% de la prueba Próctor estándar, colocado en capas máximas de 20 cm y con un adecuado control del contenido óptimo de agua de compactación; la capa se colocó desde el nivel de demolición (-3.95 m) hasta el nivel -3.25 m. Construcción de las contratrabes de cimentación, se estimó un peralte de 80cm, cuyo valor final se derivará de los análisis estructurales complementarios. Colocación entre contratrabes de una capa de tepetate compactado al 90% Próctor. Construcción de la losa de cimentación de acuerdo con los resultados de los análisis estructurales. Los peraltes de las contratrabes y de losa de cimentación dependerán de los resultados finales de los análisis estructurales.


SEGOVIA J. et al.

5 CONCLUSIONES Ante una mala práctica de construcción de pilas, el proyecto de la torre de consultorios se vio comprometido. La solución de desligar las pilas de la losa de cimentación y analizarla como una cimentación con base en una losa de concreto, reforzando el suelo con inclusiones, rehabilitó la viabilidad del proyecto y permitió concluir su construcción, Figura 16.

Figura 16 Estado de la estructura, octubre 2015

REFERENCIAS SMMS (2002), “Manual de Construcción Geotécnica, Cap. 9, Inclusiones”, Sociedad Mexicana de Mecánica de Suelos. Santoyo, E. y Ovando, E. (2001), “Catedral y Sagrario de la Ciudad de México, Corrección Geométrica y Endurecimiento del Subsuelo”, editado por TGC Geotecnia S.A. de C.V. Paniagua, W.I., Ibarra, E. and Valle, J.A. (2008), “Rigid Inclusions for Soil Improvement in a 76 Building Complex”, 33rd Annual Member´s Conference, New York, Deep Foundation Institute.


37

Technical Committee




Foundation design and construction for high-rise Towers in Mexico City Diseño de la cimentación y construcción de Torres de gran altura en la Ciudad de México Peter W. DEMING1, Sissy NIKOLAOU2, Raymond J. POLETTO3, & George J. TAMARO3 1Mueser

Rutledge Consulting Engineers, Partner, New York, NY USA Rutledge Consulting Engineers, Senior Associate, New York, NY USA 3Mueser Rutledge Consulting Engineers, Consultant, New York, NY USA

2Mueser

ABSTRACT: The unique Mexico City subsurface conditions with deep rock and natural valley topography filled with soft plastic clays, combined with the high seismic activity, makes design and construction of foundations for high-rise buildings challenging. The authors present some of the challenges stemming from more than two decades of relevant experience that include: (i) need for comprehensive site characterization including in-situ dynamic measurements; (ii) seismic hazard, ground motions that reflect the amplification of the seismic waves propagating from rock through deep soft clay; (iii) high risk that temporary excavations will cause ground movement and damage adjacent structures (iv) settlement due to the clay behavior and with respect to regional settlement; (v) soil-foundation-structure interaction; and (vi) dense urban construction that could result in building-to-building interaction. Recent projects using a performancebased design approach are discussed, raising some questions and presenting some solutions. The paper provides an overview of foundation systems used for few high-rise buildings of 40 or more stories along Paseo de la Reforma.

.

1 INTRODUCTION Foundation design for high-rise towers requires an understanding of the building loads, geotechnical conditions and regional seismicity. Structural details such as basement depth, perimeter wall alignment, column spacing, and column load variations are required for a full understanding of the soilfoundation-structure system. Design aspects of foundation type and depth of adjacent structures often drive decisions for temporary excavation support methods where deep basements are desired. Temporary excavation support requirements are considered in the foundation design, and may be incorporated into the permanent foundation system. Structural requirements create design and construction challenges in the geology of the Mexico City Federal District for design of high-rise towers. Recent designs call for deeper basements and construction directly adjacent to existing structures. Since 1947, MRCE has been involved with Mexico City projects, evaluating sites and designing foundations. The firm has provided comprehensive geotechnical and foundation design and testing services since the 1990s in downtown and surrounding suburban areas for major developers (See Figure 1). Example projects the authors have worked on include geotechnical assessment of sites in Federal District (e.g., Alameda Park site, Torre Mayor at

Chapultepec Park’s east end, and Santa Fe’s multiple developments,) and peer review services on several other tall buildings in the Zona Rosa area. The 55-story Torre Mayor (aka 505 Torre Reforma) received the prestigious American Consulting Engineers Council (ACEC) Platinum Award in 2004.

Figure 1. High-rise structures designed with MRCE support.


40

Foundation design and construction for high-rise Towers in Mexico City

2 GEOLOGIC SETTING AND SITE CONDITIONS

2.1 Tectonic and Seismologic Setting Mexico City is subject to earthquakes and other natural hazards such as floods and volcanic eruptions. The region is located on a unique volcanic high plateau at about 2240 m above sea level, bounded by volcanic sierras, alluvial fans and plains (Flores-Estrella et al., 2007).

earthquakes of magnitudes between 7 and 8, and many frequent smaller events. The ocean floor of the Cocos plate is subducting beneath the continental edge of the North American plate at a rate of about 6 cm/year. Figure 3 is a 3-D image of the subducting slab developed by seismic tomography (Caltech, 2015). The slab begins forcing its way down beneath the continental crust at a shallow angle, then levels off to nearly horizontal. Below Mexico City it plunges steeply into the mantle and ends abruptly at a depth of about 500 km.

Figure 3. Subduction image (modified from Caltech, 2015).

2.2 Geologic Setting – the Ancient Lake

Figure 2. Seismotectonic setting and major tectonic plates.

The deep basement (Valley of Mexico) is faulted and folded, contributing to basin seismic effects. The plateau is located within the Trans-Mexican Volcanic Belt (TMVB), a complex Tertiary and Quaternary feature which crosses the country from the Pacific to the Atlantic Oceans. Situated on a subduction zone, the city’s complex seismotectonic environment consists of four major tectonic plates shown on Fig. 2 (North American, Cocos, Caribbean, and Pacific) and a microplate (Rivera) that have generated

The site was originally a lake with islands connected by causeways to surrounding higher ground. Spanish colonists filled the lake between the islands in the 17th and 18th centuries. The Federal District Paseo de La Reforma area where high-rise structures under discussion are located, lies within the “Lower Transition Zone” and at the east edge of Largo Centro I (central lake zone) on the area map of Fig. 4. This is the so-called Valley of Mexico formed by volcanic materials interspersed with alluvial deposits covered in the center of the valley by lacustrine clays. Ground water pumped for domestic supply has resulted in ground settlement, causing widespread damage of structures and infrastructure.


DEMING P. et al.

Figure 4. Mexico City geotechnical zoning (after TGC).


41

42


2.3 Subsurface of Reforma (Zona Rosa) As in most areas formerly occupied by the lakes, the primary soil strata are soft clay deposits combined with layers of stiff clayey silt and silty sands. A typical subsurface cross section in the Zona Rosa area is shown in Fig. 5 (Ovando-Shelley, et al., 2007). Geotechnical performance is dominated by the compressible “Upper Clay Series” clay and silt deposits. The Upper Clay deposit (FAS) underlies the man-made fill and is a crust of desiccated low plasticity silty clay that contains volcanic ash, with water content as high as 300% and Liquid Limit up to 400%.

Figure 5. Typical cross section in central Mexico City (modified from Ovando-Shelley, et al., 2007). Designers should develop a soil profile specific to their site.

The Upper Clay series is interlayered with compact silty sands. These compressible deposits are underlain by the “Capa Dura” a compact alluvial sand layer at a depth of about 25-30 m, and the lightly preconsolidated lacustrine “Lower Clay Series.” The Federal District along Paseo De La Reforma are located lies within the “Lower Transition Zone” and the east edge of “Lago Cetro I.” Performance of soils in Lago Centro is dominated by the compressible “Upper Clay Series.” At depths on the order of 33 m the soil profile transitions to “Depositos Profundos,” thick layers of medium compact silty sand. The depth to bedrock within Lago Centro is so great that structural support on bedrock is impractical. 2.4 Water Conditions The shallow phreatic water table is on the order of 4 m deep; this is likely water perched in the Upper Clay. Water extraction from deep sources has lowered piezometric water levels in the deep deposits. Prudent designs should assume the shallow water table will be restored in the life of the structure, and consider such hydrostatic forces will act on the basement walls and base slab.

3

FOUNDATION DESIGN

3.1 Geotechnical Considerations The Upper and Lower Clay series are not suitable for high-rise building support and introduce substantial risk in temporary excavation support, especially adjacent to existing structures. Shallow mats in the clay horizons also incorporate drilled shafts or piles bearing on the “Depositos Profundos” silty sand below 33 m. There are two basic foundation systems, either structural mat or mat on piles, or deep elements alone as foundations independent of the base slab. In either case, the base slab should be designed as a pressure slab transferring hydrostatic uplift to the columns, or with a permanent drainage system to relieve hydrostatic pressure. The pressure slab is preferred, as drainage systems require maintenance, can increase effective stresses, and result in base settlement and downdrag forces at the perimeter. Design of the base mat as a structural mat for load support takes advantage of the weight credit provided by excavation and reduces dependence on the load-deformation performance of individual foundation elements. For example, the 55-story tower at 505 Reforma has four basement levels at 18 m deep, with a 2.5-m thick piled mat bearing on the “Capa Dura” sands, and drilled shafts bearing in the “Depositos Profundos” below 33 m depth. The drilled shafts were arranged at column locations and distributed around the perimeter. Drilled shafts take advantage of side friction in both clay and sand layers, and load sharing between the mat and piers provides bearing and overturning safety, and reduces settlement. Newer structures that the authors are familiar with involve greater basement depths for parking, some requiring the removal of all of the Upper Clay Series. The high-rise structures are supported on deep foundations bearing in silty sand below 33 m depth, without base mat contribution. For deep basement designs, use of slurry construction methods such as perimeter diaphragm walls (slurry walls) and cast-in-place concrete drilled shafts or Load Bearing Elements (LBE) at column locations is desirable. These deep foundation elements are constructed from the ground surface to provide hydraulic head advantage to the slurry methods, and because soil subgrade at lower levels cannot support the construction equipment. Use of diaphragm walls at the perimeter and topdown excavation support methods are popular, as they alleviate the cost of temporary excavation support bracing, provide rigid support of neighboring properties and can be designed to prevent bottom heave in soft clays. Slurry-supported construction advanced from street grade increases contact bearing stresses at the sidewalls of the deep


DEMING P. et al.

elements under fluid pressure of fresh concrete for maximum friction benefit. There remain many subsurface unknowns that must be explored, considered, and controlled by the design engineers to be confident that either the low structure or high-rise tower is safe for occupancy for 50 years or more. These are explained below. 3.2 Structural Loads Structural gravity and wind loads above grade can be far easier calculated than the vertical and horizontal reactions on the foundation elements in the soft clay soils at basement level. Knowledge of soil stratification, geotechnical strength and consolidation characteristics, and piezometric pressure at depth are required for reliable estimates of foundation support capacity, downdrag forces, and structure total/differential settlement. Load contributions are not computed in one phase, but rather evaluated several times during the design development, as the interplay between final project structure conditions and foundation performance are considered by the designers. Of particular importance is the influence of deep excavation on adjacent structures. Many low-rise structures supported on shallow footings or mat foundations suffered severe damage and even collapsed in past earthquakes. Following the violent 1985 Mexico earthquake of moment magnitude Mw 8.0 and a Mercalli Intensity of IX, local engineers have an understanding of the importance of seismic soil-structure interaction and are seeking more advanced design guidance from international specialized consultants. Using a more thorough understanding of the soil-foundation-structure system, designers can introduce more redundancy by distributing the mass and stiffness to create multiple load paths that are favorable to the safety and performance of the structure. 3.3 Friction and Bearing for Deep Foundations Even though elements constructed under slurry are typically designed for side shear support without end bearing, removal of sediment from the bottom is mandatory. Low-density sediment is readily displaced by concrete, and can coat the lower sidewalls of the element to greatly reduce side friction. Soft sediment trapped at the bottom reduces safety factor derived from end bearing, introducing an undesirable increase of risk to foundation performance. Where building columns are supported on individual LBE systems at wide spacing, the deformation of each LBE is critical for adequate frame support and bottom preparation should be carefully inspected. An LBE constructed from grade can be designed to incorporate an upper structural steel column terminating in the concrete LBE below final subgrade, or as LBE columns extending through the basement

43

space. In top-down construction, attaching floors and a base mat to a concrete LBE is feasible with special detailing. Shear dowels are drilled into or placed through the LBE elements to support floor girders and connect to a base mat (See Figure 6). The LBE is often used in conjunction with perimeter diaphragm walls since both are constructed with the same equipment, reducing construction cost. Drilled pier elements are a viable option, but because of their smaller perimeter area, several drilled piers must be clustered below a column that require a structural cap or mat foundation to distribute column loads to the deep elements.

Figure 6. Foundation mat top and bottom steel with LBE connections before concrete pour.

3.4 Deep Foundation Recommendations For high-rise tower foundation design, we provide these general recommendations:  Deep foundations (piles, slurry walls, LBEs) with and without soil grouting must be competently designed to account for friction, downdrag, end bearing and lateral resistance in the various soil deposits that they are driven or drilled through.  The depths and consistency of soil strata (especially the lake deposits) vary considerably, and may be influenced by adjacent structures due to group action and seismic lateral movements. The geotechnical exploration should define both shallow conditions which support adjacent structures and deep deposits which are planned for high-rise support.  Supporting foundation design with load testing of trial deep foundation elements at the design depth and constructed in the manner proposed for production work is highly recommended. Load testing with instrumentation to demonstrate actual side shear and end bearing values which are then applied to the final foundation geometry, provides


44


confidence in foundation performance. If an Osterberg load cell is used (See Figure 7), cleaning the bottom of the test shaft to provide competent bearing is required since the cell obtains resistance from end bearing for stressing the shaft side shear.  Use of post-grouting to improve deep foundation element performance is feasible, but the improvements should be demonstrated by testing. Tip grouting is recommended as a proactive remedy for potential soft sediment at the tip due to slurry construction. Side grouting is desirable, but may be ineffective in clay soil.

for moment and shear connections for belowgrade frame action. 4 FOUNDATION CONSTRUCTION

4.1 Support of Excavation Considerations Deep excavations may be made with open-cuts and sloped sides in soil, or they may be made with the sides supported by a sheeting system (timber, steel sheet piling or slurry walls) and bracing of various types in a staged construction of a large site.

Figure 8. Couplers at block-out in LBE in top-down construction.

Figure 7. Osterberg cells used at 505 Torre Reforma.

 Construction methods must be carefully inspected and slurry properties tightly controlled to prevent reduction of sidewall resistance and prevent sedimentation below support elements.  Diligent inspection and quality control testing are needed during the execution of foundation work to reduce owner and contractor’s risk and liability.  Floor keys should be blocked out for exposure as construction progresses downward.  Couplers are needed when multiple segments of preassembled cages or closely spaced rebar cages are used in deep piles and slurry wall or LBEs. Couplers (See Figure 8) are recommended

Open-cuts must have stable side slopes and be above the ground water. Site conditions can be determined by geotechnical borings and undisturbed soil sampling to determine design properties. Where the sides of the excavation are supported the support wall must extend deep enough below the excavation subgrade to prevent a base heave failure. A stability analysis of the perimeter system should be performed to maintain adequate safety factor. With deep basement construction, slurry walls can be extended deep below the temporary support requirements to gain vertical load capacity and enhance horizontal resistance. The walls also control temporary seepage and base heave during construction. 4.2 Construction of Deep Foundations “Top-down” construction is a somewhat complex technique which works well for high-rise towers with deep basements. The use of top-down construction of deep foundation basements offers many advantages. The upper building may advance at the same time as excavation below grade (Up-Down), so long as the tower loads are supported by the foundation method, and final foundations in staged excavations are completed in concert with the tower


DEMING P. et al.

loading (e.g., Torre Mayor). The concept, illustrated in Figure 9, allows schedule savings. Slurry walls, internally supported by permanent basement slabs, allow excavation spoils to be removed through temporary openings in the floor slabs, which are cast as the excavation progresses. Top-down construction is preferred for excavations deeper than 30 m with the following advantages:  Elimination of temporary bracing and shoring below grade if the permanent basement is cast as excavation proceeds downward (See Figure 10).  High-rise structural work can commence early.  Imperfections in deep elements exposed by excavation can be addressed before loads increase above allowable.

Figure 9. Top-down basement construction illustrated and example.

Figure 10. Staged temporary bracing of excavation at Torre Mayor.

45

5 SEISMIC DESIGN

5.1 Local Code and Performance Based Design Seismic design codes in Mexico have been progressively developing for more than 7 decades (Ordaz & Meli, 2004). Some important developments adopted internationally include importance factors, linear distribution of seismic forces with height, dynamical analysis, and higher accelerations for soft soil conditions. The Mexico City Building Code published in 2004 is used as a model code in different municipalities and states in the country (Alcocer & Castano, 2008). Some key aspects include: (i) site effects in design spectra are addressed using the predominant ground period (See Figure 11) and (ii) two limit states (service and collapse prevention) along with interstory drifts that reflect expected structural performance. For essential structures, peer review procedures are considered as risk reduction measures. The surge of high-rise construction at the turn of the last century world-wide created a need for performance-based design (PBD) approaches to guide design for heights outside the range of building code prescriptive provisions. The Pacific Earthquake Engineering Research Center (PEER) responded to this need by leading the Tall Buildings Initiative (TBI) to develop engineering design criteria that will ensure safe and usable tall buildings following future earthquakes. As a result, the “Guidelines for Performance-Based Seismic Design of Tall Buildings,” was developed (PEER, 2010). The Guidelines are an alternative to prescriptive procedures such as those in ASCE7 and IBC (International Building Code), intended for engineers and building officials involved in seismic design of individual tall buildings. The Guidelines consider seismic response characteristics of tall buildings, including relatively long fundamental vibration period, significant mass participation and lateral response in higher modes of vibration, soil-foundation-structure interaction, etc. They recommend a sliding scale of seismic hazard (or probability of an earthquake happening during the design life of the building) with corresponding performance objectives. Globally, the PEER (2010) PBD Guidelines have been extensively used for tall and mega-projects, and in recent Mexico tall buildings in combination with the local code. The PBD application in Mexico includes three seismic hazard levels and corresponding level of performance: Service Level (for a frequent event with return period Tr = 43 years), Design Level (an event that has reasonable probability to occur within 50 years of design life and return period of Tr = 125 years), and Maximum Considered Earthquake MCE (extreme very rare event with Tr = 2475 years).


46


motion and large settlement on the order of 0.6 m (end-bearing elements). The 1985 earthquake led to tighter building codes and the establishment of a Seismic Alarm System which provides a 50-sec warning for any earthquake greater than magnitude 6.0 occurring off the coast of Guerrero or Michoacán. 5.3 Seismic Hazard

Figure 11. Predominant ground period (sec) in 2004 Mexico City Code (after Ordaz & Meli, 2004).

5.2 Historic Seismicity Mexico has a long history of destructive earthquakes and volcanic eruptions. Figure 12 shows events of magnitude greater than 7.0 between 1900 and 2000.

The seismic sources affecting Mexico City are shown on Fig. 13. The wave attenuation varies for each type of source as shown on Fig. 14 for an assumed earthquake magnitude of 7.5. This figure shows attenuation of the spectral acceleration (SA) for a structure with period T of 1 sec and distance from the epicenter ranging from 0 to 500 km. Uniform hazard predictions for each PBD design level at bedrock can be made using the Probabilistic Seismic Hazard Analysis (PSHA) approach originally developed by Cornell in 1968. PSHA (See Figure 15) combines all seismic sources, incorporating their annual activity rate and the wave attenuation from the epicenter to the site using a logic tree approach where each assumption is assigned a weight to produce uniform spectra at the bedrock or some stiff interface layer.

Figure 13. Examples of shallow crustal (C) and intraslab (I) seismic sources of Mexico. Figure 12. Historic earthquakes with M>7 between 1900 and 2000 (ref: USGS). 1

0.1

SA (1.0 sec) : g

SA (1.0 sec) : g

On September 19 1985, an earthquake measuring 8.1 on the Richter scale (moment magnitude Mw of 8.0) centered between the states of Michoacán and Guerrero, in the subduction zone off Acapulco, killed more than 4000 people in Mexico City, 300 km away. Double resonance coupling between the earthquake shaking, soil layers, and buildings caused intensity IX shaking, lasting up to 3 min in some areas. Effects of soil amplification and topography were pronounced in the lake area. Surface foundation failures were observed and deep foundations experienced reduction in shear strength (friction elements) due to many cycles of strong

0.1

Atkinson & Boore (2003) - Worldwide Intraslab Zhao et al. (2006) USGS Variant Intraslab Sources Youngs (1997) USGS Variant - Intraslab Rock Zhao et al. (2006) USGS Variant Interface Sources & Boore (2003) -Worldwide Subduction Atkinson & Boore (2003) -Worldwide Subduction Atkinson Abrahamson & Silva (2008) Abrahamson & Silva (2008) Boore & Atkinson (2008) Boore & Atkinson (2008) Shallow Crustal Sources Campbell & Bozorgnia (2008) Campbell & Bozorgnia (2008) Chiou & Youngs (2008) Chiou & Youngs (2008)

Atkinson & Boore (2003) - Worldwide Intraslab Zhao et al. (2006) USGS Variant Youngs (1997) USGS Variant - Intraslab Rock 0.01 Zhao et al. (2006) USGS Variant

0.01

0.001

1

7.5 MagnitudeM= =7.5

M = 7.5 1

Intraslab Sources Interface Sources

Shallow Crustal Sources

0.001

10

1

SITE-TO-SOURCE R : km

100

10

D I S T A NSCI TE E - T O - S O U R C E R : km

500

100

500

DISTANCE

Figure 14. Wave attenuation of spectral acceleration for magnitude 7.5 and structural period of 1 second.


DEMING P. et al.

5.4 Site Response and Ground Motions Site effects are significant in the deep soil profiles of the Reforma area. One-dimensional equivalent linear analysis can be used to capture soil effects and supplemented as needed with nonlinear and twodimensional (2-D) models. Figure 16 illustrates computed soil amplification of the motion at the foundation elevation compared to the motion at the bedrock. As in this case, it is typical that the dominant natural soil periods are long, often higher than 1 sec, and the response of the soil may not be damped out depending on the thickness and properties of the fine-grained soils. Additional ground motion effects due to topography depend on the site location with respect to the valley.

Figure 15. Steps of PSHA (from Cornell 1968). 3.5 Ground Surface - 2475 yr

AF = 3.0

Amplification Factor, AF

3.0

Average AF

2.5 2.0 1.5 1.0 0.5 0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

47

There are several approaches for incorporating Soil-Foundation-Tower interaction in the PBD analysis. Listed in order of simplicity:  Fixed-base: The superstructure is assumed to be fixed at the surface and the input motion is the free-field (FF) motion at the ground surface (Fig. 17a). The results do not consider interaction and are a “baseline” for more detailed analyses.  Foundation model (service earthquake): The input motion at the bottom of a rigid bathtub (Fig 17b) can either be the FF or the FIM (Foundation Input Motion), modified to account for kinematic interaction. This approach does not account for s soil springs and the subterranean levels. Mass should include the mass of the tower below grade.  Foundation model (MCE): Springs and dashpots representing foundation-soil interaction along the foundation sides and base are included (Fig. 17c). Input motions are applied via a rigid foundation frame as either FF or FIM. Nonlinear response history analysis is performed using 3D model with the ground motion applied at the foundation base or directly through the distributed soil springs. The importance of incorporating the foundation compliance with equivalent springs depend on the foundation system selected. In Mexico City, there have been tall buildings whose foundations are mats structurally designed to transfer the load to the upper soil layers, often complemented by deeper foundations to ensure safety and account for buoyancy effects. In this case, the equivalent springs and their distribution below the mat can result in a longer structural period as compared to the fixed-base model (Fig. 17a). In smooth, code-type spectra, this translates to a lower base shear but to also larger computed deformations of the structure, since it is more flexible. When a slurry wall acts also as the perimeter foundation wall, the interaction between the wall and the surrounding soil should be properly represented by equivalent springs, especially since these springs may vary in stiffness for each soil layer. The complete soil-foundation system should be studied with the input applied at the bottom of the rigid mat (Fig. 17c).

Period, T : s

Figure 16. Amplification of seismic motion at ground surface from the soil layers above bedrock.

It is essential that the site is characterized sufficiently with dynamic measurements of soil shear wave velocity and laboratory testing. 5.5 Foundation-Tower Interaction The physical problem of Soil-Foundation-Tower Interaction is illustrated on Fig. 17 (top) and various model simulations are illustrated on Fig. 17 (bottom). SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

48


Ref: Nikolaou (2008)

Figure 17. Soil-Foundation-Tower Interaction: (top) actual problem (Nikolaou, 2008); (bottom) model simulation options (Klemencic et al., 2012).

In other designs, the load transfer relies mostly on LBEs that transfer loads to deep competent soil layers, while the slab acts essentially as slab on grade without providing stiffness or load transfer. In this case, the stiffness of the LBE controls the response and shift of the structural period. This may not be significant for very stiff rigid elements and the simplified model of Fig. 17b can be applied with some modifications on the input ground motion that should be filtered for kinematic effects (Klemencic et al., 2012). The complete soil-foundation system should be studied with the input applied at the tip and along the shaft of the deep LBEs (Fig. 17c). The responsibility of the analysis and design of the foundation systems is strongly dependent on good collaboration between the structural and geotechnical engineers who often go through several iterations to ensure compatibility of deformations and strains between the superstructure and substructure models.

design. Testing must also consider the construction methods, especially where slurry stabilization is used. The high skin friction provided by the upper clays challenges static load testing of the deep sands, but enables the Osterberg load cell method to develop sufficient reaction for measurement of the unit friction in the deep sand. The Osterberg load cell has been in very few developments in Mexico City. Future Needs: More available test data defining the unit friction capacity of the deep sands being engaged in friction for high-rise structure support. Regional seismic activity is high, and the great depth to bedrock with soft soil and clay deposits can result in an elongated period of the soil-foundation system. Soil-structure interaction analysis may not lead to a reduction in inertial forces for tall structures, as the soil may also vibrate in long periods. The PBD design approach has become a standard practice for high-rise designs in Mexico City, in combination with the local Codes. Future Needs: Better agreement on assumptions made for hazard evaluations, such as selection of seismic sources and attenuation models. Foundation-tower interaction analysis should be performed in progressive stages as design details are developed. For deep excavations which remove the upper clays, the complete soil-foundation system should be studied with the earthquake excitation applied at the bottom of the rigid mat, or at the tip and sides of the deep foundation elements where a mat is not constructed. High-rise construction loads result in high bearing stresses on mats or high shear stress at deep element interfaces. These loads are also applied to the soil adjacent to the perimeter diaphragm wall, which typically performs as a foundation in bearing. Future Needs: Static and dynamic interaction of bearing support soil should be considered where new structures are in close proximity to existing, and new foundations are constructed several stories deeper than adjacent foundations. REFERENCES

6 CONCLUSIONS AND FUTURE NEEDS The construction demands introduced by deep excavations into the soft upper clay deposits of Mexico City, and the need to support excavations adjacent to structures supported on shallow foundations, allow permanent diaphragm walls and top-down construction methods to be competitive with temporary excavation support methods. Mat foundations with shared deep element support at columns, and deep element support with only pressure slab at grade are popular foundation options due to the challenging subsurface conditions of Mexico City. The use of pile load testing to measure unit skin friction is needed for safe and practical foundation

Alcocer, S.M. & Castano, V.M. (2008). “Evolution of codes for structural design in Mexico,” UNAM Avilésa, J. & Pérez-Rochab L.E. (2010). “Regional subsidence of Mexico City and its effects on seismic response,” Soil Dynamics & Earthquake Engineering, 30(10):981-989 Caltech (2015). “The unusual case of the Mexican subduction zone,” CalTech Tectonics Obervatory. Chávez M. y Alcántara L. (1990). “Interacción sueloestructura en estaciones acelerográficas de la ciudad de México”, Memorias XV Reunión Nacional de Mecánica de Suelos, SLP, 1: 69-76 Cornell, C.A. (1968): “Engineering Seismic Risk Analysis”, BSSA, 58:1583-1606


DEMING P. et al.

Flores-Estrella, H. et al. (2007). “Seismic response of Mexico City Basin: A review of 20 years of research,” Natural Hazards, 40:357-372 Hadley, P.K. et al. (1991). “Subsoil geology and soil amplification in Mexico valley.“ Soil Dynamics & Earthquake Engineering, 10(2):101-109 Klemencic, R., McFarlane, I.S., Hawkins, N.M., Nikolaou, S. (2012). “Seismic Design of Mat Foundations,” NIST/NEHRP Design Brief No. 7. Nikolaou, S. (2008). “Site-Specific Seismic Studies for Optimal Structural Design,” Structure, Feb. Ordaz, M. & Meli, R. (2004). “Seismic design and codes in Mexico,” 13th World Conf. on Earthquake Engineering, Vancouver, Canada, Paper No. 4000 Ovando-Shelley, E., Ossa, A., Romo, M.P. (2007). “The sinking of Mexico City: Its effects on soil properties and seismic response,” Soil Dynamics & Earthquake Engineering, 27:333-343 PEER (2010). “Guidelines for Performance-Based Seismic Design of Tall Buildings,” Tall Buildings Initiative (TBI) Stone, W.C. et al. (1987). “Engineering Aspects of 9/19/1985 Mexico Earthquake,” NIST, NBS BSS16 Tamaro, G.J. et al. (2000). "Design & construction constraints imposed by unique geology in New York," DFI 8th Int. Conf., NYC TGC Geotecnia, MRCE (1994). “Geotechnical and environmental investigation for the Chapultepec Tower project, Mexico, DF” [now Torre Mayor].


49

Technical Committee




Deep foundations in Mexico City soft soils Cimentaciones profundas en suelos blandos de la Ciudad de México Gabriel AUVINET-GUICHARD1 & Juan-Félix RODRÍGUEZ-REBOLLEDO 2 1

Instituto de Ingeniería, UNAM, México 2 Universidade de Brasilia, Brasil

ABSTRACT: A general overview of the different solutions to the problem of foundation of buildings on highly compressible soft soils such as those found in Mexico City valley is presented. Special attention is given to deep foundations systems that were developed to control settlement or protruding in consolidating soils. Recent techniques for modeling the behavior of deep foundations are also reviewed.

1 INTRODUCTION The lacustrine soils of Mexico City are famous worldwide for their high water content, poor shear strength and high compressibility [35]. Since the founding of the city, Mexican engineers became aware that, in many cases, the techniques used in other countries were not directly applicable to the exceptionally difficult geotechnical conditions prevailing in the area. They had to innovate. This was not an easy task and some foundation designs proved unsatisfactory due to wrong reasoning, forgotten factors or overly optimistic assumptions. The most severe judges of the quality of the foundation systems were the earthquakes that affect periodically this area; the harshest lessons, but also the most useful, were derived from seismic events and especially those of 1985 [5]. In this paper, an overview of the contributions of many researchers, consultants and builders to foundation engineering in the soft soils of Mexico City is presented, emphasizing deep foundation systems and special solutions that have been proposed to overcome the specific difficulties of the area. Recent contributions of numerical modelling and full-scale observations to a better understanding of foundations behavior and improved design criteria, most of which have been included in Mexico City building code, are exposed. 2 GEOTECHNICAL CONDITIONS Until the end of the XVIIIth century, the valley of Mexico was a closed basin with a number of shallow lakes, amongst them: Texcoco, Xaltocan and Chalco lakes. The capital of the Aztec empire, Tenochtitlan was built on a small island of the Texcoco Lake. The

valley became an open basin when the Nochistongo cut, a channel 7km long and up to 50 deep, was dug by hand between 1637 and 1789. Progressively, the lakes were drained, initially through the Nochistongo cut and later through the Tequisquiac tunnels (1900 and 1952) and the Deep Drainage System (Emisor Central, 1975), and practically disappeared. A large part of the city was then built on lacustrine sediments which are highly compressible volcanic soft clays interbedded with layers of silt and sand and sandy gravels of alluvial origin. 2.1 Soil profile As shown in Fig. 1, the urban area of Mexico City can be divided in three main geotechnical zones: Foothills (Zone I), Transition (Zone II) and Lake (Zone III), as defined in the present building code [23]. In Zone I, very compact but heterogeneous volcanic soils and lava are found. These materials contrast with the highly compressible soft soils of Zone III. Generally, in between, a Transition Zone (Zone II) is found where clayey layers of lacustrine origin alternate with erratically distributed sandy alluvial deposits. The main difficulties for foundation of high-rise buildings are encountered in Zones II and III. An updated zoning will be available in 2015. In Fig. 2, a typical soil profile corresponding to the Lake Zone is presented. The water table is close to the surface. Three clayey layers are to be distinguished, denominated upper clay formation (Formación Arcillosa Superior, FAS), lower clay formation (Formación Arcillosa Inferior, FAI) and deep deposits (Depósitos Profundos, DP). The clays of FAS are separated from FAI by a hard layer (Capa Dura, CD), a sandy clayey stratum, some 3m thick, found at a typical depth of 30 to 35m. Generally, FAS


52

Deep foundations in Mexico City soft soils

is covered by a desiccated crust and/or an artificial fill with a thickness varying from a few decimeters to several meters. Average values of index properties

are presented in Table 1 for a typical borehole in the Lake Zone.

Figure 1. Geotechnical zoning of Mexico City [23].

Table 1. Typical average values of index properties in Lake Zone (Borehole Pc-28, Marsal, 1975). PROPERTY

FAS

CD

FAI

Water content, %

270

58

191 288

Liquid limit wL, %

300

59

Plastic limit, wP, %

86

45

68

Density of solids, Ss

2.30

2.58

2.31

Initial void ratio, e0

6.17

1.36

4.53

Unconfined compressive strength, qu, kN/m2

85

24

160

Spatial variations of soil properties in the lacustrine zone have been registered in a data base consisting of more than 10,000 pits and boreholes. This data base was incorporated into a Geographical Information System focused on geological and geotechnical features. Distribution of soil properties within any specific area of interest can be assessed from the database. Virtual soil profiles and cross sections can be defined using geostatistical estimation or simulation techniques. As an example, Fig. 3 shows the variation of soil water content within FAS along an EW axis. Figure 2. Soil profile in the Lake Zone of Mexico City (Marsal, 1975).


53

AUVINET G. et al.

Figure 3. Spatial variations of water content.

a)

2.2 Regional subsidence

0

2.3 Seismic site effects

150

200

250

300

350

0

Pore Pressure, kPa 50

100

150

200

250

300

0

Crust Crust 5

5 Hydrostatic Oct-80

Hydrostatic

Jun-92

10

Dec-94

10

Aug-05

Feb-97

Upper clay formation 20

Depth, m

Oct-99

15

15

Upper clay formation 20

25

25

30

30

35

Pronounced site effects leading to amplification of earthquakes affecting the basin of Mexico must also be taken into account in foundation design in the lake zone [52].

100

0

Depth, m

Exploitation of underground aquifers for supplying potable water to the growing population, causes a progressive depletion of the piezometric profile in most of the valley (Fig. 4). As a consequence, effective stresses within the soil increase and induce consolidation. Mexico City is thus affected by a regional subsidence that, in some locations has reached an accumulated value of 13.5m since 1862 [14]. Recent data show that the rate of subsidence tends to decrease in certain areas. However, in newly developed urban zones, such as the eastern parts of former lakes of Texcoco, Xochimilco and Chalco, the consolidation process is only in its first stage and the rate of subsidence can be as high as 40cm per year [14]. In transition zones between soft and firm soils, the subsidence induces differential settlements and soil fracturing, causing severe damages to pavements and small constructions [15,16].

b)

Pore Pressure, kPa 50

Hard Layer

Hard Layer Lower clay formation

35

Figure 4. Typical piezometric profiles in the lake zone of Mexico Valley.


35

54


3 TYPES OF FOUNDATIONS 3.1 Design criteria Foundations of buildings in the lake area of the basin of Mexico must be designed to contend with the severe conditions described in the preceding paragraphs. Using the terminology of the Building Regulations for the Federal District [23], foundations must provide adequate security with respect to multiple limit states: a) Failure limit states: floating, local or general plastic flow of soil under the foundation and structural failure of piles or other foundation elements. b) Service limit states: average vertical movement, settlement or emergence with respect to the level of the surrounding terrain, average inclination and differential deformation.

Safety with respect to these limit states must be guaranteed for short and long term static loads but also for accidental actions, particularly in seismic conditions. It should be recognized that, with few exceptions, before the 1985 Mexico City earthquake, designs almost exclusively focused on control of settlement or emersion of foundations under static conditions. The concern for controlling these movements led designers to use many types of foundation including traditional ones but also many special systems, some of them very ingenious. 3.2 Conventional foundations The solutions traditionally adopted for foundations of buildings in Mexico City range from isolated or continuous footings or slabs for buildings with few levels, to concrete point bearing piles for high-rise buildings (Fig. 5).

Figure 5. Conventional foundations on soft soils in Mexico City.

3.2.1 Surficial and compensated foundations Foundation on masonry footings or general raft, sometimes with short wood piles, was the first system tested by the founders of the city, with very little success at that as attested by the spectacular problems registered in the foundation of “Templo Mayor”, the main pyramid of the Aztecs, and of many heavy colonial temples such as the Metropolitan Cathedral. It is now accepted that surficial foundations on footings or surficial mats are only suitable for very light constructions occupying a relatively small area. It must be taken into account that a load of only 20kPa applied on a large area of the lake zone can be expected to induce a total settlement close to 1m with differential settlements of about 50cm. Moreover, these foundations are vulnerable to movements induced by adjacent buildings. Some of the problems faced when using surficial foundations can be managed recurring to compensated or “floating” foundation. The wellknown compensation technique consists of designing

the foundation, generally a box-type structure, in such a way that the mass of excavated soil will be comparable to the mass of the building [20]. Theoretically, if both weights are equal, the soil below the foundation is not submitted to any net stress increment and no significant settlement should develop. When the weight of the soil is smaller than the weight of the building, the foundation is partially compensated; in the opposite case, it is overcompensated. In practice, even perfectly compensated foundations undergo some absolute and differential vertical movements due to soil elastic deformation, to soil disturbance during construction and to static soilstructure interaction thereafter. Furthermore, constructing this type of foundation is not straightforward since a deep excavation in soft soil is generally required with the associated problems of stability of earth slopes or support systems and to bottom expansion or failure. Water tightness of the foundation is also a critical factor for compensated foundations; in many cases, this type of foundation


AUVINET G. et al.

must be equipped with a permanent pumping system to control infiltrations. 3.2.2 Point-bearing piles Precast or cast-in-place end-bearing piles embedded in a deep hard stratum are an apparently obvious solution for foundations on soft soils. Moreover, this technique has proven to be more reliable than other types of foundation in seismic conditions in Mexico City. However, foundations on point-bearing piles may present some substantial problems and their design may face severe difficulties. The bearing capacity of the hard layer in which the piles rest is a first source of uncertainty. The shortcomings of classical analytical methods for evaluating this capacity have long been recognized. Most of them assume rigid-plastic behavior of the soil ignoring the essential role of soil deformability. Bearing capacity estimations thus tend to be based principally on in situ tests (cone penetration test, pressuremeter) or on loading tests. Heterogeneity of these hard strata is difficult to assess and may originate tilting of buildings with such a foundation. In consolidating soils, negative skin friction develops on the pile shaft [4]. Moreover, an apparent emersion of the structure with respect to the surrounding area is generally observed (Fig. 6) and damage can be induced in adjacent buildings supported by other types of foundation. Consolidation has also the effect of separating the slab of the substructure from the soil. In that condition, the head of the piles is no longer confined and can be structurally vulnerable to the combined effect of seismic overturning moment and base shear [6].

55

1) shear stress developed on the shaft of a pile cannot be larger than the limit soil shear strength. 2) limit shear stress can only be reached when the soil attains the corresponding required shear deformation. 3) axial force developed in a pile due to skin friction within a pile group cannot be larger than the weight of the soil located within the tributary area of the pile. 4) unloading stresses induced by negative skin friction within the soil cannot be larger than those that are sufficient to stop the consolidation process that originates the skin friction in the first place. Curiously, many of the methods available to take into account the negative skin friction do not consider all of the above conditions, especially the last one. These factors can easily be taken into account using numerical (finite element) modelling [10, 50]. As mentioned above, foundations on point-bearing piles presented generally an acceptable behavior during the 1985 earthquakes. However, some cases of structural damages in the upper part of the piles were detected (Fig. 7). They were attributed to load concentrations in the perimeter of the structure due to overturning moment and base shear force.

Figure 7. Seismic damage in the upper part of a pointbearing pile (1985).

3.3 Special foundations Figure 6. Apparent protruding of a foundation on endbearing piles in Mexico City.

3.3.1 Objective

As recognized in Mexico City building code, when estimating the downdrag force induced on piles by negative skin friction, the following considerations should be taken into account:

Special deep foundation systems have been developed with the principal objective of avoiding both excessive settlement and apparent emersion associated to consolidation of the upper clay formation (Fig.8). Some systems also allow controlling the load transmitted to each pile.


56


Figure 8. Some special foundation systems for soft soils in Mexico City.

3.3.2 Types of special foundations The different systems all have in common the inclusion in the piles of a “fuse” (an element

presenting large deformations when a critical load is exceeded) allowing the construction to follow the regional subsidence

Table 2. Principal types of special foundations. Type

Fuse in lower part of pile

Friction piles

X

Piles with penetrating tip

X

P3 piles

X

Telescopic piles

X

Fuse in upper part of pile

Negative skin friction piles

X

Control piles

X

Overlapping piles

In Table 2, the principal systems have been regrouped according to the position of this element (in the upper or lower part of the pile, or both). The type of fuse used is characteristic of each system; in some instances, the fuse is the soil itself. Another solution, not included in the above table, consists of using piles placed within a flexible casing [55]. These piles are designed to avoid overloading of point bearing piles by negative skin friction. a) Friction piles Friction piles are generally used to transfer stresses induced by shallow or partially compensated foundations to deeper, less compressible layers of the subsoil, and to reduce settlements. Not so often, they constitute the main foundation system and the stability of the structure is dependent on the bearing capacity of the piles. A clear distinction must be

X

X

established between these two types of design (Fig. 9; [5]) Type I: Design in terms of bearing capacity In this first type of design, the number and dimensions of the piles are selected with the aim of guaranteeing that they will be able to support the load from the structure under static as well as dynamic conditions, with a safety factor generally larger than 1.5. In areas affected by regional subsidence, this type of friction pile is submitted to complex loading conditions (Fig 9). It has been shown [59, 46, 63, 4] that negative skin friction can develop on the upper part of the piles while positive friction develops in the lower part. A "neutral" level can then be defined between these two zones, where no relative displacement occurs between soil and piles.


AUVINET G. et al.

When the neutral level is in a low position (large number of piles or high strength of the lower layers), the downdrag force induced by negative skin friction may lead to significant compression stresses in the piles. Moreover, with time, the head of the piles can be expected to protrude from the surrounding ground

57

due to consolidation of the soil located between the surface and the neutral level, Fig. 10.

Figure 9. Friction piles. Train

Road

Pavement Slab

Slab apparent protrusion

Friction Hard layer piles Figure 10. Apparent protruding of a footing on friction piles (Type I design).

When this design philosophy is adopted, the bearing capacity of piles must be estimated taking into account the possibility of group behavior. When the density of piles is high, soil friction available on

the perimeter of the pile group plus its base capacity can in effect be smaller than the sum of the capacities of individual piles [45].


58


For piles working in the conditions indicated in Fig. 9a, settlements cannot be calculated by simplistic methods such as the well-known "2/3 rule" equivalent loading [42]. Depending of the position of the neutral level, the foundation can in fact present either settlement or emersion. Details of a more realistic method to estimate foundations movements adapted to these conditions were presented by Reséndiz and Auvinet [46] and recent developments of this model are summarized in paragraph 4.1. Type II: Design in terms of soil deformation In this case, only a limited number piles is used with the principal objective of reducing the settlements of a partially compensated foundation (compensated foundations with friction piles; [61, 62, 63, 64]). Since the number of piles is low, the neutral level generally coincides with the piles cap (Fig 8b). In that case, positive friction is mobilized along the full length of the piles, and the piles are in a permanent failure state, which justifies the name of “creep piles” that they were given by some authors [30]. Problems similar to those discussed for compensated foundations may occur. Reliability is low against excessive settlements in static conditions [8]. Without any doubt, foundations of this type were those that suffered most damages during the 1985 earthquake. 13% of all buildings between 5 and 15 stories, most of them on compensated foundation with friction piles, experimented settlement, tilting, punching of the soil (Fig. 11) and, in one case, total failure.

during a seismic event due to widening by lateral seismic forces of the perforation where the pile was installed and be further reduced by remoulding of the soil as cyclic shear stresses develop at the interface between soil and pile. This was confirmed by observations on an instrumented bridge on a box type foundation with friction piles [39, 40]. Instrumentation included load cells in piles and pressure cells below the slab of the substructure, as well as piezometers and accelerometers. A transfer of loads from the piles to the slab during earthquakes was clearly observed. Full scale experiments performed by Jaime et al. [32] have shown that piles fail when the combination of sustained plus cyclic loading exceeds the static bearing capacity during more than ten cycles. When the total loading exceeds this value by more than 20%, the subsequent sustained bearing capacity decreases to a value as low as 50% of the static capacity, while a penetration of the pile of 10 cm or more is observed. In the laboratory, some direct shear tests of the soil-concrete interface have also been performed [43]. The results show that static friction decreases significantly after cyclic loading For this type of design, it is thus commendable to ignore the contribution of the piles to the global bearing capacity. The bearing capacity to be considered under seismic conditions should be merely the capacity of the soil to resist the slab contact pressure. The presence of the piles should only be taken into account in static settlement estimations [23]. There has been a number of proposals aiming at increasing the efficiency of friction piles by modifying the shape of their cross section (triangular, H, etc.). Jaime et al. [32] have shown that this is generally not achieved. Among the research work aiming at improving friction piles, attempts to develop high adherence electro-metallic piles using electroosmotic treatment should also be mentioned [56, 57] b) Pile with penetrating tip.

Figure 11. Failure of a foundation on friction piles (Type II design) during the 1985 earthquake.

This type of pile [44] was conceived to increase the bearing-capacity of friction piles with a controlled contribution of the piles point. The diameter of the point is smaller than the rest of the pile in order to facilitate penetration in the hard layer under the combined effect of loading and negative skin friction and to avoid protruding. The point can be made of reinforced concrete [44, 21] or steel [45]. In the latter case, the bearing capacity of the pile can be better controlled by using a point with a pre-established failure load. Flexibility of the point has been found to be a problem during installation of piles.

A “creep pile” cannot be expected to absorb cyclic loading during earthquakes, since soil-pile adherence is already fully mobilized, and can even decrease SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

AUVINET G. et al.

c) Negative skin friction piles Those are simple point-bearing piles that penetrate freely through the foundation slab [17]. They can contribute to reduce significantly the settlements due to the negative skin friction that develops on the shaft of the piles under the combined effect of the structural load and the consolidation of the clay layer. Finite element modelling of this type of piles has been presented [50]. Spacing of piles appears to be the most significant design parameter for this type of foundation. d) Control piles The so-called “control piles” are similar to the previous ones but are equipped in their upper part with a mechanism that controls the load received by

59

each pile (Fig. 12). Piles can also be unloaded by removing the mechanism in order to correct any tilting of the building. Close to a thousand buildings are equipped with this system in Mexico City. Those systems have sometimes been installed during the life of the structure as part of an underpinning process [28, 29, 7]. The different available control mechanisms have been reviewed by different authors [36, 19, 1, 49]. In Table 3 a list of the best known systems is presented. In seismic conditions, some of these special systems can be vulnerable and suffer damage ranging from simple tilting to total collapse. Lack of maintenance can also be a problem. Several proposals have been made to improve the design of control piles [1]. Overturning of the loading frame can be avoided using a new type of anchors. The mechanism can also be adapted to absorb tensions.

Figure 12. Control mechanism (González Flores system).

Table 3. Principal types of control mechanisms for piles. Mechanism

Reference

Loading frame with deformable wood cubes

González Flores, 1948; Salazar Resines, 1978

Loading frame with jack and automatic relief valve

Pilatovsky, mentioned by J.J. Correa, 1980

Metallic tensors

González, 1957, mentioned by Aguilar, 1990

Metallic cap

Aguilar, 1960, mentioned by Aguilar, 1990

Loading frame with flat hydraulic jacks

Streu, 1963, mentioned by Correa, 1980 and Aguilar, 1990

Sand confined within a capsule

Creixell and Correa, 1975, mentioned by Aguilar, 1990

Energy dissipater

Aguirre, 1981; Reséndiz, 1976

Mechanical system of self-control

Jiménez, 1980

Mobile wedge

Girault, 1986, mentioned by Aguilar, 1991

Communicating hydraulic jacks

Zamora Millán, mentioned by A. Rico A., 1991

Constant friction cell

Támez, 1988

Cell with teeth for transmission of tensions

Rico, 1991


60


e) Telescopic piles

g) Rigid inclusions

These are tubular piles with a piston-like cylindrical point [18]. The tubular portion of the pile is partly filled with sand. When the sand reaches a certain level, an arching effect develops and both parts of the piles become solidary. If necessary, sand can be removed to free the point and avoid emersion of the foundation.

During the last decade, foundations on rigid inclusions have been used for a number of housing complexes in Mexico City [12, 13, 65, 66, 67, 68]. They consist of a simple slab resting on a layer of natural or artificial granular material (distribution layer) which in turn rests on structural elements, generally made of plain concrete or steel, previously placed in the ground (Fig. 13). In Mexico City, the distribution layer is not strictly necessary since a competent dry crust is found in the first meters of the soil profile. These "inclusions" constitute an economic foundation system for control of settlements. They have been used successfully for mid-rise buildings, which are only marginally affected by earthquakes [69]. To avoid disposal of large volumes of extracted soil, inclusions can be built using a displacement auger.

f) Overlapping piles This type of foundation [24, 25, 26] uses conventional friction piles (A Piles) combined with negative skin friction piles (B piles) lying on the hard layer. This arrangement reduces the stress increments in the soil and the corresponding settlements. Apparent emersion can also be avoided. This system has been used for the foundation of buildings and oil tanks in Mexico City [38]. It looks like this solution, used in Mexico City for many years, has been “rediscovered” recently by other authors [34]. Dried crust

Mat

Extra perimeter inclusion row 2m

19m

Upper clay formation

0.4m circular inclusions

8m

Hard Layer

Figure 13. Typical foundation on rigid inclusions.

h) Foundations on cast-in-place walls and structural cell Foundations on barrettes and cast-in-place walls are increasingly common in Mexico City. Special mention should be made of foundations consisting of four cast-in-place walls forming a structural cell as shown on Fig. 14 [37]. In this type of foundation, advantage is taken of the embedment of the cell into the soil to resist seismic actions. The width of the cell can then be significantly smaller than the one of a classical box-type foundation on friction piles, an important advantage in an urban context.

4

ADVANCES IN MODELLING OF DEEP FOUNDATIONS IN CONSOLIDATING SOILS

Different types of models have been developed along the years to represent the complex behavior of deep foundations in soft consolidating soils.

4.1 Analytical model [46, 3, 50] This model is useful to estimate the stresses in the complex field generated by loads caused by positive and negative skin friction on piles as described in Fig. 8. From the vertical stress distribution and taking into account the results of odometer tests, local


AUVINET G. et al.

settlement can be estimated. Two versions of this model are available. The first one [46] is a simplified model where equivalent horizontal areas with uniform loading at different depths are used to represent the positive

61

and negative skin friction acting on the piles shaft (Fig. 15).

Figure 14. Foundation on structural cell.

Figure 15. Simplified load distribution model for friction piles [46].

The second model [3] computes explicitly the stresses induced by each individual pile as well as by the contact pressure between the slab and the soil,

using elasticity theory (Mindlin and Geddes solutions). Due to its simplicity, the 1973 model is still one of the most commonly used by geotechnical designers.


62


An important improvement of this model has been introduced recently. To assess the negative skin friction in the upper part of a pile, consideration 4) of paragraph 3.3.2 is explicitly taken into account. According to this consideration, negative skin friction developed on the upper part of the pile cannot exceed the seepage forces due to piezometric depletion acting on the soil contained within the tributary area around the pile. The procedure implemented in software “MICRA” [41] is thus as follows: Neutral level elevation (z0), is defined solving the following equation:

Q  FN  FP  C P NP

FN  C f

z0 Df

The local neutral level for these piles will tend to occupy a higher position than for piles located within the group. Larger settlements should then be expected in the foundation perimeter, a fact confirmed by observations in many buildings. Introducing a downdrag force FN equal to  z0 AT in the model proposed by Reséndiz and Auvinet (Fig. 16) allows evaluating the local settlement (settlement at the elevation of the neutral level elevation) at different times in the future corresponding to different hypotheses regarding the evolution of pore pressure depletion profile.

(1)

where

Q

sum of permanent actions plus variable actions with medium intensity;

NP number of piles;

FP  C f

D f  LP z0

Positive friction equal to the limit

shear strength that can be developed from z0 to depth of pile tip (Df + LP);

Cp LP Df FN

estimated point bearing capacity length of piles; depth of substructure (slab)

negative skin friction estimated as follows:  C z0  f D f FN  mín. (2)   z0 AT

where:

Cf

z0 Df

estimated limit shear strength that may deve-

lop on the pile shaft from depth Df to z0.

 z0 effective stress increase at depth z0 (ignoring the presence of piles) generated by the pore pressure depletion estimated for the future at a given date. Note that  z0 AT is the resulting

Figure 16. Considerations for negative skin friction on a pile.

Comparing this local settlement to the estimated or computed regional consolidation of the soil between the tip and the bottom of the piles, it is possible to conclude whether the foundation will settle or protrude with respect to the surrounding terrain in the future (Fig. 17). An optimum design should aim at minimizing both settlement and protruding during the life of the structure.

force at the neutral level elevation due to vertical seepage forces acting on the soil around the upper part of the pile.

AT tributary area of each pile. Perimeter piles require special attention since the tributary area of these piles is not limited horizontally. In that case, the skin friction can be considered as: SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

AUVINET G. et al.

63

4.2 Numerical models

Figure 17. Evolution of general subsidence, settlement and resulting apparent movement.

local

Advances in numerical modeling of the static behavior of pile groups and inclusions in soils subjected to regional consolidation have been presented recently [50, 51]. Parametric studies were performed using 2D, axisymmetric and 3D numerical models of isolated piles and pile groups (Fig. 18),with advanced constitutive soil constitutive laws models (Soft Soil or S-Clay1, [60]). To simplify 3D models it has been found useful to resort to “slice” models taking advantage of symmetry and approximate plane strain conditions prevailing within piles groups (Fig 19).

Figure 18. 3D model of a group of piles [50].


64


Figure 19. “Slice” model for simplified 3D FEM analysis of piles groups [50].

Important conclusions were established: a) For the design of piles, it is very conservative to add the effect of negative skin friction to accidental loads. Indeed, when a pile is subjected to the effect of depletion of an initial pore pressure profile, if a load is applied to the head of the pile, it behaves similarly to a preloaded element, i.e. an important part of the force generated by negative skin friction is replaced by the surcharge and in some cases friction can become positive. As set forth by Fellenius [22], the problem of downdrag forces on piles is thus mainly a problem of deformation. This has been taken into account in the 2015 review of Mexico City building code to be published soon where it is explicitly indicated that: “This friction should be considered only for the structural revision of piles and for estimating the movements of the foundation (settlement or emersion)”. b) The depth of the neutral level (Fig. 8) tends to stabilize as the consolidation process unfolds. The depth of this level depends significantly on the initial loading conditions of the pile. The position of the neutral level depends more on the spacing between piles than on the magnitude of pore pressures depletion. c) The interaction between pile and soil is very sensitive to variation of the position of the groundwater level (WT). The transition between the dry and rainy seasons (raising WT) can generate the development of two neutral levels along the pile as was shown by Auvinet and Hanell [4]. d) For the numerical modeling of long-term behavior of a central pile which is part of a

e)

f)

g)

h)

i)

j)

group, it is not necessary to use interface elements at the soil pile contact. This is because their behavior depends more on the compressibility of the material than on its shear strength. The initial anisotropy of compressible material influences the behavior of the piles when the spacing between these elements is relatively large and when the stress level in the medium is low, i.e. in those cases in which the shape of the plastic yield surface is more important than the position of the failure criterion. Negative skin friction on point bearing piles due to depletion of pore pressures, cannot lead to “loss of confinement” (effective vertical stress smaller than the initial value) at the level of the support stratum, contrary to the concept proposed by Zeevaert [63]. Qualitatively, rigid inclusions present the same behavior as piles, however, the former are less efficient. This is because the external load on inclusions is transmitted both by the head and the shaft requiring some deformation of the distribution layer and the reinforced layer. The 3D models show that 2D axisymmetric models (cells representing the tributary volume of each pile; [13]) are applicable to foundations formed by a large number of piles. For the modeling of small groups of piles it is necessary to use 3D models, since in this case the edge and corner elements have a major influence on the overall behavior of the foundation. Stiffness of the surface layers and slab or box foundation is a factor that greatly influence the behavior of foundations formed by small groups of piles.


AUVINET G. et al.

k) The behavior of the corner and edge piles, submitted to depletion of pore pressures, is practically independent of the plan dimensions of the foundation, provided that the separation between elements is the same. l) For small groups of piles, stress concentration at the bottom of the piles due to depletion of pore pressures, generates a significant bending of the elements located in the perimeter and, to a lesser degree, of central piles. m) Limit shear strength conditions may be generated by negative friction on the shaft of the perimeter piles. For central piles, this condition is generally reached only at the tip and in the lower part of the shaft where positive skin friction prevails. Limit conditions for the negative skin friction on central piles can only be attained for large spacing between piles and high piezometric drawdown. n) Numerical modeling confirm that negative friction on central piles in consolidating soils cannot be greater than the apparent increase of the submerged weight of the mass of soil surrounding the pile above the neutral level, induced by seepage forces associated to the head gradient caused by pore pressure profile depletion. o) The analytical model by Reséndiz and Auvinet [46] modified to take into account the considerations set forth in the preceding paragraphs gives acceptable results and can continue to be used as a simple tool for the analysis of piled foundations in consolidating soils.

5 CONCLUSIONS The difficult geotechnical conditions prevailing in the lake zone of Mexico City, have led to the development of multiple solutions for foundation of high-rise buildings in soft soils affected by regional consolidation. Many valuable contributions to the analysis of foundations in these conditions have been made in recent decades. The role of numerical methods to obtain a better understanding of the behavior of these foundations has been particularly significant. New construction methods also have a strong influence on the deep foundation engineering practice. REFERENCES AND BIBLIOGRAPHY [1] Aguilar, J.M. and Rojas, E., Importantes mejoras en los dispositivos de control de pilotes, Memoria de la XVa Reunión Nacional de mecánica de suelos, San Luis Potosí, Sociedad Mexicana de Mecánica de Suelos, now Sociedad Mexicana de

65

Ingeniería Geotécncia, (SMMS now SMIG), México, 1990. [2] Aguirre, M., Dispositivo para controlar hundimientos de estructuras piloteadas, Publicación No 439, Instituto de Ingeniería, UNAM, México, D.F., 1981. [3] Auvinet, G., and Díaz Mora, C., Programa de computadora para predecir movimientos verticales de cimentaciones, Publicación N° 438 del Instituto de Ingeniería, UNAM, 71p., junio, México, 1981. [4] Auvinet, G. and Hanell, J.J. Negative skin friction on piles in Mexico City clay, Proc. Xth International Conference on Soil Mechanics and Foundation Engineering, 2 (1981), 599-604. [5] Auvinet, G. and Mendoza, M., Comportamiento de diversos tipos de cimentaciones en la zona lacustre de la Ciudad de México durante el sismo del 19 de Septiembre de 1985, Memoria, Simposio "Los sismos de 1985; casos de mecánica de suelos", SMMS, (1986), 227-240. [6] Auvinet, G. and Mendoza, M., Consideraciones respecto al diseño de cimentaciones sobre pilotes de fricción, VIIa Reunión Nacional de Ingeniería Sísmica, (1987), 19-21. [7] Auvinet, G. and Gutiérrez, E., Instrumentación de un edificio en proceso de recimentación, Memoria, Simposio sobre recimentaciones, SMMS, (1989), 137-148. [8] Auvinet, G., and Rossa, O., Reliability of Foundations on Soft Soils, Proceedings, Sixth International Conference on Applications of Statistics and Probability in Civil Engineering, CERRAICASP-6, (1981), 768-775. [9] Auvinet, G., Pecker, A. and Salencon, J., Seismic bearing capacity of shallow foundations in Mexico City during the 1985 Michoacán Earthquake, Proceedings, Eleventh World Conference on Earthquake Engineering, (CDROM), (1986). [10] Auvinet, G. and Rodríguez, J.F., Modeling of friction piles in consolidating soils, Proc. Int. Deep Foundation Congress, ASCE, (2002), 224-235. [11] Auvinet G. and Rodríguez, J.F., Behavior of endbearing piles in consolidating soils, Proc. Int. workshop, Foundation Engineering in difficult soft soil conditions, ISSMGE TC 36, (2002), 133-137. [12] Auvinet, G and Rodríguez, J.F., Inclusiones rígidas como alternativa de cimentación en suelos lacustres de la ciudad de México, Memoria, XXIII Reunión Nacional de Mecánica de Suelos e Ingeniería de Cimentaciones, Sociedad Mexicana de Mecánica de Suelos, (2006). [13] Auvinet, G. and Rodríguez, J.F., Modelling of rigid inclusions in consolidating soils, CD-ROM Proceedings, XIIIth Pan-American Conference on Soil Mechanics and Geotechnical Engineering, (2007)


66


[14] Auvinet, G., Land subsidence in Mexico City, Proceedings, ISSMGE TC36 workshop, “Geotechnical engineering in urban areas affected by land subsidence; the cases of Mexico City, Bangkok and other large cities”, (2009), 3-11. [15] Auvinet, G., Soil fracturing induced by land subsidence, in “Land subsidence, Associated Hazards and the Role of Natural Resources Development”, 339, (2011), 0-26 [16] Auvinet, G. et al., Evaluation of regional subsidence and soil fracturing in Mexico City Valley, This Conference (2015). [17] Correa, J.J., The application of negative friction piles to reduction of settlement, Fifth International Conference on Soil Mechanics and Foundation Engineering, (1961). [18] Correa, J.J., A telescopic type of pile for subsidence conditions, Proc. Specialty session on negative skin friction and settlements of piled foundations, 7th International Conference on Soil Mechanics and Foundation Engineering, (1969). [19] Correa, J.J., Estado actual del conocimiento sobre pilotes de control, Memoria de la Reunión Conjunta Consultores-Constructores Cimentaciones profundas, SMMS, (1980). [20] Cuevas, J.A., The floating foundation of the new building for the National Lottery of Mexico: an actual size study of the deformations of a flocculent structured deep soil, First International Conference on Soil Mechanics and Foundation Engineering, Harvard, (1936). [21] Ellstein, A., El pilote penetrante o P3, Memoria de la Reunión Conjunta ConsultoresConstructores Cimentaciones profundas, SMMS, (1980). [22] Fellenius, B.H., Recent advances in the design of piles for axial loads, dragloads, downdrag and settlement, ASCE and Port of NY&NJ Seminar, (1998).


Technical Committee




The use of micropiles technology in soft soil conditions El uso de tecnología de micropilotes en condiciones de suelo blando Federico PAGLIACCI1 1

Sc. in C.E. - Soilmec

ABSTRACT: Micropiles are small diameter structural elements that can be used for deep foundations and underpinning, soil consolidation and retaining supporting walls.Starting from the first application in 1952, the use of micropiles in underground engineering has increased considerably thanks to the versatility of the technology and the possibility to use less cumbersome equipment, resulting in a reduced impact on existing soil and superstructures.This technique is applicable in soft to firm ground conditions - in loose to medium dense sands and in cohesive soils.This presentation will present an overview of the various types of micropiles that have been used, the different drilling techniques that can be used according to soil nature and characteristics, general requirements for types of equipment and tooling needed, and design and construction recommendations. In addition, practical examples of the technology will be presented as mini case histories.

1 INTRODUCTION The use of micropiles in underground engineering has increased considerably thanks to the versatility of the technology and the possibility to use less cumbersome equipment, resulting in a reduced impact on existing soil and superstructures. Usually micropiles are described as: "Small diameter structural elements that can be used for deep foundations and underpinning, soil consolidation, retaining walls for deep excavations and tunnelling". A second, more recent definition is the following: “a small-diameter (less than 300 mm), replacement, drilled pile composed of placed or injected grout, and having some form of steel reinforcement to resist a high proportion of the design load.” (D. A. Bruce, 1999).

The origin of micropiles is certainly the work of Prof. Lizzi (1914-2003), who in 1952 registered the patent of the "root piles", small diameter piles for underpinning of buildings subject to excessive settlement (Figure1). Thanks to the fact that the piles could be built with very small and lightweight equipment and the subsequent possibility of operating in confined areas, the method was broadly used for underpinning historic buildings like Ponte Vecchio in Florence in 1966 and stabilising the leaning bell tower in Burano, Italy (Figure 2).

Figure 1. Typical scheme of “Root piles” for the consolidation of an ancient monument; the block of soil resulting from the reinforcement with metal "roots" acquired adequate mechanical properties to withstand the applied loads.


68

The use of micropiles technology in soft soil conditions

Figure 2. Root piles adopted for the consolidation of the leaning bell tower of Burano, Italy.

Thanks to the fact that the piles could be built with very small and lightweight equipment and the subsequent possibility of operating in confined areas, the method was broadly used for underpinning historic buildings like Ponte Vecchio in Florence in 1966 and stabilising the leaning bell tower in Burano, Italy (Figure.2). 2 MICROPILES CLASSIFICATION The micropiles construction consists of three main stages: a) drilling b) laying of the reinforcement steel element c) grouting The most commonly used method to classify micropiles is based on the type of selected grout injection method. The two main types of microplies can be classified as follows: 1) Low-pressure grout-injected microplies 2) High-pressure grout-injected micropiles

2.1 Low-pressure grout-injected micropiles (gravity backfilling) Micropiles are executed by simply filling the hole after laying the reinforcement steel element. The hole is filled from the bottom up using a water/cement mix, adding sand if needed. If tubular steel reinforcements are used, the tube itself acts as a tremie pipe through which the mix is pumped to the bottom of the hole and gradually fills it up to the surface of the tube, bonding it to the surrounding soil.

(manchettes). Grouting is a two-step process. The first step is identical to the one adopted for low-pressure injected micropiles: the space between the tube and the hole wall is filled from the bottom with the grout. After hardening of the initially placed grout, the second stage of grouting is performed. Through the nonreturn valves, by using a double packer, the grout is high-pressure injected in a selective way through each valve; this injection creates a bulb of consolidated soil that significantly increases the load bearing capacity of the micropile. 3 DRILLING TECHNIQUES The techniques used for drilling micropiles are those usually labelled "small diameter drilling" techniques. The key advantages of using these techniques is the possibility of drilling a large variety of rocks and soils, overcoming pre-existing foundations and other obstacles, boulders and rocky layers. Depending on the types of soils the following drilling techniques can be used: rotation with or without casing; rotation-percussion by means of Down-the-Hole Hammer (DTH) with or without casing; rotation-percussion by means of top hammer without casing. In the Table 1 the standard diameter and the maximum depth are reported, according to the three above listed drilling methods. A correlation between the drilling method and the type of soil, according to ASTM classification, is reported in Table 2. The same correlation, based on International Drilling Company (IDC) (Table. 3.1) is reported in Table 3.2.

2.2 High-pressure –grout injected micropiles In this case, the micropile tubular reinforcement bottom section is equipped with non-return valves


PAGLIACCI F.

69

Table 1. Micropiles standard diameters and maximum depth function of drilling technique. Standard diameter Max. drilling depth Drilling technique (mm) (m) Rotation with or without casing

80 - 400

50 - 70

Rotation-percussion by means of Downthe-Hole Hammer (DTH) with or without casing

80-230

60 - 150

Rotation-percussion by means of top hammer without casing

50-250

15 - 30

Table 2. Correlation between the drilling method and the type of soil, according to ASTM classification. Type of soil

Clay/Silt

Sand Fine

Medium

Gravel Coarse

Fine

Cobbles Boulders

Medium

Drilling Method Rotation Top hammer DTH

Rock classification

Table 3.1. IDC rocks classification. Type of rock

UCS (MPa)

1

Soft

Coal - Chalk - Marl - Weathered sandstone

2 - 50

2

Medium

Tuff - Slate - Dolomite - Limestone - Rhyolite

10 - 100

3

Hard

Limestone - Sandstone - Rhyolite

50 - 200

4

Very hard

Basalt - Diorite - Gneiss - Schist - Granite - Conglomerate > 200

Table 3.2. Correlation between the drilling method and the type of rock, according to IDC classification. Rock classification Drilling method 1 2 3 4 Rotary (rock drilling bt) Rotary (tricone) Rotary (diamond crown) DTH

3.1 Rotary drilling without casing This technique is used when there is no need to support the borehole walls or when the drilling fluid (water, bentonite, polymer) is capable of supporting the hole walls. Energy is transferred to the drill bit through the drill rods that are rotated and pushed by a rotary drive mounted on the drill rig. 3.2 Rotary drilling with casing This technique is used when there is the need to support the borehole walls or when use of drilling fluid

is not allowed by project restrictions. A double rotary drive can be used: one transfers torque to the string of inner rods, while the lower one rotates the casing. 3.3 DTH drilling without casing This technique is used when drilling rock. Percussion is applied on the bottom of the hole by a compressed-air driven hammer mounted at the bottom of the drill string through which compressed air is driven to the hammer. Once percussion is completed, air is driven back through the drill bit to clean all the drilling debris from the hole. Torque and


70


down crowd force are transferred by a rotary drive mounted on the drill rig. 3.4 DTH drilling with casing This technique is used in fractured rocks or in granular soils when the risk of collapsing of the borehole is high and drilling fluids do not guarantee support to the hole walls. The system (ODEX, TUBEX or SYMMETRIX depending on each manufacturer's trade mark) is based on the principle of slightly enlarging the hole diameter during penetration, to make it larger than the diameter of the casing which is

driven, without rotating, by the down-the-hole hammer powered by compressed air. The bit has a reamer that swings out to enlarge the hole diameter, and subsequently swings to the minimum diameter allowing the drill string and the DTH hammer to be lifted up, leaving the casing temporarily in the hole; the casing will be extracted once installation of the reinforcement cage and concreting are completed (Figure 3).

Figure 3. Scheme of the DTH drilling system and detail of the drilling bit and of the swing reamer.

3.5 TOP HAMMER drilling This technique is used in cohesionless soils and for shallow boreholes, to install a casing closed by a disposable bottom end in the soil. Basically it consists in "driving" in the soil a tube closed at the bottom. The percussion energy is supplied by a hydraulic hammer combined with a rotary head, both placed on top of the casing. 4 MICROPILE STEEL REINFORCEMENT A steel hollow pipe, a conventional steel cage (vertical bars and stirrup), or an H beam profile are generally adopted as steel reinforcement for low-pressure grout-injected micropiles. A hollow steel pipe equipped with non-return valves has to be used for high-pressure grout-injected micropiles. Non-return valves are located in the deepest pile section; in general, 2-3 valves per linear meter of pile are adopted (Figure 4).

Figure 4. Typical steel reinforcement hollow pipe for highpressure grout-injected micropiles.

5 GROUTING Micropiles are generally grouted with a cement-based grout. The adopted ratio between cement and water range from 1 to 2. The same injection system used for bored piles is adopted for low-pressure grout-injected micropiles: an injection pipe is lowered down to the bottom of the hole and the grout is then injected until it flows back to the surface; otherwise the same steel reinforcement hollow pipe is used. High-pressure grout-injected micropiles are injected with a two-step process: injection of cement sheath and injection of anchorage bulb.


PAGLIACCI F.

In the first stage, the cement mixture is pumped from the bottom to fill up the space between the hole walls and the external surface of the steel tube. This operation is performed to prevent the injection mix from leaking out during the second stage that is carried out under pressure. During the second stage –i.e. the injection of the anchorage bulb – the grout is injected through the non-return valves using a double packer (Figure. 5) lowered in the hollow tube and placed in a way that isolates every single valve. The pressurised mixture being injected breaks the cement sheath and penetrates the soil. Once the injection of all the valves is completed, the tube is washed to allow any subsequent injection if needed. As a general rule, injection of each valve is stopped upon reaching the maximum design injection pressure and/or the set flow rate. At the end of the injection operations, the tube is filled with mixture. This technique is similar to the one used for some years in soil consolidation, where the mixture is injected through PVC pipes equipped with non-return valves to fill the empty space in the soil.

71

In micropiles, the mixture is used to create a sort of "bulb" around the tube and allow transfer of loads to the soil. In the bulb injection stage, which is performed through each individual valve, the injection pressure, after the initial peak of about 6 MPa needed to "break" the sheathing mixture and then penetrate the soil, must be maintained between 2-3 MPa, with flow rate values ranging from 10 to 50 l/min. These pressure and flow rate values are necessary to let the mixture fill evenly the empty spaces in the soil and compact the soil around the drilled area thus creating a true bulb without any breaking (claquage) that may cause the mixture to infiltrate along the fracture lines far away from the stem, with little or no effect. In order to check the said pressure and flow rate parameters through each valve, it is necessary to isolate the valve. This is possible by using a packer that is lowered inside the tube down to the height of the valve to be injected. The two packers (upper and lower) are then stuck to the tube inner walls (mechanically or hydraulically) so as to isolate the tube section.

Figure 5. Hydraulic double packer.

6 BEARING CAPACITY DESIGN METHOD. Low-pressure grout-injected microplies are performed by reproducing the large diameter bored piles technique; therefore, the load bearing capacity has to be evaluated by adopting the same criteria used for bored piles. High-pressure grout-injected micropiles are designed by considering only the bearing strata, i.e. the level where the “bond length” is grouted. The design method of high-pressure injected micropiles (single stage or multiples stages) is based mainly on the theory of Bustamante-Doix, 1985. Micropiles are assumed to be formed by a free length, where no load is transferred to the soil, and a bond length “Lb” where the load-transfer mechanism develops. In this section, because of the performed grouting throughout the valves, the grouted borehole is enlarged via hydro-fracturing of the grout mass to give a grout root around the core diameter of the borehole (Figure 6).

Figure 6. Scheme of High- pressure grout-injected micropile.

In the bond length, because of the injection of the fluid through the valves, the hole diameter is enlarged to the diameter “Ds” that exceeds the drilled hole diameter “Dd”.


72


Considering that the load bearing capacity of a high-pressure injected micropile is due only to the lateral friction along the pile shaft and that therefore the base load bearing capacity is ignored, the ultimate load bearing capacity (Qu) can be calculated with the formulas (1) and (2) 𝑄𝑢 = 𝜋 ∗ 𝐷𝑏 ∗ 𝐿𝑏 ∗ 𝑞𝑎

(1)

𝐷𝑏 = 𝛼 ∗ 𝐷𝑑 (2) where: 𝐷𝑏

presence of stratified soils with very strong variations of the stress-strain ratio. The very low layers' contribution to load bearing capacity can be ignored. As a general rule, the bulb has to start at a depth of not less than 5 meters below the soil level with a length not less than 4 metres. The recommended safety factors for the working load, a function of the use of micropile, temporary or permanent, and of the type of stress, tension or compression, are reported in the Table 4. Table 4. Safety factor for the evaluation of the micropile working load Safety factor

= average actual diameter of the bulb

𝐷𝑑

= drilled hole diameter

𝛼

= non-dimensional coefficient, a function of the nature of soil, injection method and volume of injected mixture

𝐿𝑏

= bond length

𝑞𝑎

= bulb – soil ultimate adhesion

If the soil is stratified, the load bearing capacity of micropiles can be calculated by summing the effects of the different layers in which the bulb has been constructed, using the formula (3): 𝑄𝑢 = 𝜋 ∗ ∑𝑛1 𝐷𝑏𝑖 ∗ 𝐿𝑏𝑖 ∗ 𝑞𝑎𝑖

(3)

The symbols have the same meaning as above, while the "i" subscript indicates that the values pertain to the "i" layer among n layers the bulb runs through. Care must be paid to using the last formula in the

Scope of work

Tension

Compression

Temporary micropile

2

1,8

Permanet micropile

2,2

2

The values of the 𝛼 non-dimensional coefficient, a function of the nature of soil, injection method and volume of injected mixture, are reported in the tab. 5. The table details the minimum amounts of mixture to be injected. The Vp value pertains to the drilling volume not including the micropile reinforcement pipe steel volume. The bulb – soil ultimate adhesion qa, function of the Nspt value has been evaluated by BustamenteGianeselli for cohesionless and cohesive soils and it is reported in the Figures 7 and 8.

Table 5. 𝛼 non-dimensional coefficient and Minimum quantity of injected grout for different type of soils.

α coefficient Soil nature

Single stage grouting

Multiple stage grouting

Minimum quantity of injected grout Vi

Gravel

1,3

1,8

1,5 Vp

Sandy gravel

1,2

1,7

1,5 Vp

Gravelly sand

1,2

1,5

1,5 Vp

Sand (rough to fine)

1,1

1,4

1,5 Vp

Silty sand

1,1

1,4

Silt

1,1

1,4

Clay

1,2

1,8

1,5 Vp (Single stage grouting) 2,0 Vp (Multiple stage grouting) 2,0 Vp (Single stage grouting) 3,0 Vp (Multiple stage grouting) 2,0 Vp (Single stage grouting) 3,0 Vp (Multiple stage grouting)


PAGLIACCI F.

73

Figure 7. qa values for clay and silt and for simple injection (single stage grouting in tab.5) or repeated injection (multiple stage grouting in tab.5).

Figure 8. qa values for sand and gravel and for simple injection (single stage grouting in tab.5) or repeated injection (multiple stage grouting in table.5).

6.1 Numerical example of calculation of load bearing capacity As a theoretical example, the working load for a permanent, vertical compressed, micropile is calculated. The working load will be calculated for single stage grouting. The drilling diameter is assumed equal to 250 mm.

The 12 meter-bond length is assumed to be performed in a homogeneous layer of sand with Nspt value equal to 50.


74


𝑄𝑢 = 𝜋 ∗ 𝐷𝑏 ∗ 𝐿𝑏 ∗ 𝑞𝑎 = 𝜋 ∗ 0,35 * 12 * 30,6 = 403,55 ton Wl = 201,78 ton 7 MAIN APPLICATIONS

Figure 9. Calculation scheme.

The α coefficient value for single stage grouting, in sand, is equal to 1,1. Therefore, the average actual diameter of the bulb is equal to: 𝐷𝑏 = 𝛼 ∗ 𝐷𝑑 = 1,1 * 0,25 = 0,28 m The bulb – soil ultimate adhesion qa, value is obtained by entering the horizontal axis of figure 8 at the 50 value of SPT and intercepting the line 2 of the graph. The corresponding qa value is equal to 0,26 MPa (26,5 tonf/m2), as reported in Figure 10.

Micropiles can be used in several projects that foresee the construction of foundations with small drill rigs or whenever the type of soil makes it possible to use technologies like roto-percussion. The main applications are: retaining walls; underpinning and tunnelling.

7.1 Retaining walls Micropiles can be used successfully to support boreholes, especially in the presence of coarse material with boulders or blocks, that is whenever concrete diaphragm walls cannot be easily built. In these conditions, since the main stresses come from a combination of bending and shear, it is better to use low-pressure grout-injected microplies. Usually, anchors are installed to reduce the bending moment along the wall, as reported in Figure 11. When building an underground parking lot below or near existing buildings, it is possible to use micropiles as retaining walls in combination with floor slabs used as struts, all built with a top-down technique. If micropiles are constructed in the right positions, they can indeed become, with an additional reinforcement, the supporting pillars of the floors as reported in Figure 12.

Figure 10. Determination of the qa value.

The ultimate bearing capacity is therefore equal to: 𝑄𝑢 = 𝜋 ∗ 𝐷𝑏 ∗ 𝐿𝑏 ∗ 𝑞𝑎 = 𝜋 ∗ 0,28 * 12 * 26,5 = 279,59 ton The working bearing capacity, for a permanent, vertical compressed, micropile is therefore equal to: Wl = Qu /Fs = 279,59/2 = 139,79 ton Always as an example, if the grouting is performed as multiple stage grouting, the α coefficient value will be equal to 1,4 and the Db will become 0,35 m. In this case the qa value in Figure 10 will relate to the line 1 and therefore equal to 0,3 MPa (30,6 tonf/m2).

Figure 11. Micropiles and anchors for retaining wall.

The ultimate bearing capacity and the working load, for the same length and drilling diameter of the micropile will therefore be equal to:


PAGLIACCI F.

Figure12. Micropiles as retaining wall and pillars.

7.2 Underpinning As already mentioned, underpinnings were one of the earlier applications of micropiles. Several solutions

75

are available, depending on the features of the existing foundations. In case of old masonry structures, it is recommended to use a larger number of micropiles with a relatively small load bearing capacity to improve distribution and transfer of the load to the superstructure. The micropiles are usually connected to the existing foundations by making them adhere to the masonry work or via a concrete beam connected the old foundations. Several different solutions are available for concrete foundations. In Figure 13, some photos related to the project to support existing spread foundation plinths during the construction of a new auditorium under the ground level of an old industrial facility being renovated. Six micropiles for each plinth were constructed: four went through the plinth body and two were placed outside it. During the under-excavation, a concrete pillar was vertically cast-in-place for the purpose of enclosing the metal reinforcement of the micropiles supporting the superficial foundations.

Figure 13. Example of the use of microplies to support existing spread foundation plinths during the construction of a new auditorium under the ground level of an old industrial facility being renovated.

7.3 The Pont de Pierre in Bordeaux (France) The construction of the bridge started in 1811 and ended in 1821. Stability problems arose from the very beginning of the construction works. They were due to the short length of the wooden foundation piles that did not reach the deep layer of compact gravels (Figure 14). The problems were further worsened by significant settlement of the piers caused by the river stream that removed material from under the bed level

(scouring). In 1985 the settlement, under some of the piers, had reached 70 cm. The underpinning works involved only the piers subject to the most serious settlement forces. It was also decided that the 16 micropiles, 118 mm diameter, per pier should bear only 40% of the load weighing on the pier foundation. The micropiles had to bear a load of 300 tons each. The intervention stopped the increasing settlement of the pier.


76


Figure 14. The Pont de Pierre in Bordeaux (France) and a typical cross section of a pier.

Figure 15. 16 micropiles per pier have been executed from the bridge deck. The micropiles had to bear a load of 300 tons each.


Technical Committee




Pore pressure build-up due to pile driving in clayey deposits Desarrollo de presión de poro debido al hincado de pilotes en depósitos arcillosos Manuel J. MENDOZA 1, Miguel RUFIAR 2, Enrique IBARRA 2 and Marcos OROZCO 2 1

2

Research Professor, Instituto de Ingeniería, Universidad Nacional Autónoma de México Former graduate student, Instituto de Ingeniería, Universidad Nacional Autónoma de México

ABSTRACT: Experimental signals are depicted on a) the pore water pressure build-up and the effective stresses on the shaft of an instrumented pile model that are recorded in the laboratory during pile driving into reconstituted marine clayey soil and b) pore water pressure build-up, which was measured in lacustrine clay from Mexico City with a given distance between the piles during pile driving at a site where a friction pile-box foundation was built. Experimental results are analyzed in terms of different theoretical solutions and highlight the similarities and differences in the comparison between the predictions and data recorded in the laboratory and field.

1 INTRODUCTION There is currently significant interest in understanding the change in pore pressure in the clayey subsoil that surrounds a field of piles because its magnitude has considerable influence in the pile field’s axial load capacity. This study attempts to describe the change in pore pressure during the pile-driving process and that immediately after the completion of pile driving. When a pile is driven by hammering, pore pressures are induced in the surrounding soil that are so high that they surpass the pre-impact in situ effective vertical stress; this significantly affects the penetration of the pile and leads to an immediate load bearing capacity that is near zero. During the process of pile driving in the soft clayey soil of Mexico City, it is thus necessary to bind the piles with steel wires and straps because after only a few controlled hammer impacts, the pile easily penetrates the soil by several meters. Due to the subsequent dissipation of the pore pressure in the water, and the resulting gain in shear strength along with other phenomena that occur around the shaft, which are discussed in this article, the piles gradually achieve the capacity to withstand a given workload. This study reports on the pore pressure and the change in the effective stresses measured at the lateral face of a scaled–down model pile while it is driven into the soil by impact driving. These measurements were performed under controlled conditions in the laboratory using a reconstituted clayey soil of marine origin that was obtained from the sea bottom at the Sonda de Campeche in the Gulf of Mexico. The reconstituted soil was made

since a suspension condition, which was contained within a vessel with a diameter of nearly one meter. In addition, this study shows the pore pressure measurements recorded in the saturated mass of a clayey soil under an instrumented friction pile–box foundation at the instant when the piles were driven to a certain depth. The subsequent change in the pore pressure is also shown. This case corresponds to a foundation composed of a concrete box and friction piles, which is typical of the foundations that are frequently built in the lacustrine zone of Mexico City. The experimental observations in this study are analyzed using different theories to predict the increase in pore pressure that occurs during pile driving, and its dissipation over time; in particular, the similarities and differences between the laboratory and field observations, and the theoretical predictions are highlighted. 2 BACKGROUND AND OBJECTIVES In the geotechnical literature, there are various semiempirical and analytical methods for predicting the increase in pore pressure that occurs during pile driving and its dissipation with time, which leads to a resulting increase in resistance. Bogard and Matlock (1990) proposed correlations for predicting the time required to reach different resistance levels based only on local consolidation mechanisms. D’Appolonia and Lambe (1971) presented a method for the calculation of the pressure increase and its dissipation over time, while Vesic (1972) obtained an expression for estimating the increase in pore pressure near the shaft.


78

Pore pressure build-up due to pile driving in clayey deposits

The primary objective of this study is to succinctly review these methods and carry out a comparison between their predictions and two experimental observations of piles driven under controlled conditions. One of these experiments was performed using a small-scale model studied in the laboratory; the other experiment was performed using real piles in the field. The first measurements are part of a research project sponsored by the Mexican Petroleum Institute (Instituto Mexicano del Petróleo), testing an instrumented model pile. The results of this experiment have already been presented in previous papers (Mendoza et al., 2000a and Luna, 2002). The pore pressure and total pressure measurements on the model pile shaft help to define the effective stress state along the pile and its variation over time. Conversely, pore pressure measurements of the soil mass during the driving of 77 piles in an actual foundation, corresponding to Support No. 6 of the Impulsora overpass in northeastern Mexico City are also presented. The details of the stratigraphic conditions, foundation characteristics, and instrumentation have been described in previous papers (Mendoza et al., 1996; Mendoza et al., 1998a).

a function of the distance from the pile shaft and dissipates quickly over time. Poulos and Davis (1980) reported measurements of the increase in pore pressure u that were normalized with regard to the pre-driving in situ effective vertical stress ’vo and presented them as a function of the radial distance d from a driven pile of radius r (Figure 1).

3 THEORETICAL SOLUTIONS FOR THE EFFECTS OF INSTALLING PILES IN CLAY The method used to install piles affects the loadstrain behaviour of the soil-pile system due to changes in the initial state of the soil. The effects of piles driving into clay have been classified into four broad categories by De Mello (1969): • Remoulding or partial alteration of the soil structure near the pile; • Changes of the stress state of the soil in the vicinity of the pile; • Increase in pore pressure due to driving and its dissipation near the pile; and • Aging phenomena. This article focuses only on the pore pressure and its change over time. Some aspects related to aging have been presented in previous studies published by this group (Cruz, 2003). 3.1 Increase in pore pressure due to pile driving Several authors have reported measurements of excessive pore pressure in soil due to pile driving, including D’Appolonia and Lambe (1971) and Randolph et al. (1979). Some researchers have recorded the pore pressure along the pile shaft and found that it can reach or surpass the magnitude of the vertical effective stress by up to three times. However, this pressure excess decreases quickly as

Figure 1. Summary of the pore pressure measurements around a driven pile (Poulos and Davis, 1980).

Beyond a ratio d/r of 4 in non-sensitive clays and of approximately 8 for sensitive clays, the pore pressure is shown to decrease rapidly as distance increases. Beyond d/r = 30, the excess pore pressure is effectively zero. The data presented by Airhart et al. (1969) suggest that pore pressures are higher near the end of the pile (i.e., approximately three to four times the in situ effective vertical stress). Both phenomena are clearly observed in the measurements of the model pile presented and described below. Figure 1 includes the point corresponding to the measurement near the tip and at the interface (d/r=1) of the model pile after driving 3.2 Estimation of the increase in pore pressure at the pile shaft To predict the increase in pore pressure at the pile shaft due to driving, D’Appolonia and Lambe (1971) proposed the following expression (1).


MENDOZA M. et al.

u m  2s   1  K 0   u  A f (1)  vo  vo   where: u m = maximum increase in pore pressure;

K0= earth pressure coefficient; su = undrained shear strength; Af = Skempton pore pressure coefficient at failure, and ’vo = vertical effective stress in the soil. Comparisons reported by Lo and Stermac (1965) and Lo (1968) between Equation (1) and measured pile pore pressures after pile driving show an overall good agreement. Based on the expansion theory of cylindrical cavities, Vesic (1972) considered the stress-strain changes in undrained conditions and evaluated the increase in pore pressure at the shaft to be:

 

 R p   r  su  

where:

 f  0.707(3 Af  1)

u  0.817 f  2 Ln

Af 

(2)

q 1  sen   su 2 sen 

where: q = initial isotropic effective stress; ’ = internal friction angle, in terms of effective stresses.

Rp  I r where:

Ir 

1 r cos 

that the dissipation occurs only radially in a process described by the following partial differential equation.

  2u  1  u  u  ch  2     t  d  d   d

(6)

where: ch = consolidation coefficient in two dimensions, for horizontal drainage; u = excess pore pressure. From a practical perspective, solutions to Eq. (6), such as that shown in Figure 2 (Poulos and Davis, 1980), are used to estimate the time that must pass after driving before a load test can be performed. A rigorous analysis of the increase in pore pressure and the subsequent consolidation around a pile driven into clay was performed by Wroth (1979). During the driving process, the pile is modelled as the formation of a large cylindrical cavity. The changes in the stress and pore pressure were obtained via finite element analysis, where the Camclay model was incorporated. In these studies, it was concluded that the effective and total stresses adjacent to the pile immediately after driving may be directly related to the original shear strength of the soil and are essentially independent from the preconsolidation ratio.

(3)

R p = radius of the plastic zone.

E 2(1  ν) su

where:

79

(4)

I r = rigidity index; E = undrained soil elastic

modulus;  = Poisson ratio of the soil Conversely, Randolph et al. (1979) suggested that the excess pore pressure due to pile driving can be estimated by the following expression:

u  4su   'vo

Figure 2. Dissipation of the pore pressure near a pile (Poulos and Davis, 1980).

(5)

where: ’vo = the change in the effective stress due to the remoulding of the soil. In normal or marginally overconsolidated clays, ’vo can reach negative values; for sensitive clays, it reaches two to three times the value of su. 3.3 Dissipation of the excess pore pressure Soderberg (1962) proposed a relatively simple solution for predicting the dissipation of the increase in pore pressure around driven piles; he assumed

4 MEASUREMENTS ON A MODEL PILE Experimental measurements were performed on an instrumented model pile that was 90 cm in length and 2.64 cm in diameter. The descriptions of the model pile, its instrumentation, the data acquisition systems used, and the applied loads as well as the soil where it was driven have been reported in previous studies (Mendoza et al., 1998 and Mendoza et al., 2000); only a brief description is presented here.


80


The model features four axial-load and three bending transducers (CF) along its body, as shown in Figure 3. In addition, there are four total-pressure (PT) and four pore-pressure (PP) transducers along its lateral face at different heights.

Figure 4. Experimental process with the model pile.

The soil inside the O-97-5 oedometer used in this study corresponds to reconstituted marine clay from the Campeche Sound (Ibarra, 2002) having an overconsolidated condition (OCR=2) with a uniform pressure of 75 kPa along its surface. This pressure is kept constant throughout the test including during the driving process.

During driving, all model pile sensors were monitored and this activity was kept for nearly 24 hours after the driving process. The system set for applying loads to the pile consisted of a reaction frame, an electro pneumatic servo knob, a pneumatic actuator for the axial load, a load cell, and an LVDT linear displacement transducer. In addition, it featured a signal conditioning module, an analog/digital/analog (A/D/A) computer card, a computer interface, a servo amplifier for the servo knob control, and the software that controlled all of these components. The tests were performed (Rufiar, 2009) using a program that includes static and axial dynamic tests.

4.1 Brief description of the experiments

4.2 Experimental results with the instrumented pile

The experimental procedure of this study consisted of five stages, which are shown schematically in Figure 4 and are described below.

The measurements showing the increase in pore pressure due to pile driving from two tests, which are denoted A1 and A3, are presented below.

The driving of the pile was performed with the aid of a steel guide that directs the model pile while it is impact-driven. The impacts were provided via a stainless steel mass that had a mass of 2.98 kg (29.24 N). The mass was allowed to fall freely from a fixed height of 0.20 m. In general, driving of the model piles required 330 impacts.

Figure 5 shows the driving data, where the number of impacts required to drive the pile by 10 cm are reported.

Dimensions in mm

Figure 3. Schematic of the instrumented model pile.


MENDOZA M. et al.

81

an asymptotic trend, as shown in Figure 6, which confirmed the observations of Ahrart et al. (1969). A similar behaviour was observed in test A3, as shown in Figure 7, where a maximum increase on the order of 110 kPa was observed for cell PP3, which is located in the middle of the pile, which dissipated in the same manner as shown in test A1.

Figure 5. Records of pile-driving tests A1 and A2.

It is shown that 13 to 18 impacts were necessary to drive the pile by the initial 10 cm; conversely, for the 10-cm section between 0.5 m and 0.6 m, 75 to 78 impacts were required. This result clearly shows the increase in resistance of the pile shaft as it is driven into the clayey soil. Figure 6 shows the pore pressure that was measured at the shaft during and after the driving operation. It is shown that the pile driving generates pore pressures that increase with depth.

Figure 7. Variation in pore pressure, measured at the shaft of the model pile during and after driving. Test A3.

Certain man oeuvres were required in both tests that required temporarily stopping the driving a few minutes after the test had begun. For that reason, an immediate dissipation of the pore pressure occurred, as shown in the interval between 1000 and 1500 seconds in Figure 6. Because the pile driving was resumed, the pore pressure began to increase again. Figure 8 shows the normalized data from test A1 as a function of time; the normalization was performed with respect to the in situ effective vertical stress 'v0=v0-u0. It is shown that the installation of the pile causes a significant increase in pore pressure. A few seconds after driving began, an increase in water pressure u≈2.28'v0 was measured in cell PP4 near the end of the pile. In cell PP2, which is near the head of the pile, the lowest increase in pore pressure was detected (i.e., u≈0.88'v0).

Figure 6. Variation in pore pressure, measured at the model pile shaft during and after driving. Test A1.

The magnitude of the maximum increase in the pore pressure at cell PP4 was 170 kPa; for cell PP3, it was 140 kPa, and for cell PP2, it was only 60 kPa near the pile head. Thus, the u /’vo ratio varied from 2.5 to 0.9. These pressures were found to dissipate within approximately 20 hours and followed SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

82


the subsoil at different depths in different perforations; two were placed ex professo in the clay strata. The piezometers produced high-quality and rapid responses; as a result, the pore pressure of the site was immediately known. Three piezometers were of the resistive type with strain gauges (SG), and the remaining piezometers were of the vibratingwire type (VW); all were calibrated meticulously before experimentation. The ZD-2 and ZD-3 piezometers were driven directly into the clayey strata without a sand pocket; the others were placed in permeable strata. Table 1. Piezometer installation depths and type of surrounding soil. _____________________________________________________

Figure 8. Variation in pore pressure, measured at the model pile shaft after installation.

5 MEASUREMENTS IN A PROTOTYPE FOUNDATION

5.1 General description The foundation of support No. 6 of the Impulsora overpass (Mendoza et al., 1996; Mendoza et al., 1996; Mendoza, 2004), is the studied case history. It consists of a foundation box and 77 friction piles. The piles have a 50-cm square cross section and were driven to a depth of 30 m. The site is located in Zone III-Lago Virgen, which ensures that the piles are embedded in very soft soil of the First Upper Clay Formation (UCF); the First Hard Layer is at a depth of 33 m. Before pile driving began, a perforation with a diameter of 0.50 m was made for the initial 2 or 3 m of depth, which corresponded to the surface crust and artificial previous fills. The pile driving was performed with a Delmag No. 33 hammer, although it should be made clear that the energy delivered by this hammer is significantly larger than that required for this driving process. The soil resistance was broken with only two or three impacts, or even by its own weight, inducing considerably displacements to the piles. For this reason, they had to be fastened with steel cables to maintain control of their depth. 5.2 Instrumentation and monitoring system Some days before beginning the pile driving operations, six electric piezometers were installed in

Piezometer Type Depth, in m Soil type _____________________________________________________ ZD-1 SG 7.50 Sandy stratum ZD-2 SG 10.20 Clayey stratum ZE-1 VW 24.00 Sandy stratum ZD-3 SG 27.00 Clayey stratum ZE-2 VW 34.00 First hard layer ZE-3 VW 52.00 Deep deposits _____________________________________________________

The piezometers were placed around a central nucleus of the foundation, as shown in Figure 9. Different to the pore-pressure measurements, which were performed at the soil-shaft interface in the model piles, the measurements in this case history were performed in the central portion of the foundation where there were no piles The piles were concentrated toward the foundation edges, precisely under the foundation beams to achieve a better foundation performance during seismic events. Thus, the pore-pressure measurements were performed in the soil mass under the foundation at a distance of approximately 5 m from the surrounding piles.

Figure 9. Layout of the instrumented foundation

The SG resistive transducers recorded dynamic variations during the instant of pile driving using digital gauges that operated at a sample rate of 250


MENDOZA M. et al.

83

samples per second. The measurements reported here for the vibrating wire sensors were done with a portable frequency gauge, and correspond to static monitoring only. 5.3 Results The digital measurements of the pore pressures from the SG transducers detected variations in real time as any of the surrounding piles were driven. However, their amplitudes were reduced due to the distance between the pile and the piezometer. For this reason, more emphasis was placed on the medium-term monitoring of the pore pressure in the subsoil under the central part of the foundation, which originated from the full pile field rather than from an individual pile. Figure 10 shows the pore pressure measured at six different depths under the foundation and includes the measurements during driving and days after driving to demonstrate the change in pore pressure over the subsequent four weeks. The increase in pore pressure clearly manifested in the UCF, resulting from the strong distortions in the subsoil due to pile driving. Because no pre-boring was practiced, the presence of the piles generates a volumetric displacement of approximately 500 m 3; which was reflected in a land surface expansion of up to 11.5 cm. The increase in pore pressure due to pile driving is shown to be highest at depths near the pile tips and is independent of whether the piezometer is lodged in a sandy or in a clayey layer. Note that the curves describing the pressure change in the piezometers ZE-1 and ZD-3 located at depths of 24 m and 27 m, respectively, are parallel. The ZD-3 piezometer recorded an increase of 21% over the previous hydraulic pressure; the pore pressures in the zones closer to the piles must be higher than this value. The piezometers were separated by a distance of approximately ten times the length of their side. The increases in pore pressure measured by the other piezometers (ZD-1 and ZD-2) lodged in the UCF were lower, which can be interpreted as a result of smaller distortions in the shallower portions of the ground, and in the more permeable zones, where these increases in pressure dissipate more easily. Although the pile-driving process clearly had an effect on the UCF soil in terms of the variation of its pore pressure, it did not affect the pore pressure in the First Hard Layer or in the Deep Deposits, thus reflecting the high permeability of their soils. Conversely, the speed with which the excess pore pressure caused by driving dissipates with respect to the initial in situ condition calls one’s attention in Figure 10.

Figure 10. Measurements of pore pressure during and after pile driving at Support 6 of the Impulsora overpass.

6 COMPARISONS BETWEEN THE MEASUREMENTS AND PREDICTIONS Some of the solutions discussed in Section 3 were applied for comparison to the measured values from both the laboratory tests with the pile model and the field experiment using friction piles for Support No. 6 of the Impulsora overpass. The results are summarized in Table 2. Table 2. Comparisons between measured and predicted values according with various authors. _____________________________________________________ Measured

D’Appolonia Cavity and Lambe expansion theory kPa kPa kPa _____________________________________________________ Case

Model pile 140 223 170 Impulsora overpass piles 48 71 52 _____________________________________________________

In both the laboratory and field experiments, an overestimation of the increase in pore pressure due to driving is shown in the theoretical predictions. Karlsrud and Haugen (1985) reached a similar conclusion when comparing their results with other experiments. For the model pile case at the level-3 sensor, this overestimation is approximately 20% when considering the cavity expansion theory solution. With this solution, a prediction is obtained that agrees with the measurements from the Impulsora bridge piles. The solution proposed by D’Appolonia and Lambe produces results that are approximately 50% higher than those measured in both the laboratory and the field tests.


84


7 CONCLUSIONS It has been possible to verify through laboratory measurements and field observations that driving a pile using impacts in the absence of preboring generates high water pressures at the soil-shaft interface. In laboratory tests, pore pressures at the shaft were shown to reach values that were 2.5 times the pre-driving vertical effective stress; this increment shows that at the moment of impact, the effective stresses on the pile shaft will be lost transiently along with the shear strength of the soil. As a consequence, the pile can penetrate easily into soft clayey soil. It is common practice in the virgin region of Mexico City’s Zone III to bind piles to prevent them from sinking out of control after a few hammer impacts. The dissipation of such pore pressure after pile driving triggers a local consolidation process. This phenomenon leads to an adjustment of the effective stresses around the pile and an increase in the soil’s thixotropy and thus determines the load capacity generated in a friction pile over time. Thus, after up to a few months, the load capacity can withstand operational loads. Several theoretical solutions have been discussed for calculating the pore pressure that is generated around a pile after being driven into clayey soil. Among the theories tested, the cavity expansion theory provided the estimation that best agrees with the measurements, both in the laboratory and in the field. REFERENCES Cruz, E. (2003). “Static and dynamic test on pile models.” Master Thesis, División de Estudios de Posgrado, Facultad de Ingeniería, UNAM, México (in Spanish). D’Appolonia, D. J. and Lambe, T. W. (1971). “Performance of four foundations on end-bearing piles”. J.S.M.F.D., ASCE, vol. 97, SM1, pp.77-93. De Mello, V. F. B. (1969). “Foundations of buildings on clay.” State of the Art Report: 49-136. Proc. 7th Int. Conf. S.M. & F.E., México. Ibarra, E. (2002). “Reconstitution of a marine clayey soil in an oedometer for pile models testing.” Master Thesis, División de Estudios de Posgrado, Facultad de Ingeniería, UNAM, México (in Spanish). Karlsrud, K., and Haugen, T. (1985). Axial static capacity of steel model piles in overconsolidated clay. Publication No. 163. Norwegian Geotechnical Institute, Noruega, Oslo. Luna, O. J. (2002). “Design, construction and operation of friction pile models under static and cyclic loading”, Master Thesis, División de Estudios de Posgrado, Facultad de Ingeniería, UNAM, México (in Spanish).

Lo, K. Y. and Stermac, A. G. (1965). “Induced pore pressures during pile driving operations.” Proc. 6th Int. Conf. S.M. and F. E., vol 2, pp 285-290. Mendoza, M. J., Romo, M. P., Barrera, P., Olivares, A., Rojas, E., Sánchez, J., Luna, O. J. and Valle, C. (1998). “On experimental study of friction pile models for offshore platforms”, Proc. XIX Nat. Meeting on Soil Mechanics, SMMS, Puebla, pp 303312. Mendoza, M. J., Romo, M. P., Domínguez, L., Orozco, M., Noriega, I. and Velasco, J. (1996). “Instrumentation and behavior of a piled-box foundations in Mexico city, during its construction and initial operation”, Proc. XVIII Nat. Meeting on Soil Mechanics, SMMS, Morelia, pp 143-159. Mendoza, M. J., Romo, M. P., Orozco, M., Domínguez, L., Velasco, J. M. and Noriega, I. (1998). “Loads on piles, contact and pore water pressures induced by earthquake in a piled-box foundation in Mexico City”, Proc. XIX Nat. Meeting on Soil Mechanics, SMMS, Puebla, Vol. 1: 358-367. Mendoza, M. J., Luna, O. J., Ibarra, E., Olivares, A., Barrera, P. (2000). “Instrumentation of a smallscale pile model: design and manufacturing”, Proc. XX Nat. Meeting on Soil Mechanics, SMMS, Oaxaca, pp 321-328. Mendoza, M. J. (2004). Behavior of a piled-box foundation in Mexico City, under static and seismic loading, Doctoral Thesis, División de Estudios de Posgrado, Facultad de Ingeniería, UNAM, México (in Spanish). Poulos, H. G. and Davis, E. H. (1980). Pile Foundation Analysis and Design, John Wiley & Sons, New York. Randolph, M. F. and Wroth, C. P. (1979). “Driven piles in clay –effects of installation and subsequent consolidation-”, Géotechnique 29 No. 4, pp 361-393. Rufiar, M. (2009). “Behavior of instrumented pile models in marine clayey soils under axial static loading”. Master Thesis, Sección de Estudios de Posgrados e Investigación, ESIA-IPN, Mexico (in Spanish). Soderberg, L. O. (1962). “Consolidation theory applied to foundation pile time effects.” Géotechnique, Vol. 12, pp 217. Vesić, A. S. (1972). “Expansion of cavities in infinite soil mass”, Journal of the Soil Mechanics and Foundations Division, ASCE, Vol. 98, SM3, Proc. Paper 8790, March, pp 65-290.


Technical Committee




Geotechnical design of the foundation for an office building located at the transition zone Diseño geotécnico de la cimentación para un edificio de oficinas localizado en la zona de transición Fernando ARENAS1 & Alberto CUEVAS 1 1Ingenieros

Cuevas Asociados S. C.

2.3 Laboratory tests.

1 . INTRODUCTION Reference is made to an office building that has been contemplated to be built at Insurgentes Sur Avenue, between streets Eje 5 (Eugenia) and Eje 4 (Xola), Colonia Nápoles, in the Transition Zone of Mexico City. The structure has been designed to have seven basement floors, ground floor, nine stories and terraced roof with a concrete frame design. The excavation to accommodate the basement floors will reach a depth of 24.0 m below sidewalk elevation. 2 2. GEOTECHNICAL INFORMATION AVAILABLE 2.1 Geotechnical zoning The site under study is located at the so-called Transition Zone (Ref. 1) being characterized by stratigraphic discontinuities produced by crossed alluvial deposits; their frequency and distribution depends on the closeness to old gorges in the western hills. These materials are underlain by clay strata covering typical deposits of the Hill Zone. 2.2 Field works To be able to define the local stratigraphy, two combined borings were drilled, alternating the techniques of electric cone probing with the Standard Penetration Test to depths of 40.0 and 45.0 m, respectively; a boring with selected sampling was also advanced to recover undisturbed specimens using the technique of Shelby-type thin walled pipe samplers. In addition, to define the conditions of the pore-water pressure a piezometric station was installed incorporating the following instruments: a tell-tale pipe driven to a depth of 6.0 m and three Casagrande-type open piezometers installed at depths of 15.2, 20.5 and 25.2 m, respectively.

Properly protected and identified samples were moved to our laboratory to determine their index properties: natural water content, visual and manual soil classification, grain size distribution, percentage of fines and consistency limits. The undisturbed samples were subjected to unconsolidated undrained triaxial compression tests and to one-dimensional consolidation tests. 2.4 Stratigraphic interpretation. Based on the field works and on the laboratory test results the following stratigraphy was defined (Fig. 1). - From 0.0 to 0.2 m, plain concrete slab. - From 0.2 to 1.1 m, fill constituted by fine and coarse sand of pumice and andesite origin, with brick fragments. - From 1.1 to 3.5 m, superficial crust integrated by silty clay with fine, medium and coarse sharp sand, of andesite and quartz origin, with stiff to very stiff consistency. - From 3.5 to 6.0 m, sandy silt with stiff to hard consistency with some lenses of hard consistency. - From 6.0 to 8.6 m, dense pumice-type fine to medium silty sand. - From 8.6 to 17.6 m interstratifications of organic silt with stiff to hard consistency and highly plastic silt with fine to medium sand of quartz and pumice origin and slightly andesitic origin with stiff to hard consistency, and highly plastic silt, with quartz and pumice and some andesitic origin and stiff to hard consistency with interbedded fine to medium pumice-type semidense and dense silty sand lenses. - From 17.6 to 28.5 m on the average, very fine, fine and medium dense to very dense silty sands of andesite, and pumice origin. - From 28.5 to 31.0 m on the average, andesitetype, sharp medium and fine very dense gravel


86

Geotechnical design of the foundation for an office building located at the transition zone

with andesite-type coarse, medium and fine sand and some silt. - From 31.0 to 33.0 m on the average, fine, medium and coarse silt with hard to very hard consistency with fine, medium and coarse andesite and quartz sand. - From 33.0 to 34.4 m, dense fine to medium silty sand of quartz, andesite and pumice origin and subangular particles. - From 34.4 to 45.0 m, very dense fine, medium and coarse silty sand of quartz and andesite origin and sharp particles, with some sharp andesite coarse gravel.

3.4 Frictional bearing capacity. The admissible frictional load bearing capacity was calculated by using the following expression: Q fa 

σ mi P l i tg φ i FS

(2)

where: Qfa, admissible frictional bearing capacity, t; P, perimeter of drilled shaft, m; mi, mean effective stress at strata of interest, t/m2; li, length of drilled shaft at strata of interest, m;  angle of internal friction of strata of interest; FS, safety factors equal to 2 and 1.7 for static and dynamic conditions, respectively.

2.5 Hydraulic conditions. Based on monitoring of the piezometer station installed it was found that the telltale pipe registered a superficial water level beyond a depth of 4.4 m, whereas piezometers at depths of 15.2, 20.5 and 25.2 m, respectively are dry. It was determined that perched water exists.

3.5 Total admissible load bearing capacity. The total admissible load bearing capacity is determined as follows:

QT  Qfa  Qpa

(3)

where: QT, total load bearing capacity under compression, t.

3 FOUNDATION ANALYSIS 3.6 Tensile bearing capacity. 3.1 Seismic coefficient. The seismic coefficient for structural design was assumed equal to 0.32 (Ref. 2) corresponding to the Transition Zone. 3.2 Foundation solution. Based on the stratigraphic information and on the characteristics of the structure, the foundation solution consists in drilled shaft foundations resting at a depth of 35.0 meters. 3.3 Point bearing capacity. The point bearing capacity of the drilled shaft foundations was calculated by means of the following equation (Ref. 3): Q pa 

σ o N q Ap FS

he tensile bearing capacity of the drilled shaft foundations is equal to the admissible frictional resistance plus the weight of the element. 3.7 Settlements. The settlements to be expected as a result of load transmitted to the mass of soil were calculated using the following expression (Ref. 3): δ

(Q pa  0.66 Q fa ) L 10 A p Ec



0.36 Q pa D 10 A p Es

(4)

where: , settlement, cm; D, shaft diameter, cm; L, shaft length, cm; Ap, cross sectional area, m2; Ec, modulus of elasticity of concrete for f’c= 250 kg/cm2, 221,359 kg/cm2; Es, modulus of elasticity of foundation soil, 1500 kg/cm2. 3.8 Modulus of vertical reaction.

¨

(1)

where: Qpa, admissible point bearing capacity, t; o, effective stress at foundation elevation, 23.5 t/m2, Nq, bearing capacity factor proposed by Berezantsev for an angle of internal friction  = 37o, 100; Ap, area at tip of drilled shaft, m2; and FS, safety factor equal to 3 and 2 for static and dynamic conditions, respectively.

The modulus of vertical reaction was calculated by using the following expression: kv 

QT δ

(5)

where: kv, modulus of vertical reaction, t/cm; Qt, net static bearing capacity, t; settlement, cm. Table 1 summarizes the conditions for different shaft diameters.


ARENAS F. et al.

87

N.P.T +43.60 N.P.T +42.60

Nivel

8

Nivel

7

Nivel

6

Nivel

5

Nivel

4

Nivel

3

Nivel

2

Nivel

1

N.P.T +38.40

N.P.T +34.20

N.P.T +30.00

N.P.T +25.80

N.P.T +21.60

N.P.T +17.40

N.P.T +13.20

N.P.T +9.00 Estacionamiento 1

N.P.T +4.50

SM-2 N.P.T +0.00

Planta Comercial

0

SM-1 0

Estacionamiento S - 1

(CH) Arcilla café olivo de consistencia rígida.



N.P.T -4.71 5

5

(MH) Limo café con arena fina pumítica y cuarzosa de consistencia rígida. Estacionamiento S-3


N.P.T -7.85

Ceniza volcánica muy compacta. Estacionamiento S - 3

10

N.P.T -11.00


10


N.P.T -14.14 15


1 5


7

6


10

8

15

N.P.T -17.28

1 2 3 4

11


9


N.P.T -20.42

20

20 Estacionamiento S - 7

12

N.P.T -23.57 25

25

30

30

35

35

40

40

13

14

15 16 0 10 20 30 40 50 NÚMERO DE GOLPES

18

17

(GP) Gravas muy compactas gris andesíticas angulosas con arena fina a gruesa andesítica.

45 0 10 20 30 40 50 NÚMERO DE GOLPES

1 (SM) Arena pumítica compacta. 2 (SM) Arena pumítica semicompacta. 3 (OH) Limo orgánico de color negro. 4 (MH) Limo arenoso gris verdoso de consistencia dura. 5 (OH) Limo orgánico café muy rígido. 6 (MH) Limo arenoso gris verdoso muy rígido. 7 (SM) Arena pumíta compacta con limo gris verdoso. 8 (MH) Limo gris verdoso muy rígido.

9 (SM) Arena limosa muy compacta gris con gravas andesíticas subredondeadas. 10 (MH) Limo gris verdoso de consistencia semirígida. 11 (SM) Arena limosa muy compacta gris verdoso con grumos andesíticos redondeados. 12 (SM) Arena limosa muy compacta café verdoso (arena fina a media, andesítica cuarzosa, subredondeada). 13 (SM) Arena limosa muy compacta café rojizo con gravas (arena fina a gruesa; gravas andesíticas subangulosas).

14 (SM) Arena limosa muy compacta café amarillento (arena fina a gruesa andesítica, pumítica, cuarzosa subangulosa). 15 (SM) Arena limosa muy compacta café rojizo, fina a gruesa andesítica, pumítica con intercalaciones de limo arenoso café rojizo. 16 (SM) Arena muy compacta café verdoso fina a gruesa andesítica subangulosa. 17 (SM) Arena limosa muy compacta café rojizo. 18 (SM) Arena muy compacta café verdoso fina a gruesa andesítica subangulosa.

Figure 1. Stratigraphic interpretation.

Table 1. Admissible load bearing capacity versus shaft diameter. Dimeter m

Static

0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00

110.5 147.4 184.2 221.0 257.9 294.7 331.6 368.4

Qf a , t Dynamic 130.1 173.4 216.8 260.2 303.5 346.9 390.2 433.6

Qpa , t Static

Dynamic

221.5 393.7 615.2 885.9 1205.8 1574.9 1993.2 2460.8

332.2 590.6 922.8 1328.8 1808.7 2362.4 2989.9 3691.2

QT = Qf a + Qpa , t Static Dynamic

 cm

kv t/cm

332.0 541.1 799.4 1106.9 1463.7 1869.6 2324.8 2829.2

1.79 2.12 2.47 2.83 3.20 3.56 3.93 4.30

185.83 255.03 323.22 390.74 457.80 524.52 591.01 657.32

462.3 764.0 1139.6 1589.0 2112.2 2709.2 3380.1 4124.8

Symbols: Qfa Admissible frictional bearing capacity; Qpa Admissible point bearing capacity; QT, Total load bearing capacity;  Settlement; kv Modulus of vertical reaction. These resistances shall be compared against the service load. 3.9 Bearing capacity for cut-off wall. Both, the point bearing and the frictional capacity of the cut-off wall (known as Milan wall in Mexico) were calculated (resting at a depth of 28.0 m), by means of expressions 1 and 2, respectively. The results show

an admissible frictional bearing capacity of 173.1 and 203.6 t/m whereas the admissible bearing capacity varies between 100.0 and 200.0 t/m. In addition, calculation was made of the negative skin friction (FN) to be developed along the cut-off walls at


88


depths from 0.0 to 15.5 m through the use of the following equation:

FN  0.3 ω



with

Nφ 

σ o dz

(6)

where: , perimeter of the cut-off wall, m; and ʃo dz, integral of the effective stress at the length where the negative skin friction is developed, 183.75 t/m. Substituting the corresponding values, the resulting value of the negative skin friction becomes FN= 55.0 t/m. 3.10 Net admissible load bearing capacity for cut-off wall. With the previous results, the net admissible load bearing capacity of the cut-off wall is equal to 217.9 and 348.6 t/m, for static and dynamic conditions, respectively.

1  sin φ 1  sin φ

where: Pd, horizontal pressure envelope corresponding to the active condition, t/m2; γi, unit weight of each stratum, t/m3; hi, thickness of each stratum, m; ci, undrained cohesion for each stratum, t/m2; Nφi, factor depending on the angle of internal friction corresponding to each stratum; ui; pore water pressure, t/m2; qi, surcharge at surface, 2.0 t/m2; D, depth of excavation, 24.0 m; and , angle of internal friction, in degrees. 4.2 Passive horizontal pressures. The passive horizontal pressures acting at the internal part of the wall reacting against the soil were evaluated as follows (Ref. 4):

σ hi  γ i h i N φi  2 ci

3.11 Settlements of cut-off wall. The settlements were calculated with equation (4); their magnitude became equal to 2.9 cm.

(8)

N φi

(9)

where the symbols have been previously defined. 4.3 Long-term horizontal pressures.

3.12 Modulus of vertical reaction for the cut-off wall. The modulus of vertical reaction was determined from expression (5) and its magnitude amounts to kvMilán= 93.51 t/cm.

The diagram of horizontal pressures to be withstood by the cut-off wall was calculated by means of the following expression (Ref. 4): Ph  K o (γ i z  q i  u)  u

(10)

where: Ko, at-rest pressure coefficient; the other symbols have been already defined.

4 SHORT- AND LONG-TERM HORIZONTAL PRESSURES

4.4 Seismic force. 4.1 Short-term horizontal pressures. For purposes of design of the peripheral retaining wall a maximum depth of excavation of 24.0 m was considered, with cisterns included. Calculation was made of the horizontal stresses associated to the mass of soil; the envelope of the active horizontal pressure (per linear meter of width) to be developed against the wall was determined using the criterion of Terzaghi-Peck and applying undrained shear strength parameters corresponding to the most unfavorable condition (Ref. 4):

Pd 

2 ci 1.25 γ i h i  u i q (  ) dh i  u i  D N φi N φi N φi



The seismic force, Fsis,to be developed in the soil mass against the peripheral walls was evaluated pursuant to the Complementary Technical Standards applicable to seismic design; its magnitude was calculated as follows (Ref. 2): Fsis 

W 4 ao 3

(11)

where: W, weight of the active wedge (considering the surcharge of 2.0 t/m2), t; and ao, ordinate of the design spectrum for time T=0, equal to 0.08. After substituting the corresponding values the diagrams of short-term and long-term horizontal pressures were plotted as shown in Fig. 2.

(7)


ARENAS F. et al.

89

Figure 2. Diagram of short-term and long-term horizontal pressures.

E p a  M pt

5 DESIGN OF THE EXCAVATION AND OF ANCHOR SYSTEM

FS 

To excavate the enclosure accommodating the basement levels it will be necessary to stabilize the wall by means of a retaining system constituted by a 0.40-m thick reinforced concrete cut-off wall founded at a depth of 28.0 m complemented with five rows of anchors.

where: Ep, passive earth pressure, t/m; a, lever arm of the resultant of passive earth pressure at the bottom level of anchors, m; Mpt, plastic moment of sheet pile, 35.0 t-m/m; Ea, active earth pressure, t/m; d, lever arm of the resultant of the active pressure at the bottom level of anchors, m. Substituting the corresponding values a safety factor of about 2.0 is obtained. Based on the diagrams of short-term horizontal pressures the design of the anchor system was accomplished to complement the peripheral retaining wall during the excavation works and construction of the foundations; it involved five anchor levels. Fig. 3 shows a proposal for the layout and distribution of anchors.

5.1 Review for bottom failure. The safety factor against plastic flow (creep) at the bottom of the excavation (bottom failure) was determined using the criterion of bearing capacity assuming a prism of soil gravitating at the elevation of the excavation bottom having a width Be representing a hypothetical footing:

FS 

0.5 γ Be N γ Pd  q

(12)

where: , unit weight of the soil below excavation level, 1.65 t/m3; Pd, total vertical pressure at elevation of the excavation bottom, 37.4 t/m2; q, superficial surcharge, 2.0 t/m2; Nbearing capacity factor, 44; Be, width of hypothetical footing, 5.65 m. Substituting the corresponding values a safety value exceeding 2.0 is obtained. 5.2 Review for failure of embedment of cut-off wall. Determination was made of the safety factor against embedment of the cut-off wall upon completion of the bottom level of struts and the maximum depth of excavation has been reached; the expression used reads as follows:

Ea d

(13)

5.3 Capacity of post-tensioned anchors. The capacity of the anchors was determined from the following expression:

Qf 

Pi w le tg φ FS

(14)

where: Qf, frictional capacity of anchor, t; Pi, grouting pressure, 6.0 kg/cm2; w, perimeter of the cross sectional area (15 cm in diameter), 0.47 m; le, length of the anchor bulb, m; , angle of internal friction of soil where the anchor bulb will be embedded, 36 degrees; and FS, safety factor equal to 1.5.


90


5.4 Characteristics of anchors.

6 INSTRUMENTATION

Their diameter will be of 15 cm and their bore holes will be drilled with pneumatic equipment, having an inclination of 45 degrees below the horizontal; they are constituted by 0.6”-diameter strand cable. The annular void between the borehole walls and the strand is filled with a water-cement grout with proportion of 2:1 by weight with minimum strength of 150 kg/cm2, injected at a pressure of 6.0 kg/cm2. To quickly reach the strength of the grout a setting accelerator will be used and a volume stabilizer. Within a term of three days, when the grout has already reached the design strength, post-tensioning (Fp) specified in tables 2 and 3 was applied. The reaction system of the anchors was resolved with concrete blocks supporting the reaction and wedging plates. Table 2. Anchor system for panels of 2.5m, 4.0m and 5.5m. Row

6.1 Objective. The instrumentation program necessary to monitor the movements of adjacent buildings and of excavated area itself was defined as follows: 6.2 Bench mark. For purposes of determining with accuracy the ground movements to be experienced as a result of the building construction, it is convenient to install a control point outside the influence area of the works to be carried out so that when reference is made to these points the movements generated when building of the foundation can be calculated as differentials. Such bench mark shall be located at a distance of no less than 300 m from the job site. 6.3 Measurement program.

Passive length,

Bulb length,

Total length,

Fp,

No. of

m

m

m

t

strands

1

22.0

5.0

27.0

43.0

3

2

16.0

8.5

24.5

110.0

7

3

12.5

9.5

22.0

131.0

9

4

9.0

6.5

15.5

70.0

5

6.4 Superficial references and wall marks.

5

6.5

5.0

11.5

54.0

4

They are constituted by points fixed at ground surface and as reference points painted in neighboring structures; the former are installed by defining collimation lines parallel to the excavation axes that are monitored with a builder’s transit so as to detect the horizontal displacements that have occurred, whereas with an optical level and stadia rods the vertical displacements are measured.

Table 3. Anchor systems for panels of 3.4m, 6.0m and 7.0m. Row

Passive length,

Bulb length,

Total length,

Fp,

No. of

m

m

m

t

strands

1

22.0

5.0

27.0

51.0

3

2

16.0

9.5

25.5

132.0

9

3

12.5

11.5

24.0

157.0

10

4

9.0

6.5

15.5

84.5

6

5

6.5

5.0

11.5

64.5

4

Reading of the reference point shall be made once a week with results presented graphically to facilitate their interpretation.

6.5 Reference points in cracks. Gypsum marks were plastered to monitor the enlargement of cracks and fissures at those sites where they were detected by the structural inspection of neighboring buildings. On the other hand, a survey was made of cracks at sidewalks and street pavement to monitor their behavior.

N.P.T +4.50

Planta Comercial

N.P.T +0.00


N.P.T -4.71




N.P.T -7.85



N.P.T -11.00



N.P.T -14.14



N.P.T -17.28



N.P.T -20.42



N.P.T -23.57

Figure 3. Schematic cross section and a front view of the layout proposed for the anchor system.


3rd.


Session 2: Excavations

Technical Committee

TC-214

Technical Committee




A historic capitol and a deep excavation Un capitolio histórico y una excavación profunda Nasser MASSOUDI1 & Richard SLIWOSKI 2 1Bechtel 2Virginia

Power Corporation, Frederick, Maryland, USA Department of General Services, Richmond, Virginia, USA

ABSTRACT: The Virginia State Capitol was designed by Thomas Jefferson and constructed in 1785. The Capitol is of great historic and architectural significance to not only the Commonwealth of Virginia but to the nation. It is the oldest operating Capitol in the U.S. and required a total restoration/renovation as well as an expansion to continue to remain functional. Site conditions dictated that the expansion (a new visitor center) be placed underground on the south side of the Capitol building, requiring a 40-foot deep excavation approximately 5 feet from the Building. Because of its historic significance, every measure had to be taken to ensure the safety and function of the Capitol during construction. The construction techniques that were used to make the deep excavation and protect the historic Capitol included a tied back concrete slurry wall, jet grouting, compensation grouting, and instrumentation monitoring. This paper describes the history of the Capitol building, the techniques that were employed for construction of the deep excavation, results of the latest movement monitoring, and factors contributing to meeting the restrictive movement goals.

1 HISTORICAL BACKGROUND The Virginia Capitol in Richmond houses the oldest legislative body in the United States. It has played a significant role in Virginia’s history with the contributions of its many historical figures as part if its tapestry. It is designated a National Historic Landmark and is listed on the National Register of Historic Places. Virginia’s first Capital was in Jamestown and dates back to 1619. The State Capitol was relocated to Middle Plantation (Williamsburg) in 1699. It served until the American Revolutionary War. It was Governor Thomas Jefferson who urged that the Capitol be relocated to Richmond. In 1779, the Virginia legislature voted to move the Capitol from Williamsburg to Richmond. Plans soon began for a new building to serve a new state, the Commonwealth of Virginia. With the establishment of Richmond as the new capital, six squares of land were selected for the placement of permanent public buildings on Shockhoe Hill, a major hilltop overlooking the falls of the James River in Richmond. Thomas Jefferson designed the Virginia State Capitol, which is the middle structure of the present Capitol building, while serving as minister in France. Working with French draftsman Charles-Louis Clérisseau, Jefferson designed the building to

represent a dramatic departure from British influence. He modeled the building after the Maison Carrée at Nîmes in southern France, an ancient Roman temple. Jefferson set a precedent by using a temple form and the building is nationally significant as the first Classical Revival state capitol building constructed in America. He originally intended to erect three buildings, one for each of the three branches of government. However, his goal was beyond the Commonwealth's finances. On August 18, 1785 the cornerstone was laid for the Capitol building in Richmond. However, construction of the Capitol had begun without the plans. Jefferson sent the Clérisseau drawings and the plaster model created by Fouquet; they reached Richmond separately in 1786. The original model is still on display at the Capitol building and is shown in Figure 1. Once the building plans arrived, it was discovered that foundations had been laid out different than that in Jefferson’s plans. During the long period of construction, 1785 to 1798, Jefferson’s design was extensively altered by Samuel Dobie and other Richmond builders, including the foundations. Thomas Jefferson had expressed unhappiness with the changes in his design.


94

A historic capitol and a deep excavation

Figure 1. Original Capitol Building Model.

The late 1700s was just the beginning for the new state house as the location of many extraordinary moments in history. In December 1791, the Bill of Rights was approved making Virginia the 11th state to ratify the amendments; and in 1796 Jean-Antoine Houdon’s statue of George Washington was placed in the rotunda becoming one of Virginia’s most treasured artifacts. By 1857, the building was suffering from deferred maintenance and the effects of heavy use. Unfortunately, the cost of renovation was deemed too high and repairs were substituted. In 1858, a proposal to enlarge the Capitol building was submitted by Albert Lybrock. However, in 1861, the Virginia Convention voted to secede from the union. The American Civil War, in which Virginia played an important role, interrupted the project and Lybrock's proposals were never executed. In 1862, in the House of Delegates Chamber, Robert E. Lee was appointed as commander of the Army of Northern Virginia. The Capitol building would serve as the Capital of the Confederate States of America from 1861 to 1865. In April 1865, departing Confederate troops were ordered to burn the city’s warehouses and factories. The fire spread out of control and the Capitol building was one of few buildings that were spared, as shown in Figure 2.

African-American building contractor, offered a resolution to address the appearance and condition of the Capitol. During this period, there was a dispute over the leadership of the City of Richmond government. This led to a hearing on April 27, 1870 that was held in the large courtroom on the second floor of the Capitol building. Several hundred people crowded the room and balcony. The balcony was overloaded and the extra weight caused it to collapse and fall about 40 feet to the courtroom floor below. This soon came to be known as the “Capitol Disaster” which caused 62 deaths and injured 251 people. Despite calls for the building's demolition, the damage from the tragedy in 1870 was repaired. By the turn of the century, the Capitol was in a poor state of disrepair. Therefore, the General Assembly appropriated $250,000 for the Capitol renovation. After 16 months of work, the remodeling was completed, consisting of the addition of the wings to the east side of the Capitol building (House Chamber) and west side (Senate Chamber), including the addition of 24 granite steps on the South Portico of the building; the latter was in keeping with Jefferson’s original plan. This significant renovation from 1904 to 1906 was designed by noted Norfolk Architect John Kevan Peebles. Additional interior work ensued later, in 1937 and 1962. The latest renovation of the Capitol started in 2003 and is scheduled for completion in late 2006. This almost $100 million project includes the renovation and restoration of both the exterior and interior of the Capitol, and the construction of a 27,000 square foot extension below ground, connecting to the Capitol. Upon completion, the Capitol will have new mechanical and electrical systems; a new roof, elevators, stairwells, and legislative meeting space; the building will be fully accessible and ADA compliant; the legislative chambers and Rotunda will have been restored to their original color and décor of the time period; and there will be a new visitor entrance at 10th and Bank Streets with a new visitor center in the underground extension providing educational and museum exhibits. The exhibits are scheduled to be opened in 2007, at the time of Virginia’s 400th anniversary. 2 DEEP EXCAVATION

Figure 2. Capitol Building During the Civil War.

During the Reconstruction period, Virginia was under military rule. By 1870, the Capitol was overdue for repair and Delegate Ballard Edwards, a respected

From a geotechnical perspective, the most significant aspect of the recent construction is related to the installation of a 40-foot deep excavation as close as 5-foot from the Capitol building to house the new visitor center. This excavation becomes even more significant considering the requirements that it had to meet, including protecting the historic Capitol, which rests on soft deteriorated footings that are over 200 years old, and by allowing only very limited building movements of 0.25 inch. The condition of some of the footings is shown in Figure 3.


MASSOUDI N. et al.

Capitol

95

Building

Capitol Foundations

Slurry Wall

E x c a v a t i o n

A r e

a

Figure 3. Original Capitol Footings.

3 GEOLOGY Virginia Capitol is located on a hilltop. The site geology is characterized by sequences of marine and sedimentary deposits. The site soils consist of manmade fill, natural soils of various geologic ages, and bedrock. Typical subsurface stratigraphy at the site consists of approximately 2 feet of medium dense sandy fill, 20 feet of loose to very dense clayey sand with cobbles and boulders (average Standard Penetration Test N-value of 40 blows/foot), 70 feet of firm to hard clay and silt (average N-value of 15 blows/foot and average undrained strength of 2,000 pounds per square foot), 50 feet of very dense sand and gravel (SPT values typically greater than 100 blows/foot), and granite bedrock at a depth of about 140 feet below the ground surface with average unconfined compressive strength of 15,000 pounds per square inch. Perched groundwater level is about 20 feet below the surface. Only soils of upper 80 feet are of geotechnical significance for the project, consisting of clayey sand, and clay and silt, as these soils have the greatest influence upon the stability of the construction. 4 EXCAVATION SUPPORT CONSTRUCTION The main component of the excavation support system was a reinforced concrete slurry wall, with a total length of about 150 feet. It was 60 feet deep from the existing ground, with a thickness of 3.3 feet. It served both as a temporary excavation support and as the final structural wall for the new visitor center. The geometry of the wall in plan was unusual, dictated by architectural requirements. The slurry wall was tied to 4 transverse sections; these sections were designed as cantilevers. The slurry wall and its relation to the Capitol foundations are shown in Figure 4.

Figure 4. Slurry Wall in Plan.

Given very limiting movement requirements for the project, it was found necessary to improve the soils between the Capitol foundations and the slurry wall using jet grouting. The grouting was only in the upper 20 feet of the ground. More than 100 jet grout columns were installed. Each column was nominal 3 feet in diameter and was keyed into the underlying clay. Similarly, compensation grouting was deemed necessary as a precaution against excessive movements. Grouting was performed in the upper 20 feet of the site soils immediately under the Capitol foundations. The Capitol building was occupied during this work; therefore, all grouting had to be performed from outside the building, including grouting underneath the interior Capitol foundations. With the slurry wall and other grouting components in place, excavation in front of the slurry wall commenced, supported by 5 rows of temporary tiebacks one of which was drilled directly into the jet grout columns and the remainder through the slurry wall. The tiebacks consisted of three to six 7-wire strands. All tiebacks were post-grouted, pre-stressed, and tested. Tiebacks in the jet grout columns were stressed to 70 kips each, transmitting the prestressing load through steel walers to the jetgrouted mass. Tiebacks in the slurry wall were stressed to loads varying from 60 to 160 kips. As noted earlier, the tiebacks were only temporary; once the structure of the visitor center was completed all tiebacks were de-tensioned. 5 GEOTECHNICAL INSTRUMENTATION The extraordinarily limiting movement goal for the project of 0.25 inch required the availability of performance information on a continual basis to enable rapid response to developing trends. It was also a critical part of the verification and quality assurance. Hence, a real-time system of geotechnical instruments was used for monitoring the building and the construction. They consisted of three total station theodolites with about 80 optical prisms,


96

A historic capitol and a deep excavation

two in-place inclinometers (and one manual as backup), temperature sensors, and data acquisition hardware, software, computers, and Internet connection. The system proved invaluable in monitoring the performance of the work. At one time during the installation of the jet grout columns in November 2004, several Capitol footings began to settle unexpectedly. The availability of continuous and realtime results enabled a quick response in averting excessive foundation movements with undesirable consequences. These movements were later recovered. Equally, the real-time data was instrumental in controlling the grouting operation during the compensation grouting and tieback installation. Typical vertical movement monitoring results are shown in Figure 5. 0.25

Target OWT0111

Movement (inch)

0.2 0.15 0.1

Jet grouting period

0.05 0 -0.05 -0.1

excavation support, ground improvement using jet grouting, foundation improvement using compensation grouting, and a very restrictive movement tolerance of 0.25-inch for the Capitol. With the deep excavation complete and construction of the new underground visitor center underway, the final movements are well within the goal of 0.25 inch. The maximum movement recorded during the entire construction period was 0.17 inch of settlement which occurred during the jet grouting operation. This settlement, however, has since been recovered, with final movements largely in the 0.1-inch range. The established movement tolerances for the project were extraordinarily limiting compared to typical excavations, yet fitting and consistent with protecting deteriorating foundations and historic architecture. The achieved movement results are also extraordinary for the given excavation depth, perhaps including an element of good fortune. The success in controlling movements is credited to detailed design, close communication among the project team, proven construction practices, and a real-time monitoring system that permitted the evaluation of construction performance at close intervals and implementation of corrective measures within very short time periods.

-0.15 -0.2 -0.25 Sept Dec Mar Jun Sept Dec Mar Sep-04 Nov-04 Feb-05 May-05 Aug-05 Nov-05 Feb-06 04 04 05 05 05 05 06

Date

Figure 5. Typical Movement Monitoring Results

Results of movement monitoring for over 18 months of construction are shown in Figure 5. With structural components for the new visitor center completely in place, the Capitol building movements to date are well within the goal of 0.25-inch. The maximum recorded movement was 0.17 inch of settlement as observed during the jet grouting period. However, this settlement was later recovered, with final movements largely in the 0.1-inch range. 6 CONCLUSIONS The Virginia State Capitol was designed by Thomas Jefferson and is the oldest operating Capitol in the United States. With great historic and architectural significance, it is recognized as the model for public buildings throughout the United States. Having survived the American Civil War and many years of deterioration, it is undergoing a much needed restoration as well as an expansion to continue to remain functional. The construction of a 40-foot deep excavation exposed the Capitol to risks of excessive movements beyond that considered tolerable. Therefore, redundant construction measures for risk mitigation were implemented for meeting the challenge. These included a concrete slurry wall for SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

Technical Committee




The support of a 25 m deep excavation in difficult ground conditions using Single Bore Multiple Anchor technology Soporte de una excavación de 25 m de profundidad en condiciones de terreno difícil usando tecnología de anclaje múltiple con barreno único Devon MOTHERSILLE1, and Bora OKUMUSOGLU 2 1Single

Bore Multiple Anchor Ltd, London, UK 2Kasktas AS, Moscow, Russia

ABSTRACT: The paper describes the design, construction and testing of some 3600 temporary single bore multiple anchors (SBMAs) used to support the deep basement which forms part of the foundation for the Kuntsevo Plaza; a mixed-use development in the Kuntsevo district of south-western Moscow. SBMAs were used to provide support for 40m (131ft) deep diaphragm walls, constructed to retain a 25m (82ft) deep excavation in the challenging Moscow mixed soils, comprising combinations of low strength clays, sands and silts. Previous attempts to sustain the required loads of up 600kN (135kips) in the anchors had failed due to unacceptable creep. However, an understanding of the concept of progressive debonding, and the use of this knowledge in the design of efficient fixed anchor lengths, in the SBMAs, proved highly effective. In addition, the introduction of fixed anchor enhancement techniques such as post-grouting resulted in the achievement of anchor capacities more than double those previously achieved in the prevailing ground conditions and proved effective in limiting wall displacements to a maximum of 7.5mm. This project was also the genesis of intuitive and innovative cloud-based software, specifically developed for tablets, to analyze and manage the vast volumes of data produced from the stressing and testing of the ground anchors.

1 INTRODUCTION Ever since the first development plan in the 16th Century, the city of Moscow has drafted and implemented several development strategies over the centuries which have contributed to a city steeped in history and culture. More recently, in a bid to attract new investors to the city, an urban development plan was drafted that will take Moscow up to the year 2020. Kuntsevo Plaza, completed in late 2014, forms part of this strategy. The Plaza, located in a south-western district of Moscow (Figure 1), is described as a vibrant mix-use lifestyle centre rooted in urbanity and comprising five levels of amenities above the surface including offices, residential, retail, restaurants and leisure facilities. An essential feature of the new construction and the main focus of this paper is the foundation structure which supports a 25m deep excavation. The excavation accommodates the vast basement of the complex where features include extensive car parks, utilities and a substantial cinema complex. In order to facilitate construction of the basement a diaphragm wall of 45m total depth, 0.8m thick and approximately 600m overall perimeter was constructed.

Figure 1. Location plan showing Kuntsevo Plaza in a south-western district of Moscow.

2 GROUND CONDITIONS Two site investigations were carried out to establish the nature, and more importantly, the engineering properties of the ground. The extent of the site together with trial anchor, borehole and CPT locations, which are referred to later in this paper, are shown in Figure 2. Boreholes were generally driven to depths of up to 50m and some 605 samples were extracted and tested, and these included samples used to establish the aggressivity of the ground and the ground water.


98

The support of a 25m deep excavation in difficult ground conditions using Single Bore Multiple Anchor technology

ultimate bond stresses at the ground/grout interface is particularly useful from a fixed anchor design perspective but it is emphasised that the magnitude of this parameter is also a function of the contractor’s construction methodology. 3 THE APPLICABILITY OF THE SBMA CONCEPT

Figure 2. Simplified site plan showing trial anchor, borehole and CPT test areas.

The final design for the anchored structure utilized ground anchors installed across six levels. In the vicinity of the fixed anchors, associated with the first four levels, mixed soils comprising clay interbedded with sand, glacial sandy silt with lenses of gravel and fluvio-glacial deposits of water-saturated fine to medium sands were encountered. In the lower two levels, the fixed anchor zones were founded in Lower Cretaceous deposits of up to 20m thick water bearing sands with clay pockets. These sandy strata were underlain by impermeable Upper Jurassic clay deposits, in which the diaphragm wall is socketed. The natural ground elevation varied between 170.00 to 174.00m above sea level and the groundwater elevations varied between 164.72 and 167.10m, with two additional artesian aquifers located in the fluvioglacial and Lower Cretaceous sands. The geotechnical parameters established from laboratory tests included quantification of effective angle of shearing resistance, drained Young’s Modulus, effective cohesion, and unit weight of soil. These were relevant to varying degrees for the design and the computer modelling of the anchored structure, and are listed later in this paper in Table 2. It was established that the specialist anchor contractor, Kasktas A S, had undertaken a series of field trials on tremie grouted conventional anchors, which comprised anchors with a single 8m bond length, installed vertically into the two distinct founding layers. Borehole diameters of 150mm and 178mm were used and drilled to depths of between 18m and 27m. During these trials, the ultimate load was defined as the maximum load attained by the anchor before continuous upward displacement of the fixed anchor was observed. Test data confirmed that this generally occurred at 450kN. Back analysis of the failure loads generated average ultimate bond stresses at the ground/grout interface of 175 kPa and 195 kPa in the upper clay and in the lower sand layers, respectively. The establishment of in situ

Recognizing that conventional straight shafted grouted anchors failed to achieve the desired working loads in the prevailing ground conditions, alternative technology had to be implemented. With this background SBMA Ltd were approached by the specialist anchor contractor to design ground anchors that could fulfil the design requirements. It has been acknowledged by numerous researchers for over 60 years that when tensile load is applied to a steel tendon in grout, whether founded in rock or soils, the load distribution within the fixed anchor length is non-uniform. Ostermayer (1974 and 1977) and subsequent work by Barley (1995) into the performance of anchors founded in clays, sands and gravels highlighted the non-uniform distribution of bond stress and the progression of load concentration along the length of long fixed anchors. As a consequence the British Standard code of practice for Ground Anchorages (BS8081:1989) recommends a limit of 10m for the fixed anchor length. Barley (1995) established a method of evaluating the “efficiency” of an anchor in mobilizing ground strength; short fixed lengths being highly efficient in mobilizing ground strength and long fixed length being grossly inefficient (i.e. a 10m fixed anchor utilizes only 45% of the average bond stress exhibited in a 2.5m fixed anchor). Furthermore, and most importantly, the utilization of a multiple of short and efficient unit lengths within a single borehole has allowed the working capacity of soil anchors to be more than doubled. Working loads in the range 600 kN to 3800 kN have been safely achieved in ground conditions ranging from soils to highly weathered rocks in various parts of the world and this methodology was implemented in the design of the Kuntsevo ground anchors. 4 SBMA - DESIGN, FABRICATION AND CONSTRUCTION The SBMAs were designed in accordance with procedures documented in Ostermeyer and Barley (2003) which incorporates an efficiency factor to account of the non-linear distribution of bond stress, due to progressive debonding, that exists in the grouted tendon under increasing tensile load. A simple mathematical expression relating the ultimate bond stress, ult, to the length of the fixed anchor was applied;


MOTHERSILLE D. et al.

ult

L x feff

where feff is an efficiency factor, which itself is a function of L in the form; 1.6L-0.57 and L is fixed anchor length (m). The tendons, incorporating 7- wire low relaxation strand (Characteristic tensile strength = 261 kN) and greased HDPE sheathing for decoupling the free length, were fabricated on site ensuring that dimensions and preparation of materials were carried out in accordance with the relevant standards for temporary anchors. The borehole was formed through a steel-reinforced concrete stressing block (2.0m x 2.0m x 0.5m deep) for investigation tests, and for production anchors through reservation pipes placed within the reinforcement of the diaphragm wall. Boreholes were drilled with 145mm diameter auger drill with end-of-casing air flush using a diesel/hydraulic rig (Casagrande Type C8). In order to preserve the integrity of the borehole a steel casing (180mm o/d and 160mm i/d) was advanced to the full depth of hole. On completion of the borehole the complete tendon, comprising the three unit anchors [top (A), middle (B), bottom (C)], the primary grouting pipe and the post grouting pipe were installed through the casing (Figure 3).

99

applications) the document providing guidance on ground anchor practice in Russia. In relation to investigation tests, EN1537:1999 (recently revised to EN1537:2013) states ‘Investigation tests may be required to establish for the designer, in advance of the installation of the working ground anchors, the ultimate load resistance in relation to the ground conditions and materials used, to prove the competence of the contractor and/or to prove a new type of ground anchor by inducing a failure at the ground/grout interface.’ The investigation tests were carried out in phases across three separate locations on the site (Figure 2). The three trial areas were selected based on availability of suitable areas and logistics imposed by ongoing construction activities. Mothersille et al (2012) describe the first phase of the investigation tests which was carried out in trial area 2. The main objective of each testing phase was to assess the performance of the SBMAs using different grouting techniques in different locations on the site. In addition the tests provided the contractor an opportunity to become competent in the fabrication and execution of SBMAs which had never been used before in Russia. A series of fifteen test holes were proposed to accommodate different types of test, the details of which are summarized in Table 1 and reference can also be made to Figure 2 which shows the location of the test areas. Table 1. Summary of trial anchor details.

Figure 3. Schematic showing SBMA with three unit bond lengths (the primary grout pipe is omitted for clarity).

5 INVESTIGATION TESTS AND TRIAL ANCHOR PROGRAMME Bearing in mind the complexity of the ground, the importance of the project, the high consequences of failure and the fact that a new anchoring technology was introduced to the specialist contractor, it was considered prudent to carry out an extensive programme of investigation tests. The specification for the works referred to both EN1537:1999 (Execution of special geotechnical work – ground anchors), BS8081:1989 (British Standard Code of practice for Ground Anchorages) and VSN 506-88 (Design and installation of ground anchor

Trial anchor No

Test Area/ Phase

1

1/I

Description

Post-grouting test. 25m borehole inclined at 15° 2 1/I 29.5m SBMA, inclined at 15° using KECG 3 1/I 25m SBMA, inclined at 15° using TAM 4 1/I 15m SBMA, inclined at 15° using TAM 5 1/I 25m SBMA, inclined at 15° using CTG 6 2/II Post-grouting test. 25m borehole inclined at 15° 7 2/II 29.5m SBMA, inclined at 15° using KECG 8 2/II 25m long SBMA, inclined at 15° using TAM 9 2/II 15m long SBMA, inclined at 15° using TAM 10 2/II 25m long SBMA, inclined at 15° using CTG 11 3/III Post-grouting test. 25m borehole inclined at 15° 12 3/III 29.5m SBMA, inclined at 15° using KECG 13 3/III 25m SBMA, inclined at 15° using TAM 14 3/III 15m SBMA, inclined at 15° using TAM 15 3/III 25m SBMA, inclined at 15° using CTG *KECG denotes Kasktas end-of-casing grouting **TAM denotes tube-a-manchette ***CTG denotes conventional tremie grouting


100


5.1 Anchor grouting strategies Post-grouting of anchors is a well-established technique and known to significantly enhance the capacity of anchors given the appropriate ground conditions. Ostemeyer and Barley (2003) note that an enhancement factor of two can be reasonably achieved on post-grouted SBMAs installed in clay soils. However, the challenge confronting many contractors is deciding what post-grouting parameters should be applied for a particular type of ground. Often they have to rely on experience or trial and error techniques in order to generate the enhancement in capacity desired. To gain an understanding of the ground response to post-grouting and produce a rational basis for the execution of post-grouting in the production anchor works, a series of three post grouting trials were included in the investigation test programme. These were designed with the objective of establishing a range of break-out pressures for the tube-a manchette (TAM) valves, refusal pressures, volume of grout take and flow rates. Such parameters are a necessary prerequisite to establishing an effective post-grouting regime in the production anchors. The post grouting trial involved the installation of 25m long TAMs, eccentrically placed in the bore hole to simulate the position they would adopt were an anchor tendon placed in with them. The TAM installation comprised approximately 10m of perforated tubing, with ports spaced at 300mm centres (representing the unit anchor locations), and 15m of riser pipes, connecting the perforated tubes to the surface. The installation was oriented subvertically at an angle of 15° to the vertical axis. The trial work established a methodology which involved the use of 50mm diameter TAMs and an injection of a target volume of grout of 20 litres per port (spaced at 300mm c/c) with water/cement (w/c) ratio of 0.6 and an associated injection pressure of 10 bar. This was the starting point and crucial grouting parameters were monitored using a computerized Atlas Copco Logac G5 recorder system which provided data on pressure, flow rate and volume. The data derived from these tests assisted the contractor with the planning of the grouting strategies for the production anchor works. When sandy soils or substantially non-cohesive strata were encountered end-of-casing grouting techniques were employed. This method typically involves driving casing through to the end of the borehole and then applying a controlled pressured grout (typically 10-20 bars) through the casing as it is incrementally withdrawn and has proven effective in creating enhancement of bond capacity by a factor of 2 (Ostermeyer and Barley, 2003). Conventional tremie grouting was also employed to compare the behaviour of the test anchor incorporating other grouting techniques. The primary

grout retained a w/c of 0.45 for both end-of-casing grouting and tremie grouting. 5.2 Stressing and testing The jack arrangement for a three-unit SBMA includes three hydraulic rams that are synchronized by coupling to the same hydraulic powerpack, so that the same load is applied simultaneously to each unit anchor. The stressing and testing arrangement is shown in Figure 4. The ram extensions on the stressing jack were recorded using a digital vernier caliper for each stage of the cyclic loading and again during creep testing in accordance with BS8081. Measurements were corrected for reaction base-pad movement measured by dial gauges mounted on an independent reference beam.

Figure 4. Typical set for the stressing and testing of production SBMAs at Kuntsevo showing three hydraulically synchronised stressing jacks.

5.3 Observations and salient points derived from investigation tests 5.3.1

Tests loads adopted for trial anchors

At Kuntsevo the geometrical configuration of the trial anchors was identical to that proposed for the production anchors but for practical reasons the anchor had to be installed sub-vertically as opposed sub-horizontally. Furthermore, although production anchors are classified as temporary anchors (which are typically associated with a proof load factor of 1.25 and a maximum proof load of 665 kN) it was suggested at the outset, by Kasktas, that the anchors should be tested to 900 kN. Knowing that the maximum design working load was 532 kN this would mean imposing an onerous proof load factor of 1.7 on the anchors. The decision by Kasktas to use a higher proof load factor was partly due to the acute concern over minimizing settlements induced behind the wall and the effects of these on adjacent buried



services and this formed the basis of a conservative approach in the design. The trial anchors were therefore provided with an increased tendon capacity to allow them to be tensioned to a maximum test load of 1250kN without exceeding 80% tendon capacity. In order to assess the performance of the trial tests, calculations were made to derive the maximum permissible working loads for the founding stratum encountered. The maximum permissible working loads were derived using a limiting magnitude of creep displacement (2mm) in accordance with VSN 506-88 and Appendix M.10 of BS8081:1989. 5.3.2

Trial anchors

Trial anchor Nos. 2, 7 and 12 were designed in order to derive data from the stratum that would accommodate the bonded length for the lower rows of production anchors. Bearing in mind the nature of the ground encountered (described as sands) these trial anchors were grouted using end-of-casing grouting techniques. The end-casing grouting methods employed were not conventional in that the equipment was not available to allow pressure to be applied through 300mm lifts as the casing is rotated and withdrawn. Instead, Kasktas applied pressure through the casing each 1.5m length was withdrawn. This created significant pressure dissipation to the surrounding ground. Furthermore, the grout records confirmed relatively low pressures (average 2.5 bar for w/c = 0.45 grout) which suggests limited enhancement for a given volume of grout injection. This was reflected in the load/extension results which show relatively large permanent displacements (circa 100mm) in the upper units and lower permissible working loads. Trial anchor Nos. 3, 8 and 13 comprised 25m long, three unit SBMAs with an integrated TAM. The TAM was included here to provide a comparison on performance with other methods and to allow the operatives an opportunity to be familiar with the operation of a double inflatable packer system within the TAM. Based on previous experience it was judged that this system would provide a reliable method of controlling post-grouting episodes in the various founding layers. Trial anchor Nos. 4, 9 and 14 were designed to assess the performance in the shallower anchors (generally the upper two rows). An overall assessment of the tests was that they performed satisfactorily and based on the results that were derived from the tests an average maximum working load of 703 kN was assessed from back-analysis. The maximum required working load for these rows of anchors is 412 kN which was adequately exceeded. Trial anchor Nos. 5, 10 and 15 were grouted via a conventional tremie pipe with primary grout of w/c = 0.45, and no enhancement strategy employed. The intention here was to provide a yard stick to compare

101

the performance of the other grouting techniques employed. The stressing data derived for these trial anchors appeared more erratic than those acquired from the other tests. Larger permanent displacements were observed together with nonreproducible cyclical profiles in the load/extension curves. The fact that these anchors were grouted using conventional tremie pipe with no facility for enhancing capacity was clearly reflected in the relatively poor performance observed. Analysis of the grouting records for these trial anchors confirmed a positive effect of post-grouting. This was particularly apparent in the behaviour of anchor No.13 where the permanent displacement was reduced (as illustrated in Figure 5) in the lower units which received 30 litres of grout in the even numbered ports and 20 litres in the odd numbered ports. The target volumes were selected based on previous experience but based on these results it could be argued that the target volume should be increased. It is noteworthy that Bruce et al (2004) adopted grout volumes of 55 litres per port and associated target pressures of 800 psi (55 bars) for anchors installed in clays. Back-analysis of the derived data suggests maximum working loads ranging from 580-780 kN (with the exception of the trial anchor No. 15 which only had the conventional tremie grouting) exceeding the maximum working load of 532 kN demanded by the anchored structure design. 5.3.3 Summary and considerations for production anchor works The permissible working load of SBMAs even without pressure grouting, were found to be greater than the ultimate capacity (bond failure) of conventional anchors. The use of SBMAs increased the effective load capacity for the complex soil conditions, while pressure grouting improved the load displacement performance further. From an overall assessment of the results it was concluded that the use of the TAM was a necessary general requirement for all SBMAs installed on the site based on the following; • Contemporaneous geotechnical data indicated a significant cohesive component within the soil matrix. This was not obvious from the original soil descriptions but was confirmed when the spoil heaps were examined after drilling. It is commonly accepted that post-grouting is particularly effective in cohesive soils. • In the event that a production anchor exhibits excessive creep during testing, the existence of the TAM provides a means of revisiting the anchor to introduce further episodes of grout prior to re-testing. • The load/extension data for the trial anchors, grouted via tremie pipe, confirmed this method to be inappropriate for the conditions encountered.


102


• Reliance could not be placed on the effectiveness of the end-of-casing grouting methods implemented by the specialist anchor contractor. Based on the observation from the investigation tests, the following recommendations were made for the production anchor works: • The SBMAs to comprise 3 x 3m long unit anchors each with two strands which were locally noded to enhance the bond resistance at the grout/tendon interface • The use of TAMs was recommended for SBMAs installed in the mixed soils encountered at Kuntsevo Plaza. • The target grout volume should be increased to 50 litres per port with target pressure of 55 bars. • The location of the TAM should be carefully assessed and efforts should be made to maintain the TAM, by flushing with water, for subsequent regrouting if necessary 6 PRODUCTION ANCHOR SUPPORT SYSTEM AND LOCATION OF ADJACENT SERVICES Figure 6 shows a section through the ground support system for the 25m deep basement excavation, which comprised a diaphragm wall, socketed in an impervious clay layer, and 6 levels of temporary SBMAs with up to 600 kN working load. The footprint for the development demanded that the diaphragm wall be located within 4m of a reinforced concrete sewage pipe. The top 5m of the wall was designed to cantilever due to the close proximity of adjacent services as shown in the Figure 6. The services are defined as follows; i, r/c sewer pipe (2000mm diameter); ii and iii, steel gas mains (420mm and 300mm diameter); iv and vi, r/c storm water drains (400mm and 600mm diameter); v, steel water main (600mm diameter) and vii, r/c storm water culvert (3400mm x 3600mm). The primary function of the support system was to minimize the adverse effects of the deep excavation on the buried services, many of which were already vulnerable due to their age. During the design phase of the deep excavation the serviceability displacement limits for these services, especially the sewerage line, which was closest to the wall, influenced the inclination of the ground anchors, which were installed at 1m centres. In addition, the anchors were prestressed up to 600 kN to effectively restrict wall displacements, thereby reducing settlements behind the wall. So sensitive was the issue of settlements that the local authorities insisted that the main contractor consider the relocation or costly replacement of the sewage pipe if assurances could not be given that settlement would be within specified maximum settlement tolerance of 30mm.

Figure 5. Illustration of the positive effect of post-grouting (TAM) in contrast to conventional tremie grouting (CTG).

Figure 6. Typical retained section with CPT profile.



7 PRODUCTION ANCHOR PERFORMANCE The three unit SBMAs were stressed and tested using three hydraulically synchronized jacks. The setup duplicated that used for the trial anchors except for the fact that a more portable stressing stool arrangement was adopted as shown in Figure 4. The specification required that 1 in 10 of all production anchors be subjected to suitability tests (equivalent to performance tests in PTI DC35.1-14) and the remainder to acceptance tests (equivalent to proof tests in PTI DC35.1-14). The first cycle was carried out with nominal one minute reading at each load level, the second and the third cycles were carried out with 15 minute observation periods at proof load level for monitoring the displacement-time. At the proof load, creep displacements were generally observed to be nominal and well within 5% of elastic displacement of tendon as required by the standard. Service load monitoring of the anchors was carried out by undertaking lift-off checks at the lock-off load (approximately 110% working load) in accordance with criteria stipulated in BS8081. On the rare occasions that anchor did not satisfy acceptance criterion, additional post-grouting was employed and the anchor retested until satisfactory creep was observed or a stabilizing trend was established. Checks were also carried out on the apparent tendon free length to ensure that the seat of load transfer was within acceptable limits stipulated in BS8081:1989. There were occasions when the apparent tendon free length fell marginally short of the 90% design free length criterion (the lower limit) specified. The reasons for this were related to friction developed in the free length and under these circumstances the anchor was subjected to two extra cycles of loading to gauge repeatability. Although this methodology was justified and recommended in BS8081, concerns were expressed about the definition of what constituted a repeatable loadextension curve. BS8081 does not provide guidance on this so reference was made to the Australian standard (RTA QA Specification B114:2007, clause 9.6.2). In accordance with this standard repeatability is investigated by checking that extensions in the third cycle are within ± 5% of those in second cycle. One of the major challenges was attempting to execute the works in freezing climatic conditions which affected all aspects of the ground anchor execution. When temperatures dropped below zero down to -15°C, special heaters were deployed so that operatives could record load-extension data with some level of comfort. During construction heated water tanks were deployed and insulation jackets used on grout pipes. Below -15°C the site operations related to ground anchoring were abandoned. The inclement weather accounted for a total of two weeks delay in the ground anchor construction programme which was 7 months duration.

103

8 COMPUTER MODELLING OF ANCHORED STRUCTURE The computer modelling was carried out using a hardening soil model and a staged approach where the excavation was modelled to each anchor installation level with pre-stressing of anchors at that level before the excavation progressed to the next level. The initial modelling of deep excavation, with two dimensional finite element analysis using Plaxis and soil parameters summarised in Table 2, indicated a maximum lateral wall movement of 85mm. A comparison with actual field measurements using inclinometers showed that the actual lateral movements of the wall to be significantly less than this initial prediction. Further investigation into modelling and site conditions revealed that the dewatering system, designed to operate throughout the excavation period, effectively reduced the pore water down to the final excavation level. The system was employed around the whole perimeter of the diaphragm wall. This reappraisal of the ground water regime was in contrast to the initial assumption by the design team that pumps lowered the water level step by step to within one meter below of each anchoring platform level. Reconfiguring the model with the actual hydrological profile in the vicinity of the wall resulted in an increase of effective stresses in the soil in front of the wall which in turn increased the passive earth pressure and thereby reduced the predicted wall movement down to a maximum of 25mm. As shown in Figure 7, this revised analysis provides a closer approximation to the actual behaviour of the anchored structure. Table 2 Summary of soil parameters Soil classification Sandy loam Clay with layers of sand Clay with gravel and lenses of sands Sandy loam Dense sand Loamy sand Fine sand Impervious clay

 : kN/m3 20.5

E’ : MPa 20

’ : deg 17

c’ : kPa 50

19.7

18

16

52

21.0

38

19

82

21.2 20.5 20.2 21.1

29 48 14 45

20 37 27 38

75 3 20 6

20.2

24

24

56

Note:  unit weight, E’ = drained Young’s Modulus,’ = effective angle of shearing resistance, c’ = effective cohesion


104


Figure 7. Final lateral wall movement vs Plaxis predictions.

Figure 8. Development of cloud-based anchor analysis and testing software for tablet.

9 DATA ANALYSIS AND THE DEVELOPMENT OF ANCHORTEST

10 CONCLUSIONS AND CLOSING REMARKS

The analysis and management of the data produced from investigation tests and the mandatory testing of thousands of anchors proved challenging. The project specification stipulated that no anchors could be accepted into the works until the results from acceptance tests were examined and signed off by the client’s representatives and this process often created delays. These issues paved the way for the development of intuitive and innovative cloud-based software, designed specifically for tablets, in order to cope with the rigors of a field operation (AnchorTest Ltd, 2015). At Kuntsevo, the protocols for the delivery of data had already been established during the development phases of the software package, so although the final product was not used on the project, the concept design and preliminary trial versions were optimized as a direct result of the works carried out at Kuntsevo (Figure 8). The completed product is now fully enabled and available for use on all types of ground anchor applications to permit paperless real-time analysis, data management, GPS technology and incorporates the acceptance criteria for major international ground anchor codes of practice (including BS8081:1989 and PTIDC35.1-14). This innovation provides a more reliable and convenient way to process data, both in the field and in the office, and effectively moves anchor testing beyond the limits imposed by data input using spreadsheets on laptops and PCs.

The construction of the deep basement structure for the Kuntsevo Plaza development saw the first use of SBMA technology in Russia. Under difficult geotechnical and environmental conditions some 3600 anchors were successfully installed stressed and tested. The success of this project was largely based on the decision to undertake an extensive programme of investigation tests which included the use of post-grouting trials. These trials provided an opportunity for the contractor to become proficient with the use of the ground anchor technology and generated useful parameters to permit effective execution of the post-grouting operations when they were employed. Despite the restrictions imposed by the close proximity of the anchored structure to sensitive services, a well-executed wall design restricted lateral wall displacements to a maximum of 7.5mm during the anchor installation phase. The detailed evaluation of the construction process and the modelling of soil properties incorporating the effects of dewatering and the consequential increase in effective stresses and passive resistance ensured more credible predictions of lateral displacement using Plaxis. The most important benefit of the limited wall movements was the significant reduction in settlement or disturbance to the adjacent services. The conservatism in the anchored wall design generated considerable savings by avoiding top down construction, proposed compensation grouting and proposed realignment of an existing sewage pipe.



The project provided the environment for the creation of an innovative cloud-based data management and analysis tool for anchor testing which should prove highly beneficial for the industry. The successful execution of the SBMAs at Kuntsevo has set a valuable precedent and paved the way for further construction of deep basements in soils that historically were only able to sustain relative low working loads (Figure 9).

105

Drilling and American Society of Civil Engineers, Geotechnical Special Publication, 124, Orlando, FL, USA, 361–373. Mothersille DKV, Orkun AO and Okumusoglu B (2012). “The performance of Single Bore Multiple Anchor trials installed in mixed Moscow soils”, Proceedings of the Road Construction Conference, Penza, Russia, 461–470. Ostermayer H (1974). “Construction, Carrying Behaviour and Creep Characteristics of Ground Anchors”, Conference on Diaphragm Walls and Anchorages, London, UK, 141–151. Ostermayer H (1977). “Detailed design of anchorages”, Review of Diaphragm Walls, I.C.E., London, UK, 55–61. Ostermayer H and Barley AD (2003). “Fixed anchor design - Ground Anchors”, Geotechnical Engineering Handbook, Vol. 2, Pub Ernst and Sohn, Berlin, Germany, 189–205. VSN 506-88 (1989). “Design of Arrangement and Installation of Ground Anchors”, USSR Ministry of Assembly and Special Constructions, USSR Ministry of Assembly and Special Constructions, Moscow, Russia.

Figure 9. Completed wall showing 25m deep excavation and six levels of SBMAs providing effective support.

11 ACKNOWLEDGEMENTS The authors wish to thank the site staff at ENKA and Kasktas for their efforts in executing the works during difficult climatic conditions. The detailed Plaxis analysis was carried out by Mr Şenol Adatepe from Kasktas A.Ş, Turkey and detailed reports on the anchor tests were prepared by Mr Ali Orkun Bayur during his time with Kasktas A.Ş, Turkey, and their contributions are acknowledged with thanks. Special thanks are also due to Dr Rasin Duzceer and Mr Alp Gokalp of Kasktas A.Ş, Turkey for their valued contribution to this work and the success of the project. 12 REFERENCES AnchorTest Ltd (2015). “AnchorTest for iPad Software http://www.anchortest.info, London, UK Barley AD (1995). “Theory and Practice of the Single Bore Multiple Anchor System”, International Symposium on Anchors in Theory and Practice, Salzburg, Austria, 293-301 BS8081:1989. “British Standard Code of practice for Ground Anchorages”, BSI, London, UK. BS EN1537:1999. “Execution of special geotechnical work. Ground anchors”, BSI, London, UK. BS EN1537:2013. “Execution of special geotechnical work. Ground anchors”, BSI, London, UK. Bruce ME, Traylor RP, Barley AD, Bruce DA and Gomez J (2004). “Post grouted single bore multiple anchors at Hodenpyl Dam, Michigan”, ADSC: International Association of Foundation SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

Technical Committee




The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas El uso de muros MSE rellenados con concreto celular ligero en áreas sísmicas de terrenos blandos Daniel PRADEL 1&2 and Binod TIWARI3 1Adjunct

Associate Professor, University of California Los Angeles (UCLA), Los Angeles, California ([email protected]) & 2Vice-President, Shannon & Wilson, 664 W. Broadway, Glendale, California ([email protected]) 3Professor, California State University Fullerton (CSUF), Fullerton California ([email protected])

ABSTRACT: A series of numerical analyses were performed on a Mechanically Stabilized Earth (MSE) wall that used Lightweight Cellular Concrete (LCC) instead of soil as infill. These analyses were performed using the geometry and input ground motions for a wall recently built for the Silicon Valley Rapid Transit (SVRT) system near San Francisco, California. For our analyses, the LCC-MSE wall was significantly weakened in our numerical models by using shortened geogrid lengths, and lower material strengths than the constructed wall. In spite of the weakened nature of the wall analyzed herein, seismic failure of the LCC materials and supporting ground was not predicted. Our analyses show that well designed LCCMSE walls tend to move dynamically in a quasi-rigid fashion, i.e., that they move mainly laterally and do not exhibit major rocking or seismic settlements. Additionally, because of their broad base of MSE walls, these walls distribute compressive and shear stresses to the underlying ground in a relatively even manner. Our numerical analyses also show that internal reinforcement of LCC-MSE walls is important to restrain side panels during earthquakes, but that the inertial loads from the panels are quickly transferred to the LCC. Hence, that long or continuous reinforcements are not needed for seismic stability. In summary, our analyses show that LCC is an excellent material for MSE walls and that the lightening of vertical loads that LCC provides has distinct seismic advantages in soft ground seismic areas, e.g., the elimination of ground improvement.

1 INTRODUCTION Soft-ground construction poses significant geotechnical challenges, ranging from large consolidation settlements (below the structure and in nearby developments), construction staging and extended project schedules. In seismic regions, soft ground conditions often result in the significant amplification of structural demands. For freeway and railroad embankments, such demands often result in costly ground improvement to mitigate the significant consolidation settlements resulting from the heavy weight of Mechanically Stabilized Earth (MSE) walls. Recently, a novel approach for the construction of MSE walls has been used, which involves replacing the MSE’s soil infill with Lightweight Cellular Concrete (LCC). The main advantage of LCC is its low unit weight (often about half the unit weight of water). In California, examples of LCC-MSE walls include the Colton Crossing for the Union Pacific-BNSF railroad in Colton (Teig and Anderson, 2012), the San Bruno Railroad Grade Separation in San Bruno, and the Silicon Valley Berryessa Extension in San Jose which will connect the Silicon Valley Rapid Transit (SVRT) system to San Francisco’s Bay Area Rapid Transit (BART) system.

In addition to railroad projects, LCC-MSE walls have also been used and/or are being considered for road transportation projects such as the 22/405 freeway interchange in Orange County, California, as well as for various new bridge approaches.

Figure 1. Lightweight Cellular Concrete.


108

The use of MSE walls backfilled with Lightweight Cellular Concrete in soft ground seismic areas

Figure 2. Construction of an LCC- MSE wall (Cell-Crete, 2015).

One of the main advantages of LCC is a reduction of up to 75% in the static weight of a traditional MSE walls. This significant weight reduction, results in much lower consolidation settlements, as well as reduced inertial loads and dynamic compression during earthquakes. The excavation of relatively heavy on-site soils for the wall foundations, combined with the low unit weight of LCC (compared to soil) can result in a balanced design (with zero added bearing pressures) or a small net increase in the foundation soils’ vertical stresses. This has allowed designers to either completely eliminate the need for improving the soft subgrade (e.g., eliminating stone columns) under certain MSE walls or at least to significantly reduce the scope of ground improvement (e.g., limited vibroreplacement in Teig and Anderson, 2012). An advantage of using LCC as fill, is that the relatively high strength of this material (compared to conventional MSE granular fill) also results in essentially no lateral “earth” pressures on the MSE panels and abutment walls. The weight advantages of LCC are shared by other materials, such as Geofoam. However, Geofoam is combustible and reacts chemically with hydrocarbons such as diesel fuel. Hence, railroads have been reluctant to accept Geofoam structures in California, but have often accepted the use of LCC.

low thermal conductivities, and high strengths compared to that of a conventional MSE soil infill. Its nature and behavior is often described as similar to that of porous soft rock. In California, the design engineer generally specifies a minimum compressive strength for the LCC material that the supplier (or manufacturer) must deliver during construction. It is not unusual to have two or more minimum strengths specified by the design engineer for the same wall, e.g., a higher strength near the foundation soils and lower strengths in the upper portions of the wall. In our experience, densities of LCC used for MSE walls in California typically range from about 300 kg/m 3 to about 600 kg/m3. The strength of LCC correlates strongly with the density of the material; for the above range of densities, the average uniaxial compressive strengths of LCC materials, of the type used in MSE walls in California are typically between 500 and 3,000 kPa. Construction of LCC-MSE walls Except for the foundations, the LCC in MSE walls is generally formed between the facing panels (Figures 2 to 5). These panels play an important protective role as LCC is not as strong as conventional concrete and can be relatively easily damaged, e.g., by a small vehicle impact. The facing panels are anchored to the wall through reinforcements, as in conventional MSE walls. In our experience, LCC-MSE walls have typically been reinforced with metal straps, steel rods and geogrids (Figures 3 to 5). The reinforcement of the LCC in MSE walls is sometimes continuous (as in Figure 3), or limited in length to the area near the outer face (as shown in Figure 4).

2 CONSTRUCTION OF LCC-MSE WALLS Nature and Physical Properties of LCC Lightweight Cellular Concrete (ACI, 2006) is created by adding stable air cells during the manufacture of the material (Figure 1), and LCC is placed in a manner quite similar to concrete (Figure 2). Relatively few MSE walls with LCC as infill have been built in seismic areas and its use is generally considered a novel practice. The manufactured process of LCC results in a concrete preformed foam that has very low densities,

Figure 3. Placement of LCC over continuous geogrid reinforcement, within the panels of an MSE wall being built for the SVRT system near San Jose, California (design of wall detailed in GDC, 2014).


PRADEL D. et al.

Figure 4. LCC placement in an MSE wall reinforced with metal straps near the facing panels (Cell-Crete, 2015).

109

Figure 7. Detail of FLAC model including geometry of LCC and concrete slab. Stage shown is before backfilling next to wall (each underlying grid is 3.3 m = 10 feet).

3 NUMERICAL MODELING Introduction

Figure 5. LCC placement by in an MSE wall reinforced with steel rods (Cell-Crete, 2015).

Figure 6. FLAC model including geometry of LCC and subsurface soft soil profile (each underlying grid is 3.3 m = 10 feet).

To understand the seismic behavior of a geogrid reinforced LCC-MSE wall we conducted numerical analyses using the computer program FLAC version 7.0 from Itasca (2011). For the analyses shown herein, we adopted the subsurface conditions, wall geometry and seismic design loads of an LCCMSE wall that was recently completed in 2014-2015 for the Silicon Valley Rapid Transit (SVRT) in San Jose, California (shown in Figure 3). This project will extend subway service from San Francisco to San Jose. The adopted model geometry for the LCC-MSE wall is shown in Figures 6 and 7. It includes about 9 m of LCC materials built in 8 stages (about 2 m below and 7 m above final grade), and a rolling slab about 1.5 m thick at the top composed of conventional reinforced concrete. Since, in our opinion, the SVRT wall (being built in Figure 3) was well designed but very conservative, we decided to analyze a weakened version of the constructed wall. As a result, certain aspects of our numerical analyses are significantly different than those used for the original design by the design engineers as well as the final construction. For example: (a) the geotechnical consultant’s numerical analyses did not include geogrid elements in their FLAC models (GDC, 2014), (b) the final construction incorporated full length geogrid between the panels (as shown in Figure 3), (c) in our analyses geogrid reinforcement only extends 1/3 of the length between the walls, (d) we used reduced (lower values) LCC strengths than specified during construction, and (e) we used in FLAC the dynamic properties obtained from cyclic simple shear tests recently performed at California State University Fullerton (Tiwari, 2015a & 2015b) for the Cell-Crete corporation (Cell-Crete, 2015). Please note that Professor Tiwari’s triaxial and dynamic simple shear tests were performed on a large number of LCC samples that were prepared for


110


different sets of LCC densities, and that we used only the data that was the most appropriate for the adopted LCC density in our analyses, i.e., 480 kg/m3. In summary, the analyses presented herein, should be considered applicable to a hypothetical MSE wall that has a geometry similar to the one shown in Figure 3, but that is purposely much weaker. The differences between the constructed and analyzed wall are considered substantial. Adopted Properties of concrete and LCC Materials Both the heavy reinforced concrete slab and LCC materials were modeled as elastoplastic materials with a compressive strength of 20,000 kPa and 275 kPa, respectively. Please note that the adopted LCC strength is conservative for the specified density of 480 kg/m3. As previously indicated, our adopted LCC compressive strengths are significantly lower than the strength specified for the constructed wall in Figure 3, where we understand minimum LCC strengths around 550 kPa were ultimately required and vastly exceeded by the LCC supplier (Cell-Crete). Dynamic Properties of soils and LCC Materials The shear wave velocities near subject site and in San Jose, California, are known to considerable depths and have been the subject of numerous studies. Typical shear wave velocity profiles are reported in Chiu et al. (2008) for depths of 0 to 410 m (including the deep suspension logging performed by the USGS). For our numerical analyses, we used both the specific shear wave velocities obtained from the site’s subsurface characterization (GDC, 2014) as well as the deeper data in Chiu et al. (2008). The adopted shear wave velocities varied linearly in clays from 150 m/s at the surface to 300 m/s a depth of 30 m. Sand layers below 30 m were modeled with a constant 450 m/s velocity. Similarly, undrained shear strength varied from about 75 kPa near the surface to about 160 kPa at a depth of 30 m.

Figure 8. Accumulation of permanent shear strains and hysteretic damping in FLAC based on Massing (1926) rule.

Figure 9. Matching of modulus reduction and damping curves used for clays and Vucetic & Dobry experiments (1991).

To allow damping to vary with time during dynamic loading in FLAC, we adopted, for simplicity, hysteretic damping based on the Massing (1926) rule and we used as backbone the shear moduli curves at specific depths. This technique allows the accumulation of shear strains during dynamic loading as shown in Figure 8. Please note that a small amount of Rayleigh damping of 0.2% was added for numerical stability as well as energy dissipation at small loading cycles.

Figure 10. Matching of modulus reduction and damping curves used for clays and Darendeli experiments (2001).


PRADEL D. et al.

The modulus reduction curves that we used in this study were based on Vucetic & Dobry (1991) for the clay materials (i.e., in the upper 30 m) and Darendeli (2001) for the sands (between 30 m and the base of our model at 45 m). Since FLAC uses mathematical expressions instead of curves, we matched the experimental curves by Vucetic and Dobry as well as Darendeli as close as possible in the main area of concern, i.e., for shear strains of 0.05% to 0.2% (as shown in Figures 9 and 10). Note that the damping shown in Figures 9 and 10 is directly obtained from the modulus reduction curves through the use of the Massing (1926) rule. Geogrid, rolling slab and panels The adopted MSE reinforcement in our analyses consisted of Tensar UX1400 geogrid, which is one of the weakest uniaxial geogrids in the UX series. This geogrid has a tensile strength of 27 kN/m at 5% strain. The geogrids were modeled in FLAC using axial elements (known in FLAC as cable elements). Due to its thickness the concrete rolling slab, which seats on top of the LCC-MSE wall, was modeled using solid elements. To account for the inertial loads due to ballast, rails and similarly related machinery and equipment, the unit weight of these solid elements was increased appropriately. The MSE wall’s side panels were modeled in FLAC using liner beam elements having a thickness of 3 cm. These elements were given the properties of reinforced concrete and were directly attached to the end of the geogrid layers. Figure 11 depicts the computed vertical stresses and tensile forces in the geogrid layer at the end of construction. Please note that the predicted tensile forces in the geogrid layers are less than 2% of its tensile strength. These tensile values are low due to the relatively high shear strength of the LCC materials.

111

4 GROUND MOTIONS AND NUMERICAL ANALYSES To minimize wave reflections at model boundaries a quiet (viscous) boundary was specified along the base of the model, and free-field boundaries were specified along the edges, as shown in Figure 12. The ground motions used in our analyses were obtained from existing records of earthquakes with magnitudes and accelerations similar to those anticipated at the San Jose site (Abrahamson, 2012) and included both Fault Parallel and Fault Normal components for each of the records. The motions were spectrally matched (to the design spectral accelerations shown in Figure 13) using the software codes RspMatch and RspMatchEDT, pre and post processors for RspMatch (Abrahamson, 1992, and Geomotion, 2011a). The surface ground accelerations and spectral accelerations (before and after spectral matching) are shown in Figure 13.

Figure 12. Quiet and free field boundaries used in FLAC.

Surface ground motions were deconvoluted in order to obtain the input velocities at the base of our FLAC model using the program SHAKE2000 (Geomotions, 2011b). The results of our numerical simulations are shown in Figures 14 and 15. During our simulations, the computed maximum horizontal displacements ranged from 1 to 4 cm, and we predicted differential vertical movements from ¼ to 1½ cm. Figures 14 and 15 show the deformed mesh in an exaggerated scale from the beginning (top plot) to the end of the earthquake (bottom plot). As can be seen the mesh moves mainly in a horizontal and quasi-rigid manner. Rocking of the LCC-MSE wall is very minor and not noticeable in Figures 14 and 15.

Figure 11. Computed geogrid tensile forces and vertical stresses in ground, LCC wall, and rolling slab.


112


(shear, bearing, etc.). Similarly, our numerical analyses did not predict failure in the wall’s LCC materials either, i.e., we did not observe tensile, shear or compressive failures of the LCC infill. As seen in Figures 14 and 15, the side panels have inertial loads that pull on the geogrid layers and result in an increase of geogrid tensile forces. Figures 14 and 15 show that these tensile loads are almost immediately transferred to the LCC during an earthquake, and it appears that after approximately 2 m the geogrid has virtually no beneficial seismic role. Hence, the main role of geogrids appears to be for crack control purposes (e.g., cracks resulting from material shrinkage or minor differential movements resulting from varying bearing conditions). 5 CONCLUSIONS

Figure 13. Accelerations (in g) of the 1992 Landers earthquake and spectral accelerations, before and after spectral matching.

Because the reinforced concrete rolling slab at the top of the wall is heavy, it does create inertial overturning moments. However, these moments appear to be easily countered by the large base of the LCC-MSE wall which distributes them relatively evenly and prevents large vertical strains in the subgrade soils near the edges of the MSE wall. Even during the most intense portion of the earthquake (mid-plots in Figures 14 and 15), the vertical and horizontal stresses are relatively evenly distributed throughout the base of the wall, and do not lead to a shearing along the base or bearing failures below the LCC-MSE wall. In summary, our model does not predict a soft ground failure of any type

Our numerical simulations were performed on a significantly weaker version of the LCC-MSE wall shown under construction in Figure 3, that was built at a soft clay site in the City of San Jose, for the Silicon Valley Rapid Transit system near San Francisco, California. The designers (GDC, 2014) used LCC for this wall in order to significantly lighten the load of an originally proposed traditional MSE wall which used soil as infill. The switch from soil to LCC was very successful and allowed the designers to completely eliminate the need for soil improvement under this MSE wall. Our FLAC analyses indicate that on soft clay sites, LCC-MSE walls, such as the one analyzed herein, move in a quasi-rigid manner during earthquakes, i.e., they move mainly parallel to the ground surface and do not develop significant total or differential permanent vertical seismic movements. Our numerical modeling which was conducted on a weaken version of a constructed LCC-MSE wall, did not predict ground failures (shear, bearing, sliding, overturning, etc.) under static (during construction) or under design seismic conditions. Similarly, failure of the LCC materials in either compressive, tensile or shear modes of failure was not predicted. Our analyses indicate that the role of the geogrid reinforcement during earthquake loading is mainly to hold the side panels and that for seismic loading geogrid reinforcements provide little benefit to the wall beyond a distance of about 2 m from the face of the wall. Hence, the main role of geogrids in LCC walls appears to be for crack control purposes (e.g., cracks resulting from material shrinkage and/or minor differential movements resulting from varying bearing conditions), and designers may consider limiting both the reinforcement lengths and type.


PRADEL D. et al.

Figure 14. Deformed mesh, vertical stress contours and geogrid tensile loads (from beginning to end of earthquake).

113

Figure 15. Deformed mesh, horizontal stress contours and geogrid tensile loads (from beginning to end of earthquake).


114


REFERENCES Abrahamson, N.A. (1992). “Non-stationary spectral matching”, Seismological Research Letters 63(1), 30. Abrahamson, N. (2012) “Updated Peer Review of the 2005 ARS Curves for the Silicon Valley Rapid Transit Project,” dated August 13, 2012. ACI (2006). “Guide for Cast-in-Place Low Density Cellular Concrete”, ACI-523.1 R-06. Cell-Crete (2015), “Cellular Concrete Engineered Fill from Cell-Crete”, www.cell-crete.com. Chiu P., Pradel D. Et al. (2008), “Seismic Response Analyses for the Silicon Valley Rapid Transit Project”, ASCE GSP 181, Geotechnical Earthquake Engineering and Soil Dynamics IV, Sacramento, 1-10. Darendeli (2001). “Development of a new family of normalized modulus reduction and material damping curves”, Ph.D. Dissertation, Dept. of Civil Engineering, Univ. of Texas, Austin, TX. GDC (2014), “Dynamic Numerical Analyses using FLAC for Lightweight Cellular Concrete MSE Walls for BART Berryessa Extension, San Jose, CA”, report by Group Delta Consultants dated July 11, 2014. Geomotions (2011a), “RspMatchEDT user manual”, www.geomotions.com/Download/RspMatchEDTM anual.pdf. Geomotions (2011b), “SHAKE2000, user manual.”, www.geomotions.com/Download/SHAKE2000Man ual.pdf. Itasca (2011), “FLAC (Fast Lagrangian Analysis of Continua) Version 7.0”, Minneapolis, USA, www.itascacg.com. Masing, G. (1926), “Eigenspannungen and verfertigung beim messing”, Proc. 2nd Int. Congress on Applied Mechanics, Zurich. Teig J. and Anderson J. (2012). “Innovative Design for the Colton Flyover Grade Separation of IPRR and BNSF, Colton, CA” AREMA Annual Conference & Exposition. Tiwari B. (2015a), “Preliminary Result – Static Shear Strength of Lightweight Cellular Concrete Sample Batch 1”, report dated January 26. Tiwari B. (2015b), “G over Gmax & damping tests”, personal communication of April 4th. Vucetic & Dobry (1991), “Effect of Soil Plasticity on Cyclic Response”, ASCE Journal of Geotechnical Engineering 117(1), 89-107.


3rd.


Session 3: Soil improvement

Technical Committee

TC-214

Technical Committee




Soil improvement around the world – Applications and solution examples Mejoramiento de suelos alrededor del mundo – Aplicaciones y ejemplos de solución F. GERRESSEN1 1

BAUER Maschinen GmbH, Schrobenhausen, Germany

ABSTRACT: The uses of soil improvement to utilize areas, which are not suitable for foundation purpose, have a long tradition. Due to the increasing demand of foundation works in areas of unsuitable soils, the improvement of the existing soils becomes even more important in the future. Various applications and systems for various demands exist around the world. Either it is a “simple” improvement for settlement reduction or a more important aim as liquefaction mitigation. However, each system has its advantages, limits and needs. The paper will describe some basics of various techniques like e.g. Vibro-flotation, Dynamic compaction and soil mixing. The main part will focus on jobsite examples, where the systems provided solutions for different purposes, e.g. simple foundation or liquefaction mitigation.

1 GENERAL Soil improvement techniques continue to progress in addressing ground engineering problems across the world, especially in urban areas where land development and reuse need to be efficient not only in geotechnical engineering but in time, cost and energy used. These techniques provide a toolbox for geotechnical engineers who look for opportunities to modify the ground characteristics and behavior, not accepting the soil as it is. Focus can be taken as well on geo-mechanical properties like e.g. strength and deformability as on the hydraulic conductivity, which might be either increased or decreased. A big focus can also be given to the aim of liquefaction mitigation. Based on the analysis of the local conditions, potentials and limitations of a possible soil improvement one has to do a pre-selection of possible techniques out of a wide range of applications. Looking to the literature, all techniques can somehow be put into different categories, whereby the definition of the categories is not strict. For instance, once you find the vibro-flotation in the category of “vibro and impact compaction”, which is more related to the soil conditions and the way of improvement, another time you find it at the “deep vibro techniques”, which is more related to the used equipment. 2 TECHNIQUES Trying to categorize the various techniques in a way that the number of categories will not be too high, will lead to a minimum of five categories.

- Self-compaction by Vibro or impact compaction - Dewatering by vertical drains or Vacuum consolidation - Soil Mixing - Soil-displacement by rigid inclusions or stone columns - Grouting A classification of ground improvement techniques adopted by the TC 211 (Former TC17) of the International Society for Soil Mechanics and Geotechnical Engineering (ISSMG) covers 29 different methodologies also in 5 different categories. Some of the most important techniques will be described in the following. 2.1 Vibro – compaction, also known as vibroflotation (VF) …is one of the oldest soil improvement techniques, developed back in the 1930th by the Keller Company. VF is applicable in non-cohesive and slightly cohesive granular soils such as sands and gravels, as well as slag deposits. It is suitable for carrying high loads on the improved subsoil, including dynamic loads without significant settlements. It offers a particularly economical application in fully saturated soils below the groundwater table. The hydraulically (or electrically)-powered vibrator assemblies will be suspended from crawler-mounted cranes. The power will be supplied by either the enhanced hydraulics of the special cranes or by a power-pack mounted on the back of the crawler crane.


118

Soil improvement around the world – Applications and solution examples

Figure 1: Crawler crane mounted depth vibrator left picture without, right picture with power pack.

The flushing medium can either be water taken from rivers or existing groundwater. Both fresh and salt water are suitable. The flushing volume ranges typically from 50 to 90 m³/h. It is pumped at a pressure of about 6 to 8 bar to the top of the vibrator assembly. 2.2 Dynamic compaction … is a soil densification method developed by Menard back in the 1960th. The soil is compacted by repeated dropping of a heavy weight (pounder) from a predetermined height onto the ground surface. The imparted high kinetic energy, which is transmitted to deeper soil layers, forces the soil particles into a denser state of compaction. The degree of compaction depends on the weight of the pounder, the height from which the pounder is dropped, and the compaction grid is dropped. A heavy weight (pounder) is dropped in several passes in a primary, secondary and often tertiary grid. The primary grid (widest spacing) is used to achieve compaction at depth. It uses the largest weight and highest drop. The secondary and tertiary grid is used to achieve compaction at medium and shallow depths. The process is completed by compacting the surface layer across the treatment area in a final contiguous “ironing pass”.

Figure 2: Heavy-duty Crawler crane mounted with steel pounder.

2.3 Soil Mixing …is a methodology were the soil is improved by mixing it with cement, lime or other binders in situ by using a mixing tool. This methodology can be used either as wet mixing or as dry mixing. In the more frequently used wet mixing process, usually a mixture of binder and water, maybe with additives, is injected and mixed with the soil. Depending on the type of soil and binder, by the end of the mixing process, a mortar like mixture is created which hardens during the hydration process. In the dry mixing process, the binder is directly mixed with the soil and reacts directly with the existing soil and water.


GERRESSEN F.

Mixing tools show several versions, so that they are mixing around either vertical or horizontal axis, mixing in a trench and maybe are jet assisted. Figure 3 shows a classification done by Bruce 2010.

119

Depending on the technique, improvement depth of up to 80 m can be achieved.

Figure 3: Updated Deep Mixing Method classification (Bruce, 2010).

Even the history of these techniques starts predominantly in Japan (wet mixing) and Scandinavia (dry mixing) in the 1970th, nowadays soil mixing is getting more and more important all around the world. One of the latest developments in the field of wet mixing was the Cutter Soil Mixing (CSM) method, developed in a joint venture of BAUER Maschinen GmbH and Bachy Soletanche about 10 years ago. CSM is used mainly for stabilizing soft or loose soils (non-cohesive and cohesive), however the machinery used, derived from Bauer’s cutter technology, extends the applicability of the method to much harder strata when compared to other methods of soil mixing.

2.4 Rigid inclusions (RI) and stone columns (SC) …are classified by the TC211 as ground improvement methods with admixture or inclusions. They can be installed using various techniques. While the SC method is quite old, based on the development of vibro compaction back in the 1930th, the use of RI is first mentioned in the 1980th. Usually, the construction takes place using a regular grid of vertical elements, triangular or square, in soil layers with low bearing capacity and/or high compressibility, either down to a competent layer or to a defined depth where no competent layer is reachable. Due to the fact, that the elements are stiffer than the surrounding soil, they take part of the load, with the aim to provide in combination with the surrounding soil sufficient bearing capacity and/or acceptable settlement resistance. There are a few differences between the RIs and SCs. In comparison to the SCs, RIs don’t require lateral confinement from the existing soil due to their properties in term of stiffness/strength. Typical methods used for RIs are e.g. the well know Controlled Modulus Columns (CMS) or Vibro Concrete Columns (VCC). Replacement ratios are in a range of 10 to 35 % for SCs, but only 2 to 10 % for RIs. Typically RIs require a load transfer platform.

Figure 4: BCM Mixing head for CSM.


120


3 JOBSITE EXAMPLES 3.1 CSM Technology at Fréjus, France For the foundation of an industrial building on a site with up to 6 m low strength colluvial soil over a marlsandstone, a soil improvement solution had to be found. Main purpose of the design was as well the minimization of the total as the differential settlement of the structure, which covers about 3600m². The performed solution was a combination of soil cement panels, carried out by the CSM technology, overlaid by a load transfer layer, with a final placement of the concrete slab of the building (Figure 5).

samples from the final CSM panels were important to adjust execution parameters and obtain required properties of the soil cement panels. Samples taken to prove the achievement of the required UCS value of 3 MPa as a minimum, design value of 1.5 MPa UCS at working load with a safety factor of 2, showed a good success. The choice of CSM technology could be seen as the most suitable solution under the local circumstances, as it combines a number of advantages, technical, economical and environmental. For instance the ability of executing the system in all jobsite soil conditions, including the Marl-Sandstone, and it’s use as construction material with the benefit of spoil reduction. In addition, the system allows an installation with reduced vibrations. 3.2 Stone Columns at Formosa Plant, Taiwan

Figure 5. Cross section and geometry for finite element modelling.

The CSM panels were carried out as single panels at a size of 0.6 x 2.4 m in a rectangular pattern (Figure 6), with enlarged panel caps of 1.8 x 3.6 m to allow more efficient load transfer. All panels needed to key into the Marl-Sandstone for a minimum of 0.5 m.

Observing Taiwan's chronic shortage of upstream petrochemical materials, the Formosa Plastics Group proposed the Naphtha Cracking Project. Following several rejections by the government, final approval was obtained to build the No. 6 Naphtha Cracking Project. The project became part of a huge offshore industrial complex development with a total of 41 industrial plants. In addition to the industrial plant a new harbor and a new town were built as well. The development started with huge land reclamation. About 100 Million m³ of marine sand were spread out over an area of more than 200 hectares with approximately 8 km in length along the coastline and about 4 km into the sea. Heavy structure were founded on driven pre-cast piles, but the remaining area was first improved by dynamic compaction followed by stone column installation for the deeper parts up to 20 m. Main purpose for the decision of ground improvement by the use of stone columns was to reduce the liquefaction potential of the soil during earthquakes. The earthquake design criteria specified a magnitude M = 7.3 and an acceleration of a = 0.21 g. Based on the design criteria, the existing soil parameters and the chosen equipment, a 2.5 m triangular spacing, exemplary shown in figure 7, was considered as feasible. Test sections were carried out to proof and establish the right spacing and production parameters.

Figure 6: Rectangular pattern of CSM panels with panel cap enlargement.

Due to the use of soil as construction material, quality control during installation by using the operators monitor to control the installation parameters in real time as well as the test of achieved unconfined compressive strength on SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

GERRESSEN F.

121

Figure 8: Plant and earthquake location. Figure 7. Typical layout for stone column arrangement.

In total 1,500,000 linear m of stone columns were installed in a depth range of 13 to 20 m from 19972000. About 60 %, in particular the deep ones were installed using the Bauer TR vibrator and the bottom feed method, which ensures the forming of a continuous stone column over the full depth. The design could be proved already, as in 1999, two earthquakes took place. First the Chi-Chi earthquake with a magnitude of M = 7.3 on September, 21st, followed by the Chia-I earthquake with a magnitude of M = 6.8 on October, 22nd. Especially the Chi-Chi earthquake, also known as the 921 earthquake, caused disastrous damages. More than 2,400 people were killed, over 50,000 building were destroyed, and another 50,000 buildings were severely damaged.

Immediately after the earthquake event all plant operation was stopped looking for immediate survey of the structures by consultant firms CTCI and RESI. The survey showed no structural damage of structures and tanks at all at a maximum settlement of 10 mm. Therefore, the plant could get back in operation in full scale very shortly after the earthquake. Also the Geotechnical Earthquake Engineering Server (GEES), which provide constant information about worldwide earthquake impacts states in one of their reports:” Liquefaction happened at untreated ground at Formosa Plastics Industrial Park in Mailiao. This park is a reclaimed island, but no liquefaction or ground failure was noted at treated areas within this park; Formosa Heavy Industries Corporation used a combination of dynamic compaction, preloading and stone columns, with piles generally supporting buildings”.

Figure 9. Damages in Tai Chung harbor September 21st.


122


4 CONCLUSION The wide range of soil improvement techniques offer a wide range of applications to utilize soils, which are, due to their mechanical properties, usually not suitable for foundation purpose. One of the most important applications can be seen in the use of liquefaction mitigation. REFERENCES Wolfgang Brunner, A.rthur Bi, Yan Lian Chen 2002. Ground improvement by stone columns at Formosa plant Taiwan and its earthquake response, Ninth international conference on Piling and deep foundation, Nice, 2002. Franz-Werner Gerressen, Thomas Vohs, 2009. Soil Improvement, Vibroflotation, Vibroreplacement and concrete columns, Simposio International, Las Technologias y los Sistemas de Cimentation para el siglo XXI, Mexico City, 2009. Bruno Simon, General Report-Rigid Inclusions and Stone Columns, 2012, ISSMGE-TC211 International Symposium IS-GI, Brussels, 2012. Artur Peixoto, Estala Sousa, Pedro Gomes, 2012, Solution for soil foundation improvement of an industrial building. Donald Bruce, Deep Mixing in the United States: Milestones in evolution, Deep Mixing 2015, San Francisco 2015.


Technical Committee




Principles and application of soil mixing for ground improvement Principios y aplicación de la técnica soil mixing para mejoramiento de suelo Charles M. WILK1 1

ALLU Group Incorporated

ABSTRACT: Mass stabilization (MS) through soil mixing is a ground improvement technique that can prepare areas of low bearing strength soil for subsequent infrastructure development. The technique involves mixing binding agents such as portland cement, fly ash, slag, or lime into the subject soil while the soil remains in-place (insitu). The binders “cement” the soil grains together to form a cement modified soil or a soil cement. The MS-treated area now has improved bearing capacity to support infrastructure or to prevent movement of the material such as landslides. The same insitu soil mixing technique can be used to address contaminated areas. Binders or reagents are mixed into soil. The treatment protects human health and the environment by immobilizing hazardous constituents within the treated material. When used for the purpose of environmental remediation the technology is called Insitu Solidification/Stabilization (ISS). Both Mass Stabilization and ISS treatments require laboratory studies to develop a mix design of soil and binder(s) that produce the desired physical and/or chemical properties. The mix design is then transferred into the field. Successful MS and ISS treatments rely on reproduction at full-scale of the mix design and the thorough mixing attained at laboratory scale. Fifty to 70% of the cost of a MS or ISS project is in the cost of the binding agent that is to be mixed into the subject soil. Underdosing, overdosing, non-thorough mixing, and mixing in the wrong areas all create cost over-runs. This paper will discuss recent innovations in mass stabilization systems that improve the cost effectiveness of the treatment technology. Specialized equipment can impart greater mixing shear thus improving the thoroughness of mixing. Dry powder pressure feeders can conserve the “drying capacity” of binder resulting in higher strengths at lower binder dosages. Global Positioning System (GPS)-based systems can guide the mixing operator for complete mixing coverage. An integrated tracking and feeding system can record that proper dosing and mixing was accomplished and generate construction QA/QC reports for the client.

1 INTRODUCTION

2 PRINCIPLES OF MASS STABILIZATION

New or expanding infrastructure projects may encounter areas of low bearing capacity soil or areas where marginal materials such as waste, dredged material, or sediment have been placed. Mass stabilization is a geotechnical method that can be used to improve the bearing capacity or stiffness of soil. Mass stabilization involves mixing binding agents into soil subject to treatment while the soil remains in-place or insitu. Mass stabilization technique can also address contaminated soil therefore, the same technique can be used to attain both civil and environmental engineering goals. In this environmental application the technology is known as insitu solidification/stabilization (ISS). ISS protects human health and the environment by immobilizing hazardous constituents within the treated material.

2.1 Properties of Mass Stabilized soil Mass stabilization is used to improve the construction properties of a marginal soil. Marginal soil includes, but is not limited to, peaty soil, high water content soil, or soil with high proportion of silt or clay. Mixing a cementitious binder into marginal soil creates a material similar to cement-modified soil. When accompanied with compaction at the time of construction and use of a higher addition rate of cement, the mixed material may become similar to soil cement. The goals of mass stabilization may include: • Increase in the California Bearing Ratio (CBR). • Increase in Unconfined Compressive Strength (UCS) (Figure 1). • Reduction in plasticity characteristics as measured by Plasticity Index (PI). (applicable to clayey soil). • Reduction in the amount of silt and clay size particles. (through agglomeration or cementation).


124 • • • •

Principles and Application of Soil Mixing for Contaminated Properties

Increase in shearing strength. Decrease in volume-change properties (applicable to expansive soils). Reduction in hydraulic conductivity (Figure 2). Reduction of pore water.

to create concrete used in construction. In MS, cement by itself without sand or aggregate, is mixed into soil producing cement-modified soil or soil cement. Water may be added if the soil does not include sufficient water to hydrate the cement. Cementitious or pozzolanic industrial by-products such as ground granulated blast furnace slag, and fly ash have also been used to reduce binder costs. 2.3 Bench-scale Mix Designs

Figure 1. Unconfined Compressive Strength testing.

Projects usually include some level of laboratoryscale mix design. Several addition rates of a binder or combination of binders are mixed with representative samples of the soil subject to treatment. Mixing is usually done by bench-sized power mixers such as KitchenAid® or Hobart® stand mixers (Figure 3). Powered stand mixers impart significant mixing energy and shear (Figure 4) to speed the work and assure thorough mixing. Thoroughness of mixing is important for reproducibility and comparison. Mixed samples are cured and then tested to determine if the desired engineering/physical properties are achieved.

Figure 2 : Hydraulic Conductivity testing.

Figure 3. Bench-scale work by stand mixer.

Goals for ISS of contaminated soil may include reduction of the leachability of hazardous constituents, and reduction of hydraulic conductivity thus immobilizing hazardous constituents. An increase in strength usually is desired as well. First lines of paragraphs are indented 4 mm (0.16") except for paragraphs after a heading or a blank line (Primer Parrafo tag). 2.2 Binders Mass stabilization involves mixing binder into soil subject to treatment. A variety of binders have been used including portland cement, fly ash, slag, kiln dusts, and lime. The most commonly used binding agent in mass stabilization is portland cement. Most people are familiar with portland cement as the gray powder that is mixed with sand, aggregate, and water

Figure 4. Mixing shear developed at bench-scale.


WILK C.

125

A matrix from successful mix designs is compiled. The matrix often includes the relative costs of the binder(s) used in the successful mixes. Cement addition rates may range from 60 kilograms per cubic meter (kg/m3) to 250 kg/m3. Fine-tuning the mix design is important from a cost perspective. Fifty to 70% of the cost of a full scale mass stabilization project is the cost of the binder(s) used (ALLU 2015). There is a real incentive for selection of the least expensive mix design and efficient use of the binder(s) that meets the desired project performance standards. 3 INNOVATIONS IN FULL-SCALE MIXING Figure 6. Photograph of Power Mixer and Pressure Feeder on-site

3.1 Mixing Equipment Mass stabilization at full scale has been done with a variety of mixing equipment. These include deep soil mixing augers, high pressure jets, pulver mixers or road reclaimers, bare excavator buckets and power mixer attachments to excavators. Excavator-based power mixing appears to be particularly well suited for infrastructure improvement since: (a) the attached mixing head can reach the required depths of treatment, (b) excavators are commonly available, (c) excavators are easily transported to the jobsite, (d) the range of motion of an excavator’s arm reduces the amount of time lost in moving the mixing equipment in order to treat an area and (e) the attached power mixer imparts the mixing shear needed for thorough mixing and efficient use of binder(s). Power mixers are attachments to an excavator (Figures 5 and 6). Power mixers resemble a very powerful rototiller on a 7-meter stem. The power mixers are powered by the hydraulic system of the excavator. The mixing head of the power mixer imparts the mixing shear necessary to thoroughly mix the binding agent into the subject soil. A nozzle located between the mixing drums is used to inject binder(s) at the point of mixing.

3.2 Binder Injection “Wet” vs “Dry” Mixing binder into marginal soil requires some method of injecting the binder into the subject soil. Binders are injected either in wet or dry form. In “wet mixing” the binder is mixed with water prior to injection. Using portland cement as an example, the cement is mixed with enough water to produce a pump-able grout. This grout is then injected and mixed into the soil. An excess amount of water is used to produce a pump-able cement grout. “Excess” because in order to make a pump-able grout a water to cement ratio (w/c) greater than 0.5 is needed. Portland cement chemically requires only a little less than half its weight in water to fully hydrate- a w/c ratio of a little less than 0.50. Highest strengths are achieved at low w/c ratios- 0.5. (Kosmatka et al 2006) When combined with water that may already be present in the soil the ideal w/c ratio may be exceeded and even more grout and corresponding cement (and cost) may be needed to achieve the desired strength properties of the treated soil. “Soft” (marginal) areas of soil are often the result of excessive water in the soil. Mass stabilization is very often used to “dry” soil to improve its engineering properties. Dry mixing may be appropriate in areas where there is sufficient water already present in the soil to hydrate the cement. Dry mixing uses pneumatic Pressure Feeders to feed the binder to the augers, power mixers, or mixing equipment. Dry mixing is often more cost effective compared to wet mixing. Utilizing the existing water within the soil may result in a water cement ratio closer to 0.5. Higher strengths can be attained with lower addition rates of cement. Note that over 50% of the cost of a mass stabilization project is the cost of the binder.

Figure 5. Schematic of Excavator-mounted Power Mixer with dry binder Pressure Feeder.


126

Principles and Application of Soil Mixing for Contaminated Properties

3.3 Locating, Controlling, Recording for QA/QC The engineering properties of mass stabilized areas are critical for the support of buildings, highways, roads, industrial facilities, and rail systems. Efficient insitu mixing and meeting performance standards requires monitoring of the actual location, depth, binder addition, and thoroughness of mixing. The advent of GPS-equipped mixing equipment, computer monitored binder feed and metering devices, and computer recording of these conditions has made this job easier (Figure 7). Automated data recording is used in quality control and quality assurance documentation.

dosing and mixing was accomplished and generate construction QA/QC reports for the client. 5 REFERENCES ALLU OY, 2015, Mass Stabilization Manual. Kosmatka, S. H., Kerkhoff, B., Panarese, W.C., 2006, Design and Control of Concrete Mixtures 14th Edition, Portland Cement Association, Skokie, Illinois.

Figure 7: Schematic of GPS and mixing data recording system mounted on Power Mixer and Pressure Feeder. Illustration includes excavator operator’s display guiding operator- green completed “blocks” and red not-yetcompleted “blocks” and graphic of data reports.

4 CONCLUSIONS 4.1 Mass stabilization is a ground improvement technique that can prepare areas of low bearing strength soil for subsequent infrastructure development. Insitu Solidification/Stabilization is a similar technology that can address contaminated soil. MS projects begin with bench-scale testing to determine the effective binder and addition rate to use. Fifty to 70% of the cost of a MS project is the cost of the binder mixed into soil. Recent innovations in soil mixing, binder injection, location, dosing and recording systems improve the cost effectiveness of MS. Specialized excavator-mounted power mixers are capable of efficiently injecting and mixing binders to required depths. Power mixers impart greater mixing shear thus improving the thoroughness of mixing. Dry powder pressure feeders can conserve the “drying capacity” of binders resulting in higher strengths at lower binder dosages. Global Positioning System (GPS)-based systems can guide the mixing operator for complete mixing coverage. An integrated tracking and feeding system can record that proper SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

Technical Committee




Sustitución dinámica aplicada en turbas de la península de Yucatán Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan Alfredo CIRION ARANA1, Rémi CHATTE1 & Juan PAULÍN AGUIRRE 2 1

2

Menard México CIMESA, Soletanche-Bachy, México

RESUMEN: Se describe el procedimiento constructivo de Sustitución Dinámica utilizado como sistema de mejoramiento masivo de suelos con propiedades mecánicas pobres. Se explica el concepto de mejoramiento con este tipo de columnas granulares de gran diámetro, se enumeran las bases de su diseño, en la descripción de la secuencia constructiva se destacan los controles durante la ejecución para garantizar su calidad. Se da un ejemplo real de la aplicación de esta técnica de mejoramiento de suelos en suelos orgánicos tipo turba.

1 ANTECEDENTES 1.1 Descripción del proyecto base El proyecto en general consiste en un desarrollo habitacional de más de 130 viviendas a construirse en un terreno de casi 7 ha de superficie, que es colindante a una laguna artificial creada para poder tener conexión con el mar caribe. El acceso vehicular a los lotes donde se construirán las casas se realizará a través de vías de circulación clasificadas como primarias y secundarias, según su ancho de calzada y nivel de rasante, las cuales llevarán a los autos directamente a los estacionamientos de las casas ubicados en semisótanos. Debido a las propiedades mecánicas pobres de los suelos que se encuentran en el terreno, que son predominantemente suelos de origen orgánico −turbas− de hasta 5.0 m de espesor, el proyecto original desarrollado para construir estas vías consistía en la construcción de un viaducto formado por marcos estructurales de concreto reforzado prefabricado formados por pilotes hincados hasta la roca subyacente a las turbas y trabes de concreto, que soportarían una losa también de concreto reforzado que sería la que finalmente funcionaría como el arroyo para el tránsito vehicular y también como el medio para soportar las instalaciones necesarias para la urbanización del desarrollo habitacional: instalaciones eléctricas, hidráulicas, pluviales y sanitarias.

1.2 Descripción de la propuesta alternativa Debido a la complejidad y el costo asociado a la construcción de las vialidades con el sistema

estructural descrito en el inciso anterior, Menard México propuso una opción alternativa para la construcción de éstas, la cual consistió en mejorar las características mecánicas del suelo turboso existente mediante la técnica conocida como Sustitución Dinámica, de este modo dar la capacidad de carga necesaria y el adecuado comportamiento en deformación, a dicho terreno, para poder construir como terracerías convencionales la vialidades según el nivel requerido por proyecto, bajo este esquema las instalaciones se construyeron de manera tradicional, dejando las tuberías, pozos y registros enterrados en el terraplén final. Con esta técnica de mejoramiento los asentamientos totales finales son controlados a corto y largo plazo quedando dentro de límites aceptables en condiciones de servicio (menores a 2.5 cm a largo plazo) y los asentamientos diferenciales a valores permisibles para conservar el bombeo de las vialidades. La Sustitución Dinámica es una técnica de mejoramiento masivo de suelos que consiste en la formación de columnas de gran diámetro (de 2 a 3 m) formadas con material granular compactado, aplicadas a terrenos con propiedades mecánicas pobres, distribuidas en mallas regulares, de forma general la masa tiene una geometría especial −punzonante− y su peso varia entre 10 y 15 t. Las alturas de caída típicas son entre 5 y 15 m. La formación de estas columnas, también conocidas como ‘plots’, se realiza mediante la incorporación secuencial y compactación de material granular en el terreno original a través de la aplicación de impactos generados con una masa que se deja caer repetidamente desde cierta altura con la ayuda de


128

Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan

una grúa. La Figura 1 muestra la secuencia de formación de las columnas.

E; módulo de Young c'; cohesión drenada φ’; fricción drenada cu; cohesión no drenada El Nivel de Aguas Freáticas se encontró a 1.0 m de profundidad promedio. 3 BASES DE DISEÑO

Figura 1. Secuencia de formación de columnas de material granular ‘plots’, mediante Sustitución Dinámica.

2 DESCRIPCIÓN GEOTÉCNICA El predio se encuentra en la península de Yucatán, en la zona costera, prácticamente colindando con el mar caribe. La zona anteriormente estaba ocupada por manglares, formada por depósitos sedimentarios de arenas, limos, arcillas y turbas, seguidos de estrados de roca caliza. Según los estudios de mecánica de suelos que se tenían, todo parecía indicar que las condiciones del sitio eran erráticas, es decir, que existía gran variabilidad en cuanto a los espesores y tipo de materiales. La posterior verificación a través de calas realizadas en sitio en zonas específicas durante la ejecución de los trabajos, confirmó dicha situación. En las Tablas 1 y 2 se presenta un resumen de las condiciones y propiedades estratigráficas del sitio.

Dadas las condiciones disimilares de suelos que se presentan a lo largo y ancho del sitio de construcción, y según la variabilidad de las pendientes y geometría de los terraplenes finales, según el proyecto y las necesidades arquitectónicas y de ingeniería, se llevaron a cabo numerosos análisis paramétricos y de sensibilidad para definir la factibilidad de la solución, así como el espaciamiento, profundidad y diámetro de las columnas, y los características asociadas a las columnas terminadas. Como parte de estos estudios, se realizaron cálculos analíticos y modelos basados en el Método de Elementos Finitos, obteniendo resultados de asentamientos totales y diferenciales críticos para las condiciones naturales antes del mejoramiento de suelos

Tabla 1. Condiciones estratigráficas del sitio. Zi (m)

U00

U01

U02 U03

+3.8 a +1.5 +1.0 a +0.5 -0.8 a -1.3 >-2.4

Zf (m)

H (m)

Descripción

+1.0 a +0.5 -0.8 a -1.3 -2.4 a -5.8 -20.0

0.5 a 2.0

Relleno (terraplén)

1.8

Arena tipo SASCAB

1.2 a 5.0 >17.0

Turba Roca Caliza

Tabla 2. Propiedades estratigráficas del sitio. Unid

U00 U01 U02 U03

γ

E

c'

φ'

cu

(kN/m³)

(MPa)

(kPa)

(°)

(kPa)

22 18 17 24

50 30 1 500

5 2 100

38 30 15 40

Figura 2. Modelo axisimétrico con elementos finitos realizado para una columna de sustitución dinámica (CSD), PLAXIS.

15 -

Donde: Zi; Nivel inicial Zf; Nivel final H; Espesor del estrato γ; Peso Volumétrico SOCIEDAD MEXICANA DE INGENIERÍA GEOTÉCNICA A.C.

CIRION A. et al.

f.

129

La vialidad quedará terminada una vez que se coloquen las instalaciones dentro del terraplén (excavando donde sea necesario) y se construya la carpeta de rodamiento, banquetas y pasillos peatonales.

Figura 3. Modelo plano de deformación con elementos finitos realizado para el análisis de estabilidad y verificación del tratamiento propuesto, PLAXIS.

4 DESCRIPCIÓN DEL PROCESO CONSTRUCTIVO Para el mejoramiento de suelos y la construcción de los terraplenes fueron consideradas las etapas siguientes: a. Una vez realizado el desmonte y despalme del terreno, inició la construcción de los terraplenes mediante la colocación de una plataforma de trabajo de material granular. Esta plataforma tuvo el objetivo de dar soporte a los equipos de construcción de las CSD, y a la vez garantizar su estabilidad. Fue una plataforma horizontal, plana y drenada.

Figura 4. Aportación de material granular para la conformación de CSD.

b. Con la plataforma terminada, se realizó el trazo de las CSD en mallas regulares de 4.0 x 4.0 m, 4.5 x 4.5 m y 5.0 x 5.0 m. c. Con una pre-excavación en el punto de ubicación de una columna, inició la aportación de la grava y ésta se comenzó a compactar mediante una masa metálica de geometría especial que se dejó caer, en caída libre, con una grúa. Cada impacto logró la inserción del material de aportación en el suelo natural. La columna se formó poco a poco hasta alcanzar el criterio de paro definido previo al inicio de los trabajos, el cual garantizaba la correcta formación de la columna.

Figura 5. Izado de la masa ‘punzonante’ con la grúa.

d. Realizada la construcción de una CSD, se procedió a repetir el procedimiento para las CSD subsecuentes. e. Posteriormente se colocó una capa de geotextil sobre la plataforma y se realizó la construcción del terraplén final en capas de 20 cm de espesor −ver Figura 6−, compactadas al 99 % de su PVSM, hasta alcanzar el nivel de rasante de proyecto.

Figura 6. Construcción de terraplén final.


130

Dynamic replacement soil improvement technique applied in peaty soils in the peninsula of Yucatan

5 CONTROLES Y VERIFICACIONES DE EJECUCIÓN Una parte esencial para asegurar la calidad de los trabajos del tratamiento de suelos aplicado es el control de la ejecución de las columnas y su posterior verificación. Además del control de la construcción llevada a cabo a través de la generación de los registros de construcción de cada una de las CSD, también fue necesario definir criterios de ‘paro’, para tener los parámetros de referencia necesarios que indicaran la finalización de la construcción de cada CSD. Estos criterios se basaron en la medición de los parámetros siguientes: a. Volumen de material incorporado b. Penetración, d, de la masa al caer dentro de la columna.

d

nivel de masa en golpe precedente nivel de masa en golpe actual

Figura 7. Verificación de la penetración de la masa.

Por otro lado, se llevó a cabo la verificación de las columnas terminadas mediante la realización de pruebas de campo realizadas directamente en las columnas:  Pruebas de presiómetro Menard  Pruebas de placa, aplicación de cargas superficiales incrementales a través de placas metálicas circulares Las primeras, con el presiómetro, se realizaron para la medición de los módulos presiométricos en toda la altura de la columna, mientras que las segundas se realizaron para la medición de los asentamientos y la obtención de las curvas cargaasentamiento-tiempo con las cuales se calculó la deformabilidad de la columna y se calibraron los modelos de análisis.

Figura 8. Pruebas de placa realizadas sobre las CSD.

6 CONCLUSIONES La Sustitución Dinámica es una técnica de mejoramiento para suelos de propiedades mecánicas pobres, que se adecúa bien para el tratamiento de suelos blandos y blandos orgánicos −turba−. El estudio de factibilidad de este tipo de mejoramiento requiere en principio de un análisis adecuado y detallado de las características del terreno y del sistema de transmisión de cargas para poder estimar correctamente los asentamientos que se tendrán en la realidad a corto y largo plazo, así como de la definición de las características de las columnas a construir: diámetro, profundidad, separación, tipo de material de relleno, etc. En este trabajo se ha presentado la aplicación práctica de este tipo de tratamiento en suelos blandos arcillosos orgánicos depositados sobre roca caliza en la península de Yucatán, para un proyecto de construcción de vialidades para la urbanización de un desarrollo habitacional. Se han descrito las ventajas económicas que esta alternativa de este mejoramiento aportó al proyecto contra la solución estructural original, y se ha mostrado el procedimiento de construcción realizado y los controles llevados a cabo en la obra para aseguramiento del funcionamiento del sistema de mejoramiento. REFERENCIAS Norma NF P 94-160-1. Auscultation d'un élément de fondation. Parties 1 et 2. Octobre 2000. Philipponnat Gérard, Hubert Bertand. Fondations et ouvrages en terre. Ed. Eyrolles. Paris, Francia. 2000. Cassan Maurice. Les essais in situ en mécanique des sols. Ed. Eyrolles. Janvier 1987. Magnan Jean-Pierre. Théorie et Pratique des drains verticaux. Technique et Documentation Lavoisier. Paris, Francia. 2000.


Technical Committee




Transforming marginal land to support a world class development in Panama Modificación de suelos marginales para apoyar proyectos de clase mundial en Panamá Gustavo LANGONI1, Roger ARCHABAL1 1Langan

Engineering and Environmental Services

ABSTRACT: The Santa Maria Golf and Country Club development consists of approximately 280 hectares of low lying coastal wetlands on the western fringe of Panama City in the Republic of Panama. The on-going development represents one of the most ambitious master-planned developments ever to be undertaken in Central America. Development of the site presented tremendous geotechnical challenges since the design flood criteria required that the existing surface grade be raised with approximately 5 meters of fill placed over a highly compressible marine clay deposit. This equated to nearly 10,000,000 cubic meters of fill to develop the site. The compressible clay stratum varied in thickness, but was typically between 5 to 10 meters thick. The required fill loads resulted in over one meter of consolidation settlement. To further complicate the site development challenge, the developer construction schedule required a site preparation period of less than six months from fill placement. The theoretical time required for consolidation without ground improvement would have ranged between 10 and 20 years. The objectives of this paper are to: 1) present the array of geotechnical design challenges resulting from the need for thick fills over highly compressible clay; 2) share the geotechnical properties of the site gathered during the subsurface exploration; 3) present the technical concepts associated with the ground improvement recommendation consisting of prefabricated vertical wick drains and surcharge to accelerate the consolidation and precompress the soft clay stratum within the short pre-development period; 4) discuss the use of value engineering ideas to provide the most economical approaches to implement ground improvement systems, 5) provide and discuss full scale field settlement readings collected for the pilot test section of the project; 6) show the value of instrumentation and field measurement in adjusting and improving design recommendations and in obtaining significant cost savings.

1 PROJECT LOCATION AND DESCRIPTION The overall site consists of approximately 280 hectares of undeveloped land on the southeast side of Panama City in the Republic of Panama. Figure 1 shows the location of the project within Panama City. The overall site is bisected into two parcels, north and south, by the east-west Corredor Sur highway. The southern parcel is the subject of this paper and is approximately 172 hectares in area. It is located in a low lying coastal area that was originally used for farming and pasturing activities. Residential and light commercial development has recently occurred to the west and north of the site, mangroves and the Pacific Ocean are south of the site, and similar undeveloped land is east of the site. The project components consisted of 1) a championship eighteen

hole golf course, 2) numerous low to high-rise residential developments, a town center and country club facilities, 3) artificial lakes, and 4) extensive infrastructure, including internal vehicular roadways, underground utility lines, pump stations, perimeter security and screen walls, etc. The original grades throughout the parcel were relatively flat and typically ranged from el +2 m to +3 m MSL, where a soft saturated marine clay material was typically encountered at the ground surface. The proposed site grade for the project was on average el +6.5 m MSL and because of the difference in height between the original ground surface and the proposed finished grade; up to approximately 4.5 m of fill was required to raise the site grade.


132

Transforming marginal land to support a world class development in Panama

Figure 1. Project Location

These significant areal fills were predicted to cause an estimated settlement of 1 to 2 m due to the compressibility and thickness of the soft clay stratum underlying the project site. Furthermore, these consolidation settlements were estimated to occur over the course of more than 20 years after filling in some cases

Stratum Number

Description

1

Marine Clay

2

Intermixed Clay, Silt, Sand, and Gravel

3

Weathered to Sound Shale and Sandstone – Rock Formation

2 SITE SPECIFIC SUBSURFACE CONDITIONS Thick alluvial soil deposits are present throughout most coastal regions of Panama. In some areas, these layers are granular and consist of poorly graded sand, but in other areas thick layers of marine clays are found. These finer grained soil deposits form in depositional areas of rivers and tributaries creating soft deltaic conditions. Our project site, which is next to Panama Bay, has thick layers of normally consolidated marine clays. These clays are soft and very compressible. The generalized subsurface conditions throughout the development generally consisted of the following inferred strata:

2.1 Marine Clay (Stratum 1) At the majority of the test boring locations, the native marine clay was encountered at the existing ground surface, approximately el +2.5 m. This layer consisted primarily of highly plastic clay with some fractions of silt and sand in localized areas. The thickness of this stratum varied considerably from 2 m to 10 m (averaging about 6 m) and typically trended thicker from north to south. The stratum was generally very soft to soft with SPT N-values ranging from weight of rod (WOR) to less than 5 blows per foot (bpf). Laboratory index property tests were performed for samples taken from this stratum.


LANGONI G. et al.

133

* N=SPT N-value (blows/30 cm)

Figure 2. Typical Subsurface Profile

The natural water contents of the stratum varied typically between 40% and 100%. Liquid limits (LL) ranged between 30% and 110% (typically over 80%) and plasticity indices (PI) ranged between 45% and 90% (typically over 55). Some isolated zones, mostly in the north areas of the project, had a surficial layer of medium stiff clay, with SPT N-values ranging from 5 to over 10 bpf. These were most notably observed along the northeastern border of the site and near the surface at some of the borings along the north of the site adjacent to the Corredor Sur highway. These are likely isolated high areas that have been hardened by desiccation, and are not in the area discussed in this paper. 2.2 Intermixed Clay, Silt, Sand, and Gravel (Stratum 2) Below Stratum 1 in some of the borings, primarily on the eastern portion of the site, a stratum consisting of intermixed clay, silt, sand, and gravel was encountered. This stratum is typical of highly variable, typically more granular deposits found along the meandering banks of existing or ancient rivers. The encountered thickness of this stratum varied considerably from approximately 0.5 to 7 m. This heterogeneous stratum varied considerably from very loose/soft to dense/stiff with STP N-values ranging from 2 to over 100 bpf.

2.3 Weathered to Sound Sedimentary Rock Formation (Stratum 3) Below Strata 1 or 2, a consistent natural rock formation was encountered at elevations ranging around el 0 to el -10, but more consistently between el -5 and el -6.5 in the southern residential development areas. The sedimentary rock formation was identified as Shale and Clayey Sandstone. The rock formation typically has a 1 to 2 m highly weathered and fractured surface followed by a less fractured more intact rock formation, based on the Rock Quality Designation (RQD) values. However, even the highly weathered and fractured upper rock zone is competent with the SPT resistance N-values over 50 bpf to refusal and drill times typically greater than 15 to 20 minutes per foot. Rock Quality Designation (RQD) values on the collected rock cores varied considerably from 0% to 100%. Typically, the RQD for the upper 5 to 10 m of the rock is less than 50%. In addition, unconfined compression tests were performed on selected rock cores. Results of the tests show that the compressive strength of the rock varied from approximately 72 to 318 kg/cm 2 (1,000 to 4,500 psi). 3 KEY GEOTECHNICAL ISSUES For the average clay thickness on-site of 6 m and average fill height of 4.5 meters to reach finished grade, the magnitude of primary settlement was estimated to be in the range of 1 m. The above settle-


134

Transforming Marginal Land to Support a World Class Development in Panama

ments computed were for average conditions, and varied substantially based on clay thickness. For these conditions, the time required for primary settlement to occur (without any ground drainage enhancement) would have theoretically been greater than 20 years which would not be compatible with the construction schedule and phasing of the development. Due to the extensive size of the developable area, the type of soils that had to be improved, and the need for stable structural support of building elements as well as surrounding infrastructure, most ground improvement options were not feasible. Traditional surcharging alone (fill placed above finished grade to accelerate settlement), using two to three meters of fill without wick drains, was evaluated. For this case, the consolidation time would be accelerated somewhat to about 6 to 8 years from the original estimate of about 20 years. For the proposed residential and roadway developments, the estimated surcharge time required to achieve primary settlement would still have been unacceptable and would not have achieved the scheduling milestones for the project. Prefabricated vertical drains (PVD’s or wick drains), installed on relatively close spacings, within the compressible clay were determined to be the best option to accelerate consolidation. Wick drains decrease the drainage path for the pore water within the clay stratum to dissipate during the consolidation process. Once the design fill is placed, a surcharge fill pressure greater than that induced by the proposed surface improvements, can be placed above the finished ground surface so as to simulate the loads of the future construction and “preconsolidate” or “pre-compress” the ground. Once preconsolidated, the ground could theoretically support structural loads, up to the preconsolidation stress, using conventional shallow foundations. 4 GROUND IMPROVEMENT DESIGN The ground improvement system designed for the site based on the constraints previously outlined consisted of wick drains and surcharge. The design details are discussed in the following paragraphs.

Subsequently, we recommended supplemental subsurface exploration be performed and that additional undisturbed samples be obtained in order to perform more consolidation tests and better understand the consolidation parameters. A total of 15 consolidation tests were performed on the undisturbed samples collected from several test borings (identified as LB1, LB1A and LB2). Figure 4 shows the boring location plan with the location of the consolidation tests. Nine of the consolidation tests were performed on traditional horizontally cut samples extracted from the Shelby tubes. The remaining 6 of the 15 tests were performed on undisturbed samples cut along the vertical plane, in an attempt to better understand the horizontal consolidation behavior and potential impact on time-rate of settlement. The following table summarizes the results of the consolidation tests for the site. As shown in the following Table 1, the compression index (Cc) varied from 0.41 to 1.15, with an average value of about 0.80. The coefficient of consolidation value (Cv) ranged from 0.38 to 2.54 m2/year with and average weighted value of about 1.39 m2/year. This average value was greater than the 0.55 m2/year, originally used in our design and allowed for increased spacing of wick drains based on the same time constraints. The horizontal coefficient of consolidation (Ch) was compared to the Cv values, but this did not reveal any apparent behavior trends or relationship. Literature indicates that Ch is typically greater than Cv, but the samples tested showed similar coefficient of consolidation values for both. Hence, we concluded that the relationship between Cv and Ch within the clay matrix at the site is relatively close and without any specific trends. As stated above, results of the additional consolidation testing suggested that an improved coefficient of consolidation (Cv) was appropriate for the design of the ground improvement system. Additional engineering analyses was performed to evaluate the wick drain spacing and surcharge requirements to meet the project goals, specifically a preload surcharge period of 6 months using a coefficient of consolidation (Cv) of 1.39 m2/year. The analysis yielded a revised wick drain spacing of 1.37 m to achieve a 6 month preload time period.

4.1 Wick Drain Spacing At the onset of our involvement, our theoretical analysis of wick drains using the Kjellman-Barron formula suggested a wick spacing of about 1 m in order to reduce the time for the surcharge fill placement period to about 6 to 8 months (180 to 240 days), for an average clay thickness of about 6 meters. This was based on a coefficient of consolidation (Cv) of 0.55 m2/yr, from a previous preliminary geotechnical investigation by others which at that time was the only consolidation test that was performed at the site.

4.2 Surcharge Surcharge needs to be placed in order to appropriately pre-consolidate and pre-stress the underlying soft clays to stress levels above those imposed by the future surface improvements (primarily residential structures and roadways). The surcharge fill required to pre-compress the clay to anticipated post-construction stresses was generally 2 m high (over proposed finished grade) in the residential development areas and 1 m high (over


LANGONI G. et al.

proposed finished grade) in the roadway areas. Appropriately performed, the surcharged ground would allow for normal construction of the future

135

residential structures using conventional shallow foundations and typical construction techniques.

Figure 3. Boring Location Plan.


136


Table 1. Summary of Laboratory Consolidation Tests. Boring ID

Sample

Description of M aterial

Depth (m)

Water Content (%)

LB-2

ST-1A

CH, dark brown CLAY, trace f. sand

0.6

42.6

89

27

LB-2

ST-1A

CH, dark brown CLAY, trace f. sand

0.6

40.5

89

LB-1

ST-1B

CH, gray CLAY, trace f. sand

1.1

48.8

95

LB-2

ST-2B

CH, dark brown CLAY, trace m -f sand

1.7

LB-2

ST-2B

CH, dark brown CLAY, trace c-f sand

ST-3C

CH, gray CLAY, trace m -f sand; clay nodules noted

LB-1A

ST-3C

CH, dark brown CLAY, trace m -f sand; clay nodules noted

LB-1A

ST-4B

LB-2

LL PL Plasticity (%) (%) Index (PI)

Cv (m2 / yr)

Cc

Cr

Cα

62

0.317

0.083

0.001

0.38

27

62

0.300

0.062

0.001

0.44

29

66

0.413

0.090

0.003

0.38

89.1

111 33

78

0.833

0.071

0.008

0.89

1.7

88.1

111 33

78

0.948

0.138

0.009

1.19

2.4

88.9

116 26

90

0.826

0.080

0.008

0.45

2.4

88.6

116 26

90

0.827

0.070

0.008

0.86

CH, gray CLAY

2.8

95.3

100 29

71

1.007

0.106

0.015

0.51

ST-3B

CH, gray CLAY; sand pockets and shell fragm ents noted

3.0

75.6

77

31

46

0.636

0.046

0.009

1.79

LB-2

ST-3B

CH, gray silty CLAY; som e f. sand; num erous sand layers and pockets noted

3.0

45.0

n/a

n/a

n/a

0.378

0.011

0.002 190.67

LB-2

ST-4A

CH, gray CLAY, f. sand; silt lenses noted

4.0

76.2

88

29

59

0.662

0.073

0.009

1.34

LB-1A

ST-5B

CH-OH, gray ORGANIC CLAY, trace f. sand; silt pockets and shell fragm ents noted

4.6

84.5

87

30

57

1.042

0.115

0.011

2.54

LB-1A

ST-5B

CH, gray CLAY, trace f. sand; organic m aterial noted

4.6

96.5

87

30

57

1.148

0.109

0.014

1.54

LB-1

ST-2B

CH, gray CLAY; trace f. sand; shell fragm ents noted

4.8

81.9

85

31

54

0.708

0.086

0.008

0.64

ST-2B

CH, gray CLAY; trace f. sand; silt pockets noted

4.8

88.8

85

31

54

0.721

0.090

0.008

0.63

LB-1A

LB-1

5 FULL SCALE PILOT TEST PROGRAM A full-scale pilot test program was constructed and instrumented at the site to better understand the time-rate consolidation behavior of the underlying compressible strata under varied wick drain spacing and surcharges. 5.1 Test Section Construction of the pilot test area began on 18 April 2008 with the installation of the first wick drains. The pilot test area, shown in Figure 4, was located in a representative area of the project. Test boring L10E was performed within the pilot test area to verify the

1

1

1

1

1

1

thickness of the clay prior to the test program, and indeed it verified a representative clay thickness of 6 m. The test area was divided into six zones to verify the settlement changes from three different wick drain spacings of 1.00 m, 1.37 m, and 1.75 m and the two surcharge heights specified for the project of 1 m and 2 m, for roadways and residential areas respectively. Two settlement plate monitoring devices were installed in each of the varied surcharge and varied wick spacing sections. Hence, a total of twelve settlement plates were installed. In addition, six piezometers were also installed to measure the changes in pore water pressures with time.


LANGONI G. et al.

137

Figure 4. Full Scale Pilot Test Section.

Monitoring of settlement plates installed within the pilot test area was semi-weekly beginning in April 2008 and continuing until February 2010. Figure 6 graphically presents the settlement data on a linear time scale. The graphs show the varied wick spacings (1.37 m, 1.50 m, and 1.75 m) within the 1 meter and 2 meter surcharge zones, respectively.

Figure 7 presents the settlement data on a logarithmic time scale in order to better evaluate when primary consolidation is transitioning into secondary compression or “creep”.


138


Figure 5. Settlement Data on a Linear Time Scale.

Figure 6. Settlement Data on a Logarithmic Time Scale.

Subsequent to surcharge top-out, settlements of approximately 90 cm and 100 cm within the 1 m and 2 m surcharge zones, respectively, were recorded. Readings after 160 to 180 days showed a significantly reduced settlement trend. The point at which the reduced settlement trend begins is characteristic of the completion of primary

consolidation and on-set of the long-term secondary compression phase. The pilot test settlement data obtained closely corroborated the original theoretical estimated settlement values of 90 to 100 cm. Based on the settlement data obtained from the test section, the transition time from primary consolidation to secondary compression (after


LANGONI G. et al.

surcharge top-out) proved to be close to the original estimate of 180 days using a wick drain spacing of 1.37 m. The similar settlement time frame that occurred for the wick drain spacings tested, however, suggested that the drainage properties were affected by the scale of the large test area. Although the cause for this has not been clearly understood it is likely attributed to drainage paths within the soil and wick drains due to the heterogeneous nature of the soil mass under the full scale section. This prompted the construction of a second test area and an adjustment of the wick drain spacing to 1.75 m and 2 m in some areas of the project saving millions of dollars in ground improvement costs. 6 CONCLUSIONS Ground improvement using wick drains and surcharging was successful in pre-consolidating the highly compressible ground in a short time frame in order to allow for construction in about 180 days post surcharge top-out. A full scale field test section proved to be a cost-effective and valuable tool to allow for optimization of wick drain spacing. Wick drain spacing and improvement times calculated by theoretical methods using consolidation data and test borings appear to err on the conservative side. This is likely because the theoretical methods do not account for the effect of thin sandy soil seams within the compressible clay layer that have a higher hydraulic conductivity and enhance the global drainage capability. Full scale field test sections can capture the actual drainage properties of an entire area in a way that cannot be evaluated by a

139

subsurface exploration program. Properly performed, field test sections represent a very minor instrumentation and evaluation cost with a very large economic benefit in reduced ground improvement costs. For the Phase 1 construction, surcharging to above the design loads for the project allowed for successful construction of both the significant site infrastructure as well as a significant number of low rise residential structures on conventional shallow foundations. Long term creep settlements were measured to be on the order of 1 cm for the first year of monitoring. Although the monitoring period was not sufficient to make a definitive statement about long term creep settlement, the readings after the first year showed little signs of creep movement and are expected to decrease with time. Further, the settlements measured by monitoring pins on actual constructed structures were negligible and consistent with results from the full scale test sections. REFERENCES Dhar, A.S., Ameen, S.F. (2001). “Ground Improvement Using Pre-loading with Prefabricated Vertical Drains”. Holtz R. and Kovacs W. (1981). “An Introduction to Geotechnical Engineering”. Lee et. al. (2001). “Soft Soil Engineering”, p.4-8. Terzaghi K., Peck R., and Mesri G. (1996). “Soil Mechanics in Engineering Practice”, Third Edition, p. 100-116.



Technical Committee

TC-214

Author Index Page

Page ARCHABAL Roger ARENAS Fernando AUVINET-GUICHARD Gabriel CHATTE Rémi CHIARABELLI Marco CIRION ARANA Alfredo CUEVAS Alberto DEMING Peter W. GERRESSEN F. HUANG Yanbo IBARRA Enrique LANGONI Gustavo LIN Cheng LIN Guoming LÓPEZ Germán MARINUCCI Antonio MASSOUDI Nasser MENDOZA Manuel J. MOTHERSILLE Devon NIKOLAOU Sissy OKUMUSOGLU Bora OROZCO Marcos PAGLIACCI Federico PANIAGUA Walter PAULÍN AGUIRRE Juan POLETTO Raymond J. PRADEL Daniel

131 85 51 127 19 127 85 39 117 3 77 131 3 3 31 19 93 77 97 39 97 77 67 31 127 39 107

RODRÍGUEZ-REBOLLEDO Juan-Félix RUFIAR Miguel SEGOVIA José SLIWOSKI Richard TAKUMA Takefumi TAMARO George J. TIWARI Binod WILK Charles M.


51 77 31 93 11 39 107 123

3er Simposio Internacional de Cimentaciones Profundas 05-11-2015)

Recommend Documents